gstreamer/gst/gstpad.c

4304 lines
117 KiB
C
Raw Normal View History

/* GStreamer
* Copyright (C) 1999,2000 Erik Walthinsen <omega@cse.ogi.edu>
* 2000 Wim Taymans <wtay@chello.be>
*
* gstpad.c: Pads for linking elements together
*
* This library is free software; you can redistribute it and/or
* modify it under the terms of the GNU Library General Public
* License as published by the Free Software Foundation; either
* version 2 of the License, or (at your option) any later version.
*
* This library is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* Library General Public License for more details.
*
* You should have received a copy of the GNU Library General Public
* License along with this library; if not, write to the
* Free Software Foundation, Inc., 59 Temple Place - Suite 330,
* Boston, MA 02111-1307, USA.
*/
/**
* SECTION:gstpad
* @short_description: Object contained by elements that allows links to
* other elements
* @see_also: #GstPadTemplate, #GstElement, #GstEvent
*
* A #GstElement is linked to other elements via "pads", which are extremely
* light-weight generic link points.
* After two pads are retrieved from an element with gst_element_get_pad(),
* the pads can be link with gst_pad_link(). (For quick links,
* you can also use gst_element_link(), which will make the obvious
* link for you if it's straightforward.)
*
* Pads are typically created from a #GstPadTemplate with
* gst_pad_new_from_template().
*
* Pads have #GstCaps attached to it to describe the media type they are
* capable of dealing with. gst_pad_get_caps() and gst_pad_set_caps() are
* used to manipulate the caps of the pads.
* Pads created from a pad template cannot set capabilities that are
* incompatible with the pad template capabilities.
*
* Pads without pad templates can be created with gst_pad_new(),
* which takes a direction and a name as an argument. If the name is NULL,
* then a guaranteed unique name will be assigned to it.
*
* gst_pad_get_parent() will retrieve the #GstElement that owns the pad.
*
* A #GstElement creating a pad will typically use the various
* gst_pad_set_*_function() calls to register callbacks for various events
* on the pads.
*
* GstElements will use gst_pad_push() and gst_pad_pull_range() to push out
* or pull in a buffer.
*
* To send a #GstEvent on a pad, use gst_pad_send_event() and
* gst_pad_push_event().
*
* Last reviewed on 2006-07-06 (0.10.9)
*/
#include "gst_private.h"
#include "gstpad.h"
#include "gstpadtemplate.h"
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
#include "gstenumtypes.h"
#include "gstmarshal.h"
#include "gstutils.h"
#include "gstinfo.h"
#include "gsterror.h"
#include "gstvalue.h"
#include "glib-compat-private.h"
GST_DEBUG_CATEGORY_STATIC (debug_dataflow);
#define GST_CAT_DEFAULT GST_CAT_PADS
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
/* Pad signals and args */
enum
{
PAD_LINKED,
PAD_UNLINKED,
PAD_REQUEST_LINK,
PAD_HAVE_DATA,
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
/* FILL ME */
LAST_SIGNAL
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
};
enum
{
PAD_PROP_0,
PAD_PROP_CAPS,
PAD_PROP_DIRECTION,
PAD_PROP_TEMPLATE,
/* FILL ME */
};
static void gst_pad_class_init (GstPadClass * klass);
static void gst_pad_init (GstPad * pad);
static void gst_pad_dispose (GObject * object);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
static void gst_pad_finalize (GObject * object);
static void gst_pad_set_property (GObject * object, guint prop_id,
const GValue * value, GParamSpec * pspec);
static void gst_pad_get_property (GObject * object, guint prop_id,
GValue * value, GParamSpec * pspec);
static GstFlowReturn handle_pad_block (GstPad * pad);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
static GstCaps *gst_pad_get_caps_unlocked (GstPad * pad);
static void gst_pad_set_pad_template (GstPad * pad, GstPadTemplate * templ);
gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. Original commit message from CVS: 2005-06-27 Andy Wingo <wingo@pobox.com> * gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. * gst/gsttrace.c (gst_alloc_trace_list_sorted): New helper, returns a sorted copy of the trace list. (gst_alloc_trace_print_live): New API, only prints traces with live objects. Sort the list. (gst_alloc_trace_print_all): Sort the list. (gst_alloc_trace_print): Align columns. * gst/elements/gstttypefindelement.c: * gst/elements/gsttee.c: * gst/base/gstbasesrc.c: * gst/base/gstbasesink.c: * gst/base/gstbasetransform.c: * gst/gstqueue.c: Adapt for pad activation changes. * gst/gstpipeline.c (gst_pipeline_init): Unref after parenting sched. (gst_pipeline_dispose): Drop ref on sched. * gst/gstpad.c (gst_pad_init): Set the default activate func. (gst_pad_activate_default): Push mode by default. (pre_activate_switch, post_activate_switch): New stubs, things to do before and after switching activation modes on pads. (gst_pad_set_active): Take a boolean and not a mode, dispatch to the pad's activate function to choose which mode to activate. Shortcut on deactivation and call the right function directly. (gst_pad_activate_pull): New API, (de)activates a pad in pull mode. (gst_pad_activate_push): New API, same for push mode. (gst_pad_set_activate_function) (gst_pad_set_activatepull_function) (gst_pad_set_activatepush_function): Setters for new API. * gst/gstminiobject.c (gst_mini_object_new, gst_mini_object_free): Trace all miniobjects. (gst_mini_object_make_writable): Unref the arg if we copy, like gst_caps_make_writable. * gst/gstmessage.c (_gst_message_initialize): No trace init. * gst/gstghostpad.c (gst_proxy_pad_do_activate) (gst_proxy_pad_do_activatepull, gst_proxy_pad_do_activatepush): Adapt for new pad API. * gst/gstevent.c (_gst_event_initialize): Don't initialize trace. * gst/gstelement.h: * gst/gstelement.c (gst_element_iterate_src_pads) (gst_element_iterate_sink_pads): New API functions. * gst/gstelement.c (iterator_fold_with_resync): New utility, should fold into gstiterator.c in some form. (gst_element_pads_activate): Simplified via use of fold and delegation of decisions to gstpad->activate. * gst/gstbus.c (gst_bus_source_finalize): Set the bus to NULL, help in debugging. * gst/gstbuffer.c (_gst_buffer_initialize): Ref the buffer type class once in init, like gstmessage. Didn't run into this issue but it seems correct. Don't initialize a trace, gstminiobject does that. * check/pipelines/simple_launch_lines.c (test_stop_from_app): New test, runs fakesrc ! fakesink, stopping on ::handoff via a message to the bus. (assert_live_count): New util function, uses alloc traces to check cleanup. * check/gst/gstghostpad.c (test_ghost_pads): More refcount checks. To be modified when unlink drops the internal pad.
2005-06-27 18:35:05 +00:00
static gboolean gst_pad_activate_default (GstPad * pad);
static gboolean gst_pad_acceptcaps_default (GstPad * pad, GstCaps * caps);
#ifndef GST_DISABLE_LOADSAVE
static xmlNodePtr gst_pad_save_thyself (GstObject * object, xmlNodePtr parent);
#endif
static GstObjectClass *parent_class = NULL;
static guint gst_pad_signals[LAST_SIGNAL] = { 0 };
/* quarks for probe signals */
static GQuark buffer_quark;
static GQuark event_quark;
typedef struct
{
const gint ret;
const gchar *name;
GQuark quark;
} GstFlowQuarks;
static GstFlowQuarks flow_quarks[] = {
{GST_FLOW_CUSTOM_SUCCESS, "custom-success", 0},
{GST_FLOW_RESEND, "resend", 0},
{GST_FLOW_OK, "ok", 0},
{GST_FLOW_NOT_LINKED, "not-linked", 0},
{GST_FLOW_WRONG_STATE, "wrong-state", 0},
{GST_FLOW_UNEXPECTED, "unexpected", 0},
{GST_FLOW_NOT_NEGOTIATED, "not-negotiated", 0},
{GST_FLOW_ERROR, "error", 0},
{GST_FLOW_NOT_SUPPORTED, "not-supported", 0},
{GST_FLOW_CUSTOM_ERROR, "custom-error", 0},
{0, NULL, 0}
};
/**
* gst_flow_get_name:
* @ret: a #GstFlowReturn to get the name of.
*
* Gets a string representing the given flow return.
*
* Returns: a static string with the name of the flow return.
*/
G_CONST_RETURN gchar *
gst_flow_get_name (GstFlowReturn ret)
{
gint i;
ret = CLAMP (ret, GST_FLOW_CUSTOM_ERROR, GST_FLOW_CUSTOM_SUCCESS);
for (i = 0; flow_quarks[i].name; i++) {
if (ret == flow_quarks[i].ret)
return flow_quarks[i].name;
}
return "unknown";
}
/**
* gst_flow_to_quark:
* @ret: a #GstFlowReturn to get the quark of.
*
* Get the unique quark for the given GstFlowReturn.
*
* Returns: the quark associated with the flow return or 0 if an
* invalid return was specified.
*/
GQuark
gst_flow_to_quark (GstFlowReturn ret)
{
gint i;
ret = CLAMP (ret, GST_FLOW_CUSTOM_ERROR, GST_FLOW_CUSTOM_SUCCESS);
for (i = 0; flow_quarks[i].name; i++) {
if (ret == flow_quarks[i].ret)
return flow_quarks[i].quark;
}
return 0;
}
GType
gst_pad_get_type (void)
{
static GType gst_pad_type = 0;
if (G_UNLIKELY (gst_pad_type == 0)) {
static const GTypeInfo pad_info = {
sizeof (GstPadClass), NULL, NULL,
(GClassInitFunc) gst_pad_class_init, NULL, NULL,
sizeof (GstPad),
0,
(GInstanceInitFunc) gst_pad_init, NULL
};
gint i;
gst_pad_type = g_type_register_static (GST_TYPE_OBJECT, "GstPad",
&pad_info, 0);
buffer_quark = g_quark_from_static_string ("buffer");
event_quark = g_quark_from_static_string ("event");
for (i = 0; flow_quarks[i].name; i++) {
flow_quarks[i].quark = g_quark_from_static_string (flow_quarks[i].name);
}
GST_DEBUG_CATEGORY_INIT (debug_dataflow, "GST_DATAFLOW",
GST_DEBUG_BOLD | GST_DEBUG_FG_GREEN, "dataflow inside pads");
}
return gst_pad_type;
}
static gboolean
_gst_do_pass_data_accumulator (GSignalInvocationHint * ihint,
GValue * return_accu, const GValue * handler_return, gpointer dummy)
{
gboolean ret = g_value_get_boolean (handler_return);
GST_DEBUG ("accumulated %d", ret);
g_value_set_boolean (return_accu, ret);
return ret;
}
static gboolean
default_have_data (GstPad * pad, GstMiniObject * o)
{
return TRUE;
}
static void
gst_pad_class_init (GstPadClass * klass)
{
GObjectClass *gobject_class;
GstObjectClass *gstobject_class;
gobject_class = G_OBJECT_CLASS (klass);
gstobject_class = GST_OBJECT_CLASS (klass);
parent_class = g_type_class_peek_parent (klass);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
gobject_class->dispose = GST_DEBUG_FUNCPTR (gst_pad_dispose);
gobject_class->finalize = GST_DEBUG_FUNCPTR (gst_pad_finalize);
gobject_class->set_property = GST_DEBUG_FUNCPTR (gst_pad_set_property);
gobject_class->get_property = GST_DEBUG_FUNCPTR (gst_pad_get_property);
/**
* GstPad::linked:
* @pad: the pad that emitted the signal
* @peer: the peer pad that has been connected
*
* Signals that a pad has been linked to the peer pad.
*/
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
gst_pad_signals[PAD_LINKED] =
g_signal_new ("linked", G_TYPE_FROM_CLASS (klass), G_SIGNAL_RUN_LAST,
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
G_STRUCT_OFFSET (GstPadClass, linked), NULL, NULL,
gst_marshal_VOID__OBJECT, G_TYPE_NONE, 1, GST_TYPE_PAD);
/**
* GstPad::unlinked:
* @pad: the pad that emitted the signal
* @peer: the peer pad that has been disconnected
*
* Signals that a pad has been unlinked from the peer pad.
*/
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
gst_pad_signals[PAD_UNLINKED] =
g_signal_new ("unlinked", G_TYPE_FROM_CLASS (klass), G_SIGNAL_RUN_LAST,
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
G_STRUCT_OFFSET (GstPadClass, unlinked), NULL, NULL,
gst_marshal_VOID__OBJECT, G_TYPE_NONE, 1, GST_TYPE_PAD);
/**
* GstPad::request-link:
* @pad: the pad that emitted the signal
* @peer: the peer pad for which a connection is requested
*
* Signals that a pad connection has been requested.
*/
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
gst_pad_signals[PAD_REQUEST_LINK] =
g_signal_new ("request-link", G_TYPE_FROM_CLASS (klass),
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
G_SIGNAL_RUN_LAST, G_STRUCT_OFFSET (GstPadClass, request_link), NULL,
NULL, gst_marshal_VOID__OBJECT, G_TYPE_NONE, 0);
/**
* GstPad::have-data:
* @pad: the pad that emitted the signal
* @mini_obj: new data
*
* Signals that new data is available on the pad. This signal is used
* internally for implementing pad probes.
* See gst_pad_add_*_probe functions.
*
* Returns: %TRUE to keep the data, %FALSE to drop it
*/
gst_pad_signals[PAD_HAVE_DATA] =
g_signal_new ("have-data", G_TYPE_FROM_CLASS (klass),
G_SIGNAL_RUN_LAST | G_SIGNAL_DETAILED,
G_STRUCT_OFFSET (GstPadClass, have_data),
_gst_do_pass_data_accumulator,
NULL, gst_marshal_BOOLEAN__POINTER, G_TYPE_BOOLEAN, 1,
GST_TYPE_MINI_OBJECT);
g_object_class_install_property (gobject_class, PAD_PROP_CAPS,
g_param_spec_boxed ("caps", "Caps", "The capabilities of the pad",
GST_TYPE_CAPS, G_PARAM_READABLE));
g_object_class_install_property (gobject_class, PAD_PROP_DIRECTION,
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
g_param_spec_enum ("direction", "Direction", "The direction of the pad",
GST_TYPE_PAD_DIRECTION, GST_PAD_UNKNOWN,
G_PARAM_READWRITE | G_PARAM_CONSTRUCT_ONLY));
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
/* FIXME, Make G_PARAM_CONSTRUCT_ONLY when we fix ghostpads. */
g_object_class_install_property (gobject_class, PAD_PROP_TEMPLATE,
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
g_param_spec_object ("template", "Template",
"The GstPadTemplate of this pad", GST_TYPE_PAD_TEMPLATE,
G_PARAM_READWRITE));
#ifndef GST_DISABLE_LOADSAVE
gstobject_class->save_thyself = GST_DEBUG_FUNCPTR (gst_pad_save_thyself);
#endif
gstobject_class->path_string_separator = ".";
klass->have_data = default_have_data;
}
static void
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
gst_pad_init (GstPad * pad)
{
GST_PAD_DIRECTION (pad) = GST_PAD_UNKNOWN;
GST_PAD_PEER (pad) = NULL;
WARNING: Don't grab this updated unless you're really, REALLY sure. Original commit message from CVS: WARNING: Don't grab this updated unless you're really, REALLY sure. WARNING: Wait for the next one. Whole lotta changes here, including a few random bits: examples/*/Makefile: updated to use `libtool gcc`, not just `gcc` gst/ gstbuffer.h: updated to new flag style gst.c, gstdebug.h: added new debugging for function ptrs gstpipeline.c: set type of parent_class to the class, not the object gstthread.c: ditto plugins/ cdparanoia/cdparanoia.c: added an argument type, updated some defaults cobin/spindentity.c: updated to new do/while loopfunction style mp3encode/lame/gstlame.c: argument types, whole lotta lame options tests/: various changes Now, for the big changes: Once again, the scheduling system has changed. And once again, it broke a whole bunch of things. The gist of the change is that there is now a function pointer for gst_pad_push and gst_pad_pull, instead of a hard-wired function. Well, currently they are functions, but that's for debugging purposes only, they just call the function pointer after spewing lots of DEBUG(). This changed the GstPad structure a bit, and the GstPad API as well. Where elements used to provide chain() and pull() functions, they provide chain() and get() functions. gst_pad_set_pull[region]_function has been changed to get_pad_set_get[region]_function. This means all the elements out there that used to have pull functions need to be updated. The calls to that function have been changed in the normal elements, but the names of the functions passed is still _pull[region](), which is an aesthetic issue more than anything. As for what doesn't work yet, just about anything dealing with Connections is hosed, meaning threaded stuff won't work. This will be fixed about 12 hours from now, after I've slept, etc. The simplefake.c test works in both cothreaded and chained cases, but not much else will work due to the Connection problem. Needless to say, don't grab this unless you *need* these features *now*, else wait to update this stuff until tomorrow. I'm going to sleep now.
2000-12-16 10:18:09 +00:00
GST_PAD_CHAINFUNC (pad) = NULL;
WARNING: Don't grab this updated unless you're really, REALLY sure. Original commit message from CVS: WARNING: Don't grab this updated unless you're really, REALLY sure. WARNING: Wait for the next one. Whole lotta changes here, including a few random bits: examples/*/Makefile: updated to use `libtool gcc`, not just `gcc` gst/ gstbuffer.h: updated to new flag style gst.c, gstdebug.h: added new debugging for function ptrs gstpipeline.c: set type of parent_class to the class, not the object gstthread.c: ditto plugins/ cdparanoia/cdparanoia.c: added an argument type, updated some defaults cobin/spindentity.c: updated to new do/while loopfunction style mp3encode/lame/gstlame.c: argument types, whole lotta lame options tests/: various changes Now, for the big changes: Once again, the scheduling system has changed. And once again, it broke a whole bunch of things. The gist of the change is that there is now a function pointer for gst_pad_push and gst_pad_pull, instead of a hard-wired function. Well, currently they are functions, but that's for debugging purposes only, they just call the function pointer after spewing lots of DEBUG(). This changed the GstPad structure a bit, and the GstPad API as well. Where elements used to provide chain() and pull() functions, they provide chain() and get() functions. gst_pad_set_pull[region]_function has been changed to get_pad_set_get[region]_function. This means all the elements out there that used to have pull functions need to be updated. The calls to that function have been changed in the normal elements, but the names of the functions passed is still _pull[region](), which is an aesthetic issue more than anything. As for what doesn't work yet, just about anything dealing with Connections is hosed, meaning threaded stuff won't work. This will be fixed about 12 hours from now, after I've slept, etc. The simplefake.c test works in both cothreaded and chained cases, but not much else will work due to the Connection problem. Needless to say, don't grab this unless you *need* these features *now*, else wait to update this stuff until tomorrow. I'm going to sleep now.
2000-12-16 10:18:09 +00:00
GST_PAD_LINKFUNC (pad) = NULL;
GST_PAD_CAPS (pad) = NULL;
GST_PAD_GETCAPSFUNC (pad) = NULL;
GST_PAD_ACTIVATEFUNC (pad) = GST_DEBUG_FUNCPTR (gst_pad_activate_default);
GST_PAD_EVENTFUNC (pad) = GST_DEBUG_FUNCPTR (gst_pad_event_default);
GST_PAD_QUERYTYPEFUNC (pad) =
GST_DEBUG_FUNCPTR (gst_pad_get_query_types_default);
GST_PAD_QUERYFUNC (pad) = GST_DEBUG_FUNCPTR (gst_pad_query_default);
GST_PAD_INTLINKFUNC (pad) =
GST_DEBUG_FUNCPTR (gst_pad_get_internal_links_default);
GST_PAD_ACCEPTCAPSFUNC (pad) = GST_DEBUG_FUNCPTR (gst_pad_acceptcaps_default);
pad->do_buffer_signals = 0;
pad->do_event_signals = 0;
#if 0
GST_PAD_SET_FLUSHING (pad);
#else
/* FIXME, should be set flushing initially, see #339326 */
GST_PAD_UNSET_FLUSHING (pad);
#endif
pad->preroll_lock = g_mutex_new ();
pad->preroll_cond = g_cond_new ();
pad->stream_rec_lock = g_new (GStaticRecMutex, 1);
g_static_rec_mutex_init (pad->stream_rec_lock);
pad->block_cond = g_cond_new ();
}
static void
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
gst_pad_dispose (GObject * object)
{
GstPad *pad = GST_PAD (object);
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_DEBUG_OBJECT (GST_CAT_REFCOUNTING, pad, "dispose");
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
/* we don't hold a ref to the peer so we can just set the
* peer to NULL. */
GST_PAD_PEER (pad) = NULL;
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
/* clear the caps */
gst_caps_replace (&GST_PAD_CAPS (pad), NULL);
gst_pad_set_pad_template (pad, NULL);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
G_OBJECT_CLASS (parent_class)->dispose (object);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
}
static void
gst_pad_finalize (GObject * object)
{
GstPad *pad = GST_PAD (object);
GstTask *task;
/* in case the task is still around, clean it up */
if ((task = GST_PAD_TASK (pad))) {
gst_task_join (task);
GST_PAD_TASK (pad) = NULL;
gst_object_unref (task);
}
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
if (pad->stream_rec_lock) {
g_static_rec_mutex_free (pad->stream_rec_lock);
g_free (pad->stream_rec_lock);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
pad->stream_rec_lock = NULL;
}
if (pad->preroll_lock) {
g_mutex_free (pad->preroll_lock);
g_cond_free (pad->preroll_cond);
pad->preroll_lock = NULL;
pad->preroll_cond = NULL;
}
if (pad->block_cond) {
g_cond_free (pad->block_cond);
pad->block_cond = NULL;
}
G_OBJECT_CLASS (parent_class)->finalize (object);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
}
static void
gst_pad_set_property (GObject * object, guint prop_id,
const GValue * value, GParamSpec * pspec)
{
g_return_if_fail (GST_IS_PAD (object));
switch (prop_id) {
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
case PAD_PROP_DIRECTION:
GST_PAD_DIRECTION (object) = g_value_get_enum (value);
break;
case PAD_PROP_TEMPLATE:
gst_pad_set_pad_template (GST_PAD_CAST (object),
(GstPadTemplate *) g_value_get_object (value));
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
break;
default:
G_OBJECT_WARN_INVALID_PROPERTY_ID (object, prop_id, pspec);
break;
}
}
static void
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
gst_pad_get_property (GObject * object, guint prop_id,
GValue * value, GParamSpec * pspec)
{
g_return_if_fail (GST_IS_PAD (object));
switch (prop_id) {
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
case PAD_PROP_CAPS:
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_OBJECT_LOCK (object);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
g_value_set_boxed (value, GST_PAD_CAPS (object));
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_OBJECT_UNLOCK (object);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
break;
case PAD_PROP_DIRECTION:
g_value_set_enum (value, GST_PAD_DIRECTION (object));
break;
case PAD_PROP_TEMPLATE:
g_value_set_object (value, GST_PAD_PAD_TEMPLATE (object));
break;
default:
G_OBJECT_WARN_INVALID_PROPERTY_ID (object, prop_id, pspec);
break;
}
}
/**
* gst_pad_new:
* @name: the name of the new pad.
* @direction: the #GstPadDirection of the pad.
*
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
* Creates a new pad with the given name in the given direction.
* If name is NULL, a guaranteed unique name (across all pads)
* will be assigned.
* This function makes a copy of the name so you can safely free the name.
*
* Returns: a new #GstPad, or NULL in case of an error.
*
* MT safe.
*/
GstPad *
gst_pad_new (const gchar * name, GstPadDirection direction)
{
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
return g_object_new (GST_TYPE_PAD,
"name", name, "direction", direction, NULL);
}
/**
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
* gst_pad_new_from_template:
* @templ: the pad template to use
* @name: the name of the element
*
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
* Creates a new pad with the given name from the given template.
* If name is NULL, a guaranteed unique name (across all pads)
* will be assigned.
* This function makes a copy of the name so you can safely free the name.
*
* Returns: a new #GstPad, or NULL in case of an error.
*/
GstPad *
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
gst_pad_new_from_template (GstPadTemplate * templ, const gchar * name)
{
gst/gstelement.c (gst_element_dispose): Protect against multiple invocations. Original commit message from CVS: 2004-02-24 Andy Wingo <wingo@pobox.com> * gst/gstelement.c (gst_element_dispose): Protect against multiple invocations. * gst/schedulers/gstoptimalscheduler.c I added a mess of prototypes at the top of the file by way of documentation. Some of the operations on chains and groups were re-organized. (create_group): Added a type argument so if the group is enabled, the setup_group_scheduler knows what to do. (group_elements): Added a type argument here, too, to be passed on to create_group. (group_element_set_enabled): If an unlinked PLAYING element is added to a bin, we have to create a new group to hold the element, and this function will be called before the group is added to the chain. Thus we have a valid case for group->chain==NULL. Instead of calling chain_group_set_enabled, just set the flag on the group (the chain's status will be set when the group is added to it). (gst_opt_scheduler_state_transition, chain_group_set_enabled): Setup the group scheduler when the group is enabled, not specifically when an element goes PAUSED->PLAYING. This means PLAYING elements can be added, linked, and scheduled into a PLAYING pipeline, as was intended. (add_to_group): Don't ref the group twice. I don't know when this double-ref got in here. Removing it has the potential to cause segfaults if other parts of the scheduler are buggy. If you find that the scheduler is segfaulting for you, put in an extra ref here and see if that hacks over the underlying issue. Of course, then find out what code is unreffing a group it doesn't own... (create_group): Make the extra refcount floating, and remove it after adding the element. This means that... (unref_group): Destroy when the refcount reaches 0, not 1, like every other refcounted object in the known universe. (remove_from_group): When a group becomes empty, set it to be not active, and remove it from its chain. Don't unref it again, there's no floating reference any more. (destroy_group): We have to remove the group from the chain in remove_from_group (rather than here) to break refcounting cycles (the chain always has a ref on the group). So assert that group->chain==NULL. (ref_group_by_count): Removed, it was commented out anyway. (merge_chains): Use the remove_from_chain and add_to_chain primitives to do the reparenting, instead of rolling our own implementation. (add_to_chain): The first non-disabled group in the chain's group list will be the entry point for the chain. Because buffers can accumulate in loop elements' peer bufpens, we preferentially schedule loop groups before get groups to avoid unnecessary execution of get-based groups when the bufpens are already full. (gst_opt_scheduler_schedule_run_queue): Debug fixes. (get_group_schedule_function): Ditto. (loop_group_schedule_function): Ditto. (gst_opt_scheduler_loop_wrapper): Ditto. (gst_opt_scheduler_iterate): Ditto. I understand the opt scheduler now, yippee! * gst/gstpad.c: All throughout, added FIXMEs to look at for 0.9. (gst_pad_get_name, gst_pad_set_chain_function) (gst_pad_set_get_function, gst_pad_set_event_function) (gst_pad_set_event_mask_function, gst_pad_get_event_masks) (gst_pad_get_event_masks_default, gst_pad_set_convert_function) (gst_pad_set_query_function, gst_pad_get_query_types) (gst_pad_get_query_types_default) (gst_pad_set_internal_link_function) (gst_pad_set_formats_function, gst_pad_set_link_function) (gst_pad_set_fixate_function, gst_pad_set_getcaps_function) (gst_pad_set_bufferalloc_function, gst_pad_unlink) (gst_pad_renegotiate, gst_pad_set_parent, gst_pad_get_parent) (gst_pad_add_ghost_pad, gst_pad_proxy_getcaps) (gst_pad_proxy_pad_link, gst_pad_proxy_fixate) (gst_pad_get_pad_template_caps, gst_pad_check_compatibility) (gst_pad_get_peer, gst_pad_get_allowed_caps) (gst_pad_alloc_buffer, gst_pad_push, gst_pad_pull) (gst_pad_selectv, gst_pad_select, gst_pad_template_get_caps) (gst_pad_event_default_dispatch, gst_pad_event_default) (gst_pad_dispatcher, gst_pad_send_event, gst_pad_convert_default) (gst_pad_convert, gst_pad_query_default, gst_pad_query) (gst_pad_get_formats_default, gst_pad_get_formats): Better argument checks, and some doc fixes. (gst_pad_custom_new_from_template): Um, does anyone use these functions? Actually make a custom pad instead of a normal one. (gst_pad_try_set_caps): Transpose some checks. (gst_pad_try_set_caps_nonfixed): Same, and use a macro to check if the pad is in negotiation. (gst_pad_try_relink_filtered): Use pad_link_prepare. * gst/gstelement.c: Remove prototypes also defined in gstclock.h. * gst/gstelement.h: * gst/gstclock.h: Un-deprecate the old clocking API, as discussed on the list.
2004-02-25 13:16:12 +00:00
g_return_val_if_fail (GST_IS_PAD_TEMPLATE (templ), NULL);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
return g_object_new (GST_TYPE_PAD,
"name", name, "direction", templ->direction, "template", templ, NULL);
}
/**
* gst_pad_new_from_static_template:
* @templ: the #GstStaticPadTemplate to use
* @name: the name of the element
*
* Creates a new pad with the given name from the given static template.
* If name is NULL, a guaranteed unique name (across all pads)
* will be assigned.
* This function makes a copy of the name so you can safely free the name.
*
* Returns: a new #GstPad, or NULL in case of an error.
*/
GstPad *
gst_pad_new_from_static_template (GstStaticPadTemplate * templ,
const gchar * name)
{
GstPad *pad;
GstPadTemplate *template;
template = gst_static_pad_template_get (templ);
pad = gst_pad_new_from_template (template, name);
gst_object_unref (template);
return pad;
}
/**
* gst_pad_get_direction:
* @pad: a #GstPad to get the direction of.
*
* Gets the direction of the pad. The direction of the pad is
* decided at construction time so this function does not take
* the LOCK.
*
* Returns: the #GstPadDirection of the pad.
*
* MT safe.
*/
GstPadDirection
gst_pad_get_direction (GstPad * pad)
{
GstPadDirection result;
/* PAD_UNKNOWN is a little silly but we need some sort of
* error return value */
g_return_val_if_fail (GST_IS_PAD (pad), GST_PAD_UNKNOWN);
GST_OBJECT_LOCK (pad);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
result = GST_PAD_DIRECTION (pad);
GST_OBJECT_UNLOCK (pad);
return result;
}
gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. Original commit message from CVS: 2005-06-27 Andy Wingo <wingo@pobox.com> * gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. * gst/gsttrace.c (gst_alloc_trace_list_sorted): New helper, returns a sorted copy of the trace list. (gst_alloc_trace_print_live): New API, only prints traces with live objects. Sort the list. (gst_alloc_trace_print_all): Sort the list. (gst_alloc_trace_print): Align columns. * gst/elements/gstttypefindelement.c: * gst/elements/gsttee.c: * gst/base/gstbasesrc.c: * gst/base/gstbasesink.c: * gst/base/gstbasetransform.c: * gst/gstqueue.c: Adapt for pad activation changes. * gst/gstpipeline.c (gst_pipeline_init): Unref after parenting sched. (gst_pipeline_dispose): Drop ref on sched. * gst/gstpad.c (gst_pad_init): Set the default activate func. (gst_pad_activate_default): Push mode by default. (pre_activate_switch, post_activate_switch): New stubs, things to do before and after switching activation modes on pads. (gst_pad_set_active): Take a boolean and not a mode, dispatch to the pad's activate function to choose which mode to activate. Shortcut on deactivation and call the right function directly. (gst_pad_activate_pull): New API, (de)activates a pad in pull mode. (gst_pad_activate_push): New API, same for push mode. (gst_pad_set_activate_function) (gst_pad_set_activatepull_function) (gst_pad_set_activatepush_function): Setters for new API. * gst/gstminiobject.c (gst_mini_object_new, gst_mini_object_free): Trace all miniobjects. (gst_mini_object_make_writable): Unref the arg if we copy, like gst_caps_make_writable. * gst/gstmessage.c (_gst_message_initialize): No trace init. * gst/gstghostpad.c (gst_proxy_pad_do_activate) (gst_proxy_pad_do_activatepull, gst_proxy_pad_do_activatepush): Adapt for new pad API. * gst/gstevent.c (_gst_event_initialize): Don't initialize trace. * gst/gstelement.h: * gst/gstelement.c (gst_element_iterate_src_pads) (gst_element_iterate_sink_pads): New API functions. * gst/gstelement.c (iterator_fold_with_resync): New utility, should fold into gstiterator.c in some form. (gst_element_pads_activate): Simplified via use of fold and delegation of decisions to gstpad->activate. * gst/gstbus.c (gst_bus_source_finalize): Set the bus to NULL, help in debugging. * gst/gstbuffer.c (_gst_buffer_initialize): Ref the buffer type class once in init, like gstmessage. Didn't run into this issue but it seems correct. Don't initialize a trace, gstminiobject does that. * check/pipelines/simple_launch_lines.c (test_stop_from_app): New test, runs fakesrc ! fakesink, stopping on ::handoff via a message to the bus. (assert_live_count): New util function, uses alloc traces to check cleanup. * check/gst/gstghostpad.c (test_ghost_pads): More refcount checks. To be modified when unlink drops the internal pad.
2005-06-27 18:35:05 +00:00
static gboolean
gst_pad_activate_default (GstPad * pad)
{
return gst_pad_activate_push (pad, TRUE);
}
static void
pre_activate (GstPad * pad, GstActivateMode new_mode)
gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. Original commit message from CVS: 2005-06-27 Andy Wingo <wingo@pobox.com> * gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. * gst/gsttrace.c (gst_alloc_trace_list_sorted): New helper, returns a sorted copy of the trace list. (gst_alloc_trace_print_live): New API, only prints traces with live objects. Sort the list. (gst_alloc_trace_print_all): Sort the list. (gst_alloc_trace_print): Align columns. * gst/elements/gstttypefindelement.c: * gst/elements/gsttee.c: * gst/base/gstbasesrc.c: * gst/base/gstbasesink.c: * gst/base/gstbasetransform.c: * gst/gstqueue.c: Adapt for pad activation changes. * gst/gstpipeline.c (gst_pipeline_init): Unref after parenting sched. (gst_pipeline_dispose): Drop ref on sched. * gst/gstpad.c (gst_pad_init): Set the default activate func. (gst_pad_activate_default): Push mode by default. (pre_activate_switch, post_activate_switch): New stubs, things to do before and after switching activation modes on pads. (gst_pad_set_active): Take a boolean and not a mode, dispatch to the pad's activate function to choose which mode to activate. Shortcut on deactivation and call the right function directly. (gst_pad_activate_pull): New API, (de)activates a pad in pull mode. (gst_pad_activate_push): New API, same for push mode. (gst_pad_set_activate_function) (gst_pad_set_activatepull_function) (gst_pad_set_activatepush_function): Setters for new API. * gst/gstminiobject.c (gst_mini_object_new, gst_mini_object_free): Trace all miniobjects. (gst_mini_object_make_writable): Unref the arg if we copy, like gst_caps_make_writable. * gst/gstmessage.c (_gst_message_initialize): No trace init. * gst/gstghostpad.c (gst_proxy_pad_do_activate) (gst_proxy_pad_do_activatepull, gst_proxy_pad_do_activatepush): Adapt for new pad API. * gst/gstevent.c (_gst_event_initialize): Don't initialize trace. * gst/gstelement.h: * gst/gstelement.c (gst_element_iterate_src_pads) (gst_element_iterate_sink_pads): New API functions. * gst/gstelement.c (iterator_fold_with_resync): New utility, should fold into gstiterator.c in some form. (gst_element_pads_activate): Simplified via use of fold and delegation of decisions to gstpad->activate. * gst/gstbus.c (gst_bus_source_finalize): Set the bus to NULL, help in debugging. * gst/gstbuffer.c (_gst_buffer_initialize): Ref the buffer type class once in init, like gstmessage. Didn't run into this issue but it seems correct. Don't initialize a trace, gstminiobject does that. * check/pipelines/simple_launch_lines.c (test_stop_from_app): New test, runs fakesrc ! fakesink, stopping on ::handoff via a message to the bus. (assert_live_count): New util function, uses alloc traces to check cleanup. * check/gst/gstghostpad.c (test_ghost_pads): More refcount checks. To be modified when unlink drops the internal pad.
2005-06-27 18:35:05 +00:00
{
switch (new_mode) {
case GST_ACTIVATE_PUSH:
case GST_ACTIVATE_PULL:
GST_OBJECT_LOCK (pad);
GST_DEBUG_OBJECT (pad, "setting ACTIVATE_MODE %d, unset flushing",
new_mode);
GST_PAD_UNSET_FLUSHING (pad);
GST_PAD_ACTIVATE_MODE (pad) = new_mode;
GST_OBJECT_UNLOCK (pad);
break;
case GST_ACTIVATE_NONE:
GST_OBJECT_LOCK (pad);
GST_DEBUG_OBJECT (pad, "setting ACTIVATE_MODE NONE, set flushing");
GST_PAD_SET_FLUSHING (pad);
/* unlock blocked pads so element can resume and stop */
GST_PAD_BLOCK_BROADCAST (pad);
GST_OBJECT_UNLOCK (pad);
break;
gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. Original commit message from CVS: 2005-06-27 Andy Wingo <wingo@pobox.com> * gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. * gst/gsttrace.c (gst_alloc_trace_list_sorted): New helper, returns a sorted copy of the trace list. (gst_alloc_trace_print_live): New API, only prints traces with live objects. Sort the list. (gst_alloc_trace_print_all): Sort the list. (gst_alloc_trace_print): Align columns. * gst/elements/gstttypefindelement.c: * gst/elements/gsttee.c: * gst/base/gstbasesrc.c: * gst/base/gstbasesink.c: * gst/base/gstbasetransform.c: * gst/gstqueue.c: Adapt for pad activation changes. * gst/gstpipeline.c (gst_pipeline_init): Unref after parenting sched. (gst_pipeline_dispose): Drop ref on sched. * gst/gstpad.c (gst_pad_init): Set the default activate func. (gst_pad_activate_default): Push mode by default. (pre_activate_switch, post_activate_switch): New stubs, things to do before and after switching activation modes on pads. (gst_pad_set_active): Take a boolean and not a mode, dispatch to the pad's activate function to choose which mode to activate. Shortcut on deactivation and call the right function directly. (gst_pad_activate_pull): New API, (de)activates a pad in pull mode. (gst_pad_activate_push): New API, same for push mode. (gst_pad_set_activate_function) (gst_pad_set_activatepull_function) (gst_pad_set_activatepush_function): Setters for new API. * gst/gstminiobject.c (gst_mini_object_new, gst_mini_object_free): Trace all miniobjects. (gst_mini_object_make_writable): Unref the arg if we copy, like gst_caps_make_writable. * gst/gstmessage.c (_gst_message_initialize): No trace init. * gst/gstghostpad.c (gst_proxy_pad_do_activate) (gst_proxy_pad_do_activatepull, gst_proxy_pad_do_activatepush): Adapt for new pad API. * gst/gstevent.c (_gst_event_initialize): Don't initialize trace. * gst/gstelement.h: * gst/gstelement.c (gst_element_iterate_src_pads) (gst_element_iterate_sink_pads): New API functions. * gst/gstelement.c (iterator_fold_with_resync): New utility, should fold into gstiterator.c in some form. (gst_element_pads_activate): Simplified via use of fold and delegation of decisions to gstpad->activate. * gst/gstbus.c (gst_bus_source_finalize): Set the bus to NULL, help in debugging. * gst/gstbuffer.c (_gst_buffer_initialize): Ref the buffer type class once in init, like gstmessage. Didn't run into this issue but it seems correct. Don't initialize a trace, gstminiobject does that. * check/pipelines/simple_launch_lines.c (test_stop_from_app): New test, runs fakesrc ! fakesink, stopping on ::handoff via a message to the bus. (assert_live_count): New util function, uses alloc traces to check cleanup. * check/gst/gstghostpad.c (test_ghost_pads): More refcount checks. To be modified when unlink drops the internal pad.
2005-06-27 18:35:05 +00:00
}
}
static void
post_activate (GstPad * pad, GstActivateMode new_mode)
gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. Original commit message from CVS: 2005-06-27 Andy Wingo <wingo@pobox.com> * gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. * gst/gsttrace.c (gst_alloc_trace_list_sorted): New helper, returns a sorted copy of the trace list. (gst_alloc_trace_print_live): New API, only prints traces with live objects. Sort the list. (gst_alloc_trace_print_all): Sort the list. (gst_alloc_trace_print): Align columns. * gst/elements/gstttypefindelement.c: * gst/elements/gsttee.c: * gst/base/gstbasesrc.c: * gst/base/gstbasesink.c: * gst/base/gstbasetransform.c: * gst/gstqueue.c: Adapt for pad activation changes. * gst/gstpipeline.c (gst_pipeline_init): Unref after parenting sched. (gst_pipeline_dispose): Drop ref on sched. * gst/gstpad.c (gst_pad_init): Set the default activate func. (gst_pad_activate_default): Push mode by default. (pre_activate_switch, post_activate_switch): New stubs, things to do before and after switching activation modes on pads. (gst_pad_set_active): Take a boolean and not a mode, dispatch to the pad's activate function to choose which mode to activate. Shortcut on deactivation and call the right function directly. (gst_pad_activate_pull): New API, (de)activates a pad in pull mode. (gst_pad_activate_push): New API, same for push mode. (gst_pad_set_activate_function) (gst_pad_set_activatepull_function) (gst_pad_set_activatepush_function): Setters for new API. * gst/gstminiobject.c (gst_mini_object_new, gst_mini_object_free): Trace all miniobjects. (gst_mini_object_make_writable): Unref the arg if we copy, like gst_caps_make_writable. * gst/gstmessage.c (_gst_message_initialize): No trace init. * gst/gstghostpad.c (gst_proxy_pad_do_activate) (gst_proxy_pad_do_activatepull, gst_proxy_pad_do_activatepush): Adapt for new pad API. * gst/gstevent.c (_gst_event_initialize): Don't initialize trace. * gst/gstelement.h: * gst/gstelement.c (gst_element_iterate_src_pads) (gst_element_iterate_sink_pads): New API functions. * gst/gstelement.c (iterator_fold_with_resync): New utility, should fold into gstiterator.c in some form. (gst_element_pads_activate): Simplified via use of fold and delegation of decisions to gstpad->activate. * gst/gstbus.c (gst_bus_source_finalize): Set the bus to NULL, help in debugging. * gst/gstbuffer.c (_gst_buffer_initialize): Ref the buffer type class once in init, like gstmessage. Didn't run into this issue but it seems correct. Don't initialize a trace, gstminiobject does that. * check/pipelines/simple_launch_lines.c (test_stop_from_app): New test, runs fakesrc ! fakesink, stopping on ::handoff via a message to the bus. (assert_live_count): New util function, uses alloc traces to check cleanup. * check/gst/gstghostpad.c (test_ghost_pads): More refcount checks. To be modified when unlink drops the internal pad.
2005-06-27 18:35:05 +00:00
{
switch (new_mode) {
case GST_ACTIVATE_PUSH:
case GST_ACTIVATE_PULL:
/* nop */
break;
case GST_ACTIVATE_NONE:
/* ensures that streaming stops */
GST_PAD_STREAM_LOCK (pad);
/* while we're at it set activation mode */
GST_OBJECT_LOCK (pad);
GST_DEBUG_OBJECT (pad, "setting ACTIVATE_MODE %d", new_mode);
GST_PAD_ACTIVATE_MODE (pad) = new_mode;
GST_OBJECT_UNLOCK (pad);
GST_PAD_STREAM_UNLOCK (pad);
break;
gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. Original commit message from CVS: 2005-06-27 Andy Wingo <wingo@pobox.com> * gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. * gst/gsttrace.c (gst_alloc_trace_list_sorted): New helper, returns a sorted copy of the trace list. (gst_alloc_trace_print_live): New API, only prints traces with live objects. Sort the list. (gst_alloc_trace_print_all): Sort the list. (gst_alloc_trace_print): Align columns. * gst/elements/gstttypefindelement.c: * gst/elements/gsttee.c: * gst/base/gstbasesrc.c: * gst/base/gstbasesink.c: * gst/base/gstbasetransform.c: * gst/gstqueue.c: Adapt for pad activation changes. * gst/gstpipeline.c (gst_pipeline_init): Unref after parenting sched. (gst_pipeline_dispose): Drop ref on sched. * gst/gstpad.c (gst_pad_init): Set the default activate func. (gst_pad_activate_default): Push mode by default. (pre_activate_switch, post_activate_switch): New stubs, things to do before and after switching activation modes on pads. (gst_pad_set_active): Take a boolean and not a mode, dispatch to the pad's activate function to choose which mode to activate. Shortcut on deactivation and call the right function directly. (gst_pad_activate_pull): New API, (de)activates a pad in pull mode. (gst_pad_activate_push): New API, same for push mode. (gst_pad_set_activate_function) (gst_pad_set_activatepull_function) (gst_pad_set_activatepush_function): Setters for new API. * gst/gstminiobject.c (gst_mini_object_new, gst_mini_object_free): Trace all miniobjects. (gst_mini_object_make_writable): Unref the arg if we copy, like gst_caps_make_writable. * gst/gstmessage.c (_gst_message_initialize): No trace init. * gst/gstghostpad.c (gst_proxy_pad_do_activate) (gst_proxy_pad_do_activatepull, gst_proxy_pad_do_activatepush): Adapt for new pad API. * gst/gstevent.c (_gst_event_initialize): Don't initialize trace. * gst/gstelement.h: * gst/gstelement.c (gst_element_iterate_src_pads) (gst_element_iterate_sink_pads): New API functions. * gst/gstelement.c (iterator_fold_with_resync): New utility, should fold into gstiterator.c in some form. (gst_element_pads_activate): Simplified via use of fold and delegation of decisions to gstpad->activate. * gst/gstbus.c (gst_bus_source_finalize): Set the bus to NULL, help in debugging. * gst/gstbuffer.c (_gst_buffer_initialize): Ref the buffer type class once in init, like gstmessage. Didn't run into this issue but it seems correct. Don't initialize a trace, gstminiobject does that. * check/pipelines/simple_launch_lines.c (test_stop_from_app): New test, runs fakesrc ! fakesink, stopping on ::handoff via a message to the bus. (assert_live_count): New util function, uses alloc traces to check cleanup. * check/gst/gstghostpad.c (test_ghost_pads): More refcount checks. To be modified when unlink drops the internal pad.
2005-06-27 18:35:05 +00:00
}
}
/**
* gst_pad_set_active:
* @pad: the #GstPad to activate or deactivate.
gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. Original commit message from CVS: 2005-06-27 Andy Wingo <wingo@pobox.com> * gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. * gst/gsttrace.c (gst_alloc_trace_list_sorted): New helper, returns a sorted copy of the trace list. (gst_alloc_trace_print_live): New API, only prints traces with live objects. Sort the list. (gst_alloc_trace_print_all): Sort the list. (gst_alloc_trace_print): Align columns. * gst/elements/gstttypefindelement.c: * gst/elements/gsttee.c: * gst/base/gstbasesrc.c: * gst/base/gstbasesink.c: * gst/base/gstbasetransform.c: * gst/gstqueue.c: Adapt for pad activation changes. * gst/gstpipeline.c (gst_pipeline_init): Unref after parenting sched. (gst_pipeline_dispose): Drop ref on sched. * gst/gstpad.c (gst_pad_init): Set the default activate func. (gst_pad_activate_default): Push mode by default. (pre_activate_switch, post_activate_switch): New stubs, things to do before and after switching activation modes on pads. (gst_pad_set_active): Take a boolean and not a mode, dispatch to the pad's activate function to choose which mode to activate. Shortcut on deactivation and call the right function directly. (gst_pad_activate_pull): New API, (de)activates a pad in pull mode. (gst_pad_activate_push): New API, same for push mode. (gst_pad_set_activate_function) (gst_pad_set_activatepull_function) (gst_pad_set_activatepush_function): Setters for new API. * gst/gstminiobject.c (gst_mini_object_new, gst_mini_object_free): Trace all miniobjects. (gst_mini_object_make_writable): Unref the arg if we copy, like gst_caps_make_writable. * gst/gstmessage.c (_gst_message_initialize): No trace init. * gst/gstghostpad.c (gst_proxy_pad_do_activate) (gst_proxy_pad_do_activatepull, gst_proxy_pad_do_activatepush): Adapt for new pad API. * gst/gstevent.c (_gst_event_initialize): Don't initialize trace. * gst/gstelement.h: * gst/gstelement.c (gst_element_iterate_src_pads) (gst_element_iterate_sink_pads): New API functions. * gst/gstelement.c (iterator_fold_with_resync): New utility, should fold into gstiterator.c in some form. (gst_element_pads_activate): Simplified via use of fold and delegation of decisions to gstpad->activate. * gst/gstbus.c (gst_bus_source_finalize): Set the bus to NULL, help in debugging. * gst/gstbuffer.c (_gst_buffer_initialize): Ref the buffer type class once in init, like gstmessage. Didn't run into this issue but it seems correct. Don't initialize a trace, gstminiobject does that. * check/pipelines/simple_launch_lines.c (test_stop_from_app): New test, runs fakesrc ! fakesink, stopping on ::handoff via a message to the bus. (assert_live_count): New util function, uses alloc traces to check cleanup. * check/gst/gstghostpad.c (test_ghost_pads): More refcount checks. To be modified when unlink drops the internal pad.
2005-06-27 18:35:05 +00:00
* @active: whether or not the pad should be active.
*
* Activates or deactivates the given pad.
gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. Original commit message from CVS: 2005-06-27 Andy Wingo <wingo@pobox.com> * gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. * gst/gsttrace.c (gst_alloc_trace_list_sorted): New helper, returns a sorted copy of the trace list. (gst_alloc_trace_print_live): New API, only prints traces with live objects. Sort the list. (gst_alloc_trace_print_all): Sort the list. (gst_alloc_trace_print): Align columns. * gst/elements/gstttypefindelement.c: * gst/elements/gsttee.c: * gst/base/gstbasesrc.c: * gst/base/gstbasesink.c: * gst/base/gstbasetransform.c: * gst/gstqueue.c: Adapt for pad activation changes. * gst/gstpipeline.c (gst_pipeline_init): Unref after parenting sched. (gst_pipeline_dispose): Drop ref on sched. * gst/gstpad.c (gst_pad_init): Set the default activate func. (gst_pad_activate_default): Push mode by default. (pre_activate_switch, post_activate_switch): New stubs, things to do before and after switching activation modes on pads. (gst_pad_set_active): Take a boolean and not a mode, dispatch to the pad's activate function to choose which mode to activate. Shortcut on deactivation and call the right function directly. (gst_pad_activate_pull): New API, (de)activates a pad in pull mode. (gst_pad_activate_push): New API, same for push mode. (gst_pad_set_activate_function) (gst_pad_set_activatepull_function) (gst_pad_set_activatepush_function): Setters for new API. * gst/gstminiobject.c (gst_mini_object_new, gst_mini_object_free): Trace all miniobjects. (gst_mini_object_make_writable): Unref the arg if we copy, like gst_caps_make_writable. * gst/gstmessage.c (_gst_message_initialize): No trace init. * gst/gstghostpad.c (gst_proxy_pad_do_activate) (gst_proxy_pad_do_activatepull, gst_proxy_pad_do_activatepush): Adapt for new pad API. * gst/gstevent.c (_gst_event_initialize): Don't initialize trace. * gst/gstelement.h: * gst/gstelement.c (gst_element_iterate_src_pads) (gst_element_iterate_sink_pads): New API functions. * gst/gstelement.c (iterator_fold_with_resync): New utility, should fold into gstiterator.c in some form. (gst_element_pads_activate): Simplified via use of fold and delegation of decisions to gstpad->activate. * gst/gstbus.c (gst_bus_source_finalize): Set the bus to NULL, help in debugging. * gst/gstbuffer.c (_gst_buffer_initialize): Ref the buffer type class once in init, like gstmessage. Didn't run into this issue but it seems correct. Don't initialize a trace, gstminiobject does that. * check/pipelines/simple_launch_lines.c (test_stop_from_app): New test, runs fakesrc ! fakesink, stopping on ::handoff via a message to the bus. (assert_live_count): New util function, uses alloc traces to check cleanup. * check/gst/gstghostpad.c (test_ghost_pads): More refcount checks. To be modified when unlink drops the internal pad.
2005-06-27 18:35:05 +00:00
* Normally called from within core state change functions.
gst/: Simplify pad activation. Original commit message from CVS: * gst/base/Makefile.am: * gst/base/README: * gst/base/gstbasesink.c: (gst_basesink_get_type), (gst_basesink_base_init), (gst_basesink_class_init), (gst_basesink_pad_getcaps), (gst_basesink_init), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesink.h: * gst/base/gstbasetransform.c: (gst_base_transform_get_type), (gst_base_transform_base_init), (gst_base_transform_finalize), (gst_base_transform_class_init), (gst_base_transform_init), (gst_base_transform_proxy_getcaps), (gst_base_transform_setcaps), (gst_base_transform_event), (gst_base_transform_getrange), (gst_base_transform_chain), (gst_base_transform_handle_buffer), (gst_base_transform_set_property), (gst_base_transform_get_property), (gst_base_transform_sink_activate), (gst_base_transform_src_activate), (gst_base_transform_change_state): * gst/base/gstbasetransform.h: * gst/elements/gstidentity.c: (gst_identity_finalize), (gst_identity_class_init), (gst_identity_init), (gst_identity_event), (gst_identity_check_perfect), (gst_identity_transform), (gst_identity_set_property), (gst_identity_get_property), (gst_identity_change_state): * gst/elements/gstidentity.h: * gst/gstelement.c: (gst_element_get_state_func), (gst_element_lost_state), (gst_element_pads_activate): * gst/gstpad.c: (gst_pad_set_active), (gst_pad_peer_set_active), (gst_pad_check_pull_range), (gst_pad_pull_range): * gst/gstpad.h: Simplify pad activation. Added function to check if pull_range can be performed. Error out when pulling inactive or flushing pads. Removed const from refcounted types as it does not make sense. Simplify pad templates in basesink Added base class for simple 1-to-1 transforms. Make identity subclass the base transform.
2005-03-29 16:18:12 +00:00
*
gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. Original commit message from CVS: 2005-06-27 Andy Wingo <wingo@pobox.com> * gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. * gst/gsttrace.c (gst_alloc_trace_list_sorted): New helper, returns a sorted copy of the trace list. (gst_alloc_trace_print_live): New API, only prints traces with live objects. Sort the list. (gst_alloc_trace_print_all): Sort the list. (gst_alloc_trace_print): Align columns. * gst/elements/gstttypefindelement.c: * gst/elements/gsttee.c: * gst/base/gstbasesrc.c: * gst/base/gstbasesink.c: * gst/base/gstbasetransform.c: * gst/gstqueue.c: Adapt for pad activation changes. * gst/gstpipeline.c (gst_pipeline_init): Unref after parenting sched. (gst_pipeline_dispose): Drop ref on sched. * gst/gstpad.c (gst_pad_init): Set the default activate func. (gst_pad_activate_default): Push mode by default. (pre_activate_switch, post_activate_switch): New stubs, things to do before and after switching activation modes on pads. (gst_pad_set_active): Take a boolean and not a mode, dispatch to the pad's activate function to choose which mode to activate. Shortcut on deactivation and call the right function directly. (gst_pad_activate_pull): New API, (de)activates a pad in pull mode. (gst_pad_activate_push): New API, same for push mode. (gst_pad_set_activate_function) (gst_pad_set_activatepull_function) (gst_pad_set_activatepush_function): Setters for new API. * gst/gstminiobject.c (gst_mini_object_new, gst_mini_object_free): Trace all miniobjects. (gst_mini_object_make_writable): Unref the arg if we copy, like gst_caps_make_writable. * gst/gstmessage.c (_gst_message_initialize): No trace init. * gst/gstghostpad.c (gst_proxy_pad_do_activate) (gst_proxy_pad_do_activatepull, gst_proxy_pad_do_activatepush): Adapt for new pad API. * gst/gstevent.c (_gst_event_initialize): Don't initialize trace. * gst/gstelement.h: * gst/gstelement.c (gst_element_iterate_src_pads) (gst_element_iterate_sink_pads): New API functions. * gst/gstelement.c (iterator_fold_with_resync): New utility, should fold into gstiterator.c in some form. (gst_element_pads_activate): Simplified via use of fold and delegation of decisions to gstpad->activate. * gst/gstbus.c (gst_bus_source_finalize): Set the bus to NULL, help in debugging. * gst/gstbuffer.c (_gst_buffer_initialize): Ref the buffer type class once in init, like gstmessage. Didn't run into this issue but it seems correct. Don't initialize a trace, gstminiobject does that. * check/pipelines/simple_launch_lines.c (test_stop_from_app): New test, runs fakesrc ! fakesink, stopping on ::handoff via a message to the bus. (assert_live_count): New util function, uses alloc traces to check cleanup. * check/gst/gstghostpad.c (test_ghost_pads): More refcount checks. To be modified when unlink drops the internal pad.
2005-06-27 18:35:05 +00:00
* If @active, makes sure the pad is active. If it is already active, either in
* push or pull mode, just return. Otherwise dispatches to the pad's activate
* function to perform the actual activation.
gst/: Simplify pad activation. Original commit message from CVS: * gst/base/Makefile.am: * gst/base/README: * gst/base/gstbasesink.c: (gst_basesink_get_type), (gst_basesink_base_init), (gst_basesink_class_init), (gst_basesink_pad_getcaps), (gst_basesink_init), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesink.h: * gst/base/gstbasetransform.c: (gst_base_transform_get_type), (gst_base_transform_base_init), (gst_base_transform_finalize), (gst_base_transform_class_init), (gst_base_transform_init), (gst_base_transform_proxy_getcaps), (gst_base_transform_setcaps), (gst_base_transform_event), (gst_base_transform_getrange), (gst_base_transform_chain), (gst_base_transform_handle_buffer), (gst_base_transform_set_property), (gst_base_transform_get_property), (gst_base_transform_sink_activate), (gst_base_transform_src_activate), (gst_base_transform_change_state): * gst/base/gstbasetransform.h: * gst/elements/gstidentity.c: (gst_identity_finalize), (gst_identity_class_init), (gst_identity_init), (gst_identity_event), (gst_identity_check_perfect), (gst_identity_transform), (gst_identity_set_property), (gst_identity_get_property), (gst_identity_change_state): * gst/elements/gstidentity.h: * gst/gstelement.c: (gst_element_get_state_func), (gst_element_lost_state), (gst_element_pads_activate): * gst/gstpad.c: (gst_pad_set_active), (gst_pad_peer_set_active), (gst_pad_check_pull_range), (gst_pad_pull_range): * gst/gstpad.h: Simplify pad activation. Added function to check if pull_range can be performed. Error out when pulling inactive or flushing pads. Removed const from refcounted types as it does not make sense. Simplify pad templates in basesink Added base class for simple 1-to-1 transforms. Make identity subclass the base transform.
2005-03-29 16:18:12 +00:00
*
gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. Original commit message from CVS: 2005-06-27 Andy Wingo <wingo@pobox.com> * gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. * gst/gsttrace.c (gst_alloc_trace_list_sorted): New helper, returns a sorted copy of the trace list. (gst_alloc_trace_print_live): New API, only prints traces with live objects. Sort the list. (gst_alloc_trace_print_all): Sort the list. (gst_alloc_trace_print): Align columns. * gst/elements/gstttypefindelement.c: * gst/elements/gsttee.c: * gst/base/gstbasesrc.c: * gst/base/gstbasesink.c: * gst/base/gstbasetransform.c: * gst/gstqueue.c: Adapt for pad activation changes. * gst/gstpipeline.c (gst_pipeline_init): Unref after parenting sched. (gst_pipeline_dispose): Drop ref on sched. * gst/gstpad.c (gst_pad_init): Set the default activate func. (gst_pad_activate_default): Push mode by default. (pre_activate_switch, post_activate_switch): New stubs, things to do before and after switching activation modes on pads. (gst_pad_set_active): Take a boolean and not a mode, dispatch to the pad's activate function to choose which mode to activate. Shortcut on deactivation and call the right function directly. (gst_pad_activate_pull): New API, (de)activates a pad in pull mode. (gst_pad_activate_push): New API, same for push mode. (gst_pad_set_activate_function) (gst_pad_set_activatepull_function) (gst_pad_set_activatepush_function): Setters for new API. * gst/gstminiobject.c (gst_mini_object_new, gst_mini_object_free): Trace all miniobjects. (gst_mini_object_make_writable): Unref the arg if we copy, like gst_caps_make_writable. * gst/gstmessage.c (_gst_message_initialize): No trace init. * gst/gstghostpad.c (gst_proxy_pad_do_activate) (gst_proxy_pad_do_activatepull, gst_proxy_pad_do_activatepush): Adapt for new pad API. * gst/gstevent.c (_gst_event_initialize): Don't initialize trace. * gst/gstelement.h: * gst/gstelement.c (gst_element_iterate_src_pads) (gst_element_iterate_sink_pads): New API functions. * gst/gstelement.c (iterator_fold_with_resync): New utility, should fold into gstiterator.c in some form. (gst_element_pads_activate): Simplified via use of fold and delegation of decisions to gstpad->activate. * gst/gstbus.c (gst_bus_source_finalize): Set the bus to NULL, help in debugging. * gst/gstbuffer.c (_gst_buffer_initialize): Ref the buffer type class once in init, like gstmessage. Didn't run into this issue but it seems correct. Don't initialize a trace, gstminiobject does that. * check/pipelines/simple_launch_lines.c (test_stop_from_app): New test, runs fakesrc ! fakesink, stopping on ::handoff via a message to the bus. (assert_live_count): New util function, uses alloc traces to check cleanup. * check/gst/gstghostpad.c (test_ghost_pads): More refcount checks. To be modified when unlink drops the internal pad.
2005-06-27 18:35:05 +00:00
* If not @active, checks the pad's current mode and calls
* gst_pad_activate_push() or gst_pad_activate_pull(), as appropriate, with a
* FALSE argument.
*
* Returns: #TRUE if the operation was successful.
*
* MT safe.
*/
gboolean
gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. Original commit message from CVS: 2005-06-27 Andy Wingo <wingo@pobox.com> * gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. * gst/gsttrace.c (gst_alloc_trace_list_sorted): New helper, returns a sorted copy of the trace list. (gst_alloc_trace_print_live): New API, only prints traces with live objects. Sort the list. (gst_alloc_trace_print_all): Sort the list. (gst_alloc_trace_print): Align columns. * gst/elements/gstttypefindelement.c: * gst/elements/gsttee.c: * gst/base/gstbasesrc.c: * gst/base/gstbasesink.c: * gst/base/gstbasetransform.c: * gst/gstqueue.c: Adapt for pad activation changes. * gst/gstpipeline.c (gst_pipeline_init): Unref after parenting sched. (gst_pipeline_dispose): Drop ref on sched. * gst/gstpad.c (gst_pad_init): Set the default activate func. (gst_pad_activate_default): Push mode by default. (pre_activate_switch, post_activate_switch): New stubs, things to do before and after switching activation modes on pads. (gst_pad_set_active): Take a boolean and not a mode, dispatch to the pad's activate function to choose which mode to activate. Shortcut on deactivation and call the right function directly. (gst_pad_activate_pull): New API, (de)activates a pad in pull mode. (gst_pad_activate_push): New API, same for push mode. (gst_pad_set_activate_function) (gst_pad_set_activatepull_function) (gst_pad_set_activatepush_function): Setters for new API. * gst/gstminiobject.c (gst_mini_object_new, gst_mini_object_free): Trace all miniobjects. (gst_mini_object_make_writable): Unref the arg if we copy, like gst_caps_make_writable. * gst/gstmessage.c (_gst_message_initialize): No trace init. * gst/gstghostpad.c (gst_proxy_pad_do_activate) (gst_proxy_pad_do_activatepull, gst_proxy_pad_do_activatepush): Adapt for new pad API. * gst/gstevent.c (_gst_event_initialize): Don't initialize trace. * gst/gstelement.h: * gst/gstelement.c (gst_element_iterate_src_pads) (gst_element_iterate_sink_pads): New API functions. * gst/gstelement.c (iterator_fold_with_resync): New utility, should fold into gstiterator.c in some form. (gst_element_pads_activate): Simplified via use of fold and delegation of decisions to gstpad->activate. * gst/gstbus.c (gst_bus_source_finalize): Set the bus to NULL, help in debugging. * gst/gstbuffer.c (_gst_buffer_initialize): Ref the buffer type class once in init, like gstmessage. Didn't run into this issue but it seems correct. Don't initialize a trace, gstminiobject does that. * check/pipelines/simple_launch_lines.c (test_stop_from_app): New test, runs fakesrc ! fakesink, stopping on ::handoff via a message to the bus. (assert_live_count): New util function, uses alloc traces to check cleanup. * check/gst/gstghostpad.c (test_ghost_pads): More refcount checks. To be modified when unlink drops the internal pad.
2005-06-27 18:35:05 +00:00
gst_pad_set_active (GstPad * pad, gboolean active)
{
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
GstActivateMode old;
gboolean ret = FALSE;
g_return_val_if_fail (GST_IS_PAD (pad), FALSE);
GST_OBJECT_LOCK (pad);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
old = GST_PAD_ACTIVATE_MODE (pad);
GST_OBJECT_UNLOCK (pad);
gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. Original commit message from CVS: 2005-06-27 Andy Wingo <wingo@pobox.com> * gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. * gst/gsttrace.c (gst_alloc_trace_list_sorted): New helper, returns a sorted copy of the trace list. (gst_alloc_trace_print_live): New API, only prints traces with live objects. Sort the list. (gst_alloc_trace_print_all): Sort the list. (gst_alloc_trace_print): Align columns. * gst/elements/gstttypefindelement.c: * gst/elements/gsttee.c: * gst/base/gstbasesrc.c: * gst/base/gstbasesink.c: * gst/base/gstbasetransform.c: * gst/gstqueue.c: Adapt for pad activation changes. * gst/gstpipeline.c (gst_pipeline_init): Unref after parenting sched. (gst_pipeline_dispose): Drop ref on sched. * gst/gstpad.c (gst_pad_init): Set the default activate func. (gst_pad_activate_default): Push mode by default. (pre_activate_switch, post_activate_switch): New stubs, things to do before and after switching activation modes on pads. (gst_pad_set_active): Take a boolean and not a mode, dispatch to the pad's activate function to choose which mode to activate. Shortcut on deactivation and call the right function directly. (gst_pad_activate_pull): New API, (de)activates a pad in pull mode. (gst_pad_activate_push): New API, same for push mode. (gst_pad_set_activate_function) (gst_pad_set_activatepull_function) (gst_pad_set_activatepush_function): Setters for new API. * gst/gstminiobject.c (gst_mini_object_new, gst_mini_object_free): Trace all miniobjects. (gst_mini_object_make_writable): Unref the arg if we copy, like gst_caps_make_writable. * gst/gstmessage.c (_gst_message_initialize): No trace init. * gst/gstghostpad.c (gst_proxy_pad_do_activate) (gst_proxy_pad_do_activatepull, gst_proxy_pad_do_activatepush): Adapt for new pad API. * gst/gstevent.c (_gst_event_initialize): Don't initialize trace. * gst/gstelement.h: * gst/gstelement.c (gst_element_iterate_src_pads) (gst_element_iterate_sink_pads): New API functions. * gst/gstelement.c (iterator_fold_with_resync): New utility, should fold into gstiterator.c in some form. (gst_element_pads_activate): Simplified via use of fold and delegation of decisions to gstpad->activate. * gst/gstbus.c (gst_bus_source_finalize): Set the bus to NULL, help in debugging. * gst/gstbuffer.c (_gst_buffer_initialize): Ref the buffer type class once in init, like gstmessage. Didn't run into this issue but it seems correct. Don't initialize a trace, gstminiobject does that. * check/pipelines/simple_launch_lines.c (test_stop_from_app): New test, runs fakesrc ! fakesink, stopping on ::handoff via a message to the bus. (assert_live_count): New util function, uses alloc traces to check cleanup. * check/gst/gstghostpad.c (test_ghost_pads): More refcount checks. To be modified when unlink drops the internal pad.
2005-06-27 18:35:05 +00:00
if (active) {
switch (old) {
case GST_ACTIVATE_PUSH:
case GST_ACTIVATE_PULL:
ret = TRUE;
break;
case GST_ACTIVATE_NONE:
ret = (GST_PAD_ACTIVATEFUNC (pad)) (pad);
break;
}
} else {
switch (old) {
case GST_ACTIVATE_PUSH:
ret = gst_pad_activate_push (pad, FALSE);
break;
case GST_ACTIVATE_PULL:
ret = gst_pad_activate_pull (pad, FALSE);
break;
case GST_ACTIVATE_NONE:
ret = TRUE;
break;
}
}
if (!ret) {
GST_OBJECT_LOCK (pad);
if (!active) {
g_critical ("Failed to deactivate pad %s:%s, very bad",
GST_DEBUG_PAD_NAME (pad));
} else {
GST_WARNING ("Failed to activate pad %s:%s", GST_DEBUG_PAD_NAME (pad));
}
GST_OBJECT_UNLOCK (pad);
}
gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. Original commit message from CVS: 2005-06-27 Andy Wingo <wingo@pobox.com> * gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. * gst/gsttrace.c (gst_alloc_trace_list_sorted): New helper, returns a sorted copy of the trace list. (gst_alloc_trace_print_live): New API, only prints traces with live objects. Sort the list. (gst_alloc_trace_print_all): Sort the list. (gst_alloc_trace_print): Align columns. * gst/elements/gstttypefindelement.c: * gst/elements/gsttee.c: * gst/base/gstbasesrc.c: * gst/base/gstbasesink.c: * gst/base/gstbasetransform.c: * gst/gstqueue.c: Adapt for pad activation changes. * gst/gstpipeline.c (gst_pipeline_init): Unref after parenting sched. (gst_pipeline_dispose): Drop ref on sched. * gst/gstpad.c (gst_pad_init): Set the default activate func. (gst_pad_activate_default): Push mode by default. (pre_activate_switch, post_activate_switch): New stubs, things to do before and after switching activation modes on pads. (gst_pad_set_active): Take a boolean and not a mode, dispatch to the pad's activate function to choose which mode to activate. Shortcut on deactivation and call the right function directly. (gst_pad_activate_pull): New API, (de)activates a pad in pull mode. (gst_pad_activate_push): New API, same for push mode. (gst_pad_set_activate_function) (gst_pad_set_activatepull_function) (gst_pad_set_activatepush_function): Setters for new API. * gst/gstminiobject.c (gst_mini_object_new, gst_mini_object_free): Trace all miniobjects. (gst_mini_object_make_writable): Unref the arg if we copy, like gst_caps_make_writable. * gst/gstmessage.c (_gst_message_initialize): No trace init. * gst/gstghostpad.c (gst_proxy_pad_do_activate) (gst_proxy_pad_do_activatepull, gst_proxy_pad_do_activatepush): Adapt for new pad API. * gst/gstevent.c (_gst_event_initialize): Don't initialize trace. * gst/gstelement.h: * gst/gstelement.c (gst_element_iterate_src_pads) (gst_element_iterate_sink_pads): New API functions. * gst/gstelement.c (iterator_fold_with_resync): New utility, should fold into gstiterator.c in some form. (gst_element_pads_activate): Simplified via use of fold and delegation of decisions to gstpad->activate. * gst/gstbus.c (gst_bus_source_finalize): Set the bus to NULL, help in debugging. * gst/gstbuffer.c (_gst_buffer_initialize): Ref the buffer type class once in init, like gstmessage. Didn't run into this issue but it seems correct. Don't initialize a trace, gstminiobject does that. * check/pipelines/simple_launch_lines.c (test_stop_from_app): New test, runs fakesrc ! fakesink, stopping on ::handoff via a message to the bus. (assert_live_count): New util function, uses alloc traces to check cleanup. * check/gst/gstghostpad.c (test_ghost_pads): More refcount checks. To be modified when unlink drops the internal pad.
2005-06-27 18:35:05 +00:00
return ret;
}
gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. Original commit message from CVS: 2005-06-27 Andy Wingo <wingo@pobox.com> * gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. * gst/gsttrace.c (gst_alloc_trace_list_sorted): New helper, returns a sorted copy of the trace list. (gst_alloc_trace_print_live): New API, only prints traces with live objects. Sort the list. (gst_alloc_trace_print_all): Sort the list. (gst_alloc_trace_print): Align columns. * gst/elements/gstttypefindelement.c: * gst/elements/gsttee.c: * gst/base/gstbasesrc.c: * gst/base/gstbasesink.c: * gst/base/gstbasetransform.c: * gst/gstqueue.c: Adapt for pad activation changes. * gst/gstpipeline.c (gst_pipeline_init): Unref after parenting sched. (gst_pipeline_dispose): Drop ref on sched. * gst/gstpad.c (gst_pad_init): Set the default activate func. (gst_pad_activate_default): Push mode by default. (pre_activate_switch, post_activate_switch): New stubs, things to do before and after switching activation modes on pads. (gst_pad_set_active): Take a boolean and not a mode, dispatch to the pad's activate function to choose which mode to activate. Shortcut on deactivation and call the right function directly. (gst_pad_activate_pull): New API, (de)activates a pad in pull mode. (gst_pad_activate_push): New API, same for push mode. (gst_pad_set_activate_function) (gst_pad_set_activatepull_function) (gst_pad_set_activatepush_function): Setters for new API. * gst/gstminiobject.c (gst_mini_object_new, gst_mini_object_free): Trace all miniobjects. (gst_mini_object_make_writable): Unref the arg if we copy, like gst_caps_make_writable. * gst/gstmessage.c (_gst_message_initialize): No trace init. * gst/gstghostpad.c (gst_proxy_pad_do_activate) (gst_proxy_pad_do_activatepull, gst_proxy_pad_do_activatepush): Adapt for new pad API. * gst/gstevent.c (_gst_event_initialize): Don't initialize trace. * gst/gstelement.h: * gst/gstelement.c (gst_element_iterate_src_pads) (gst_element_iterate_sink_pads): New API functions. * gst/gstelement.c (iterator_fold_with_resync): New utility, should fold into gstiterator.c in some form. (gst_element_pads_activate): Simplified via use of fold and delegation of decisions to gstpad->activate. * gst/gstbus.c (gst_bus_source_finalize): Set the bus to NULL, help in debugging. * gst/gstbuffer.c (_gst_buffer_initialize): Ref the buffer type class once in init, like gstmessage. Didn't run into this issue but it seems correct. Don't initialize a trace, gstminiobject does that. * check/pipelines/simple_launch_lines.c (test_stop_from_app): New test, runs fakesrc ! fakesink, stopping on ::handoff via a message to the bus. (assert_live_count): New util function, uses alloc traces to check cleanup. * check/gst/gstghostpad.c (test_ghost_pads): More refcount checks. To be modified when unlink drops the internal pad.
2005-06-27 18:35:05 +00:00
/**
* gst_pad_activate_pull:
* @pad: the #GstPad to activate or deactivate.
* @active: whether or not the pad should be active.
*
* Activates or deactivates the given pad in pull mode via dispatching to the
* pad's activatepullfunc. For use from within pad activation functions only.
* When called on sink pads, will first proxy the call to the peer pad, which is
* expected to activate its internally linked pads from within its activate_pull
* function.
*
* If you don't know what this is, you probably don't want to call it.
*
* Returns: TRUE if the operation was successful.
gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. Original commit message from CVS: 2005-06-27 Andy Wingo <wingo@pobox.com> * gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. * gst/gsttrace.c (gst_alloc_trace_list_sorted): New helper, returns a sorted copy of the trace list. (gst_alloc_trace_print_live): New API, only prints traces with live objects. Sort the list. (gst_alloc_trace_print_all): Sort the list. (gst_alloc_trace_print): Align columns. * gst/elements/gstttypefindelement.c: * gst/elements/gsttee.c: * gst/base/gstbasesrc.c: * gst/base/gstbasesink.c: * gst/base/gstbasetransform.c: * gst/gstqueue.c: Adapt for pad activation changes. * gst/gstpipeline.c (gst_pipeline_init): Unref after parenting sched. (gst_pipeline_dispose): Drop ref on sched. * gst/gstpad.c (gst_pad_init): Set the default activate func. (gst_pad_activate_default): Push mode by default. (pre_activate_switch, post_activate_switch): New stubs, things to do before and after switching activation modes on pads. (gst_pad_set_active): Take a boolean and not a mode, dispatch to the pad's activate function to choose which mode to activate. Shortcut on deactivation and call the right function directly. (gst_pad_activate_pull): New API, (de)activates a pad in pull mode. (gst_pad_activate_push): New API, same for push mode. (gst_pad_set_activate_function) (gst_pad_set_activatepull_function) (gst_pad_set_activatepush_function): Setters for new API. * gst/gstminiobject.c (gst_mini_object_new, gst_mini_object_free): Trace all miniobjects. (gst_mini_object_make_writable): Unref the arg if we copy, like gst_caps_make_writable. * gst/gstmessage.c (_gst_message_initialize): No trace init. * gst/gstghostpad.c (gst_proxy_pad_do_activate) (gst_proxy_pad_do_activatepull, gst_proxy_pad_do_activatepush): Adapt for new pad API. * gst/gstevent.c (_gst_event_initialize): Don't initialize trace. * gst/gstelement.h: * gst/gstelement.c (gst_element_iterate_src_pads) (gst_element_iterate_sink_pads): New API functions. * gst/gstelement.c (iterator_fold_with_resync): New utility, should fold into gstiterator.c in some form. (gst_element_pads_activate): Simplified via use of fold and delegation of decisions to gstpad->activate. * gst/gstbus.c (gst_bus_source_finalize): Set the bus to NULL, help in debugging. * gst/gstbuffer.c (_gst_buffer_initialize): Ref the buffer type class once in init, like gstmessage. Didn't run into this issue but it seems correct. Don't initialize a trace, gstminiobject does that. * check/pipelines/simple_launch_lines.c (test_stop_from_app): New test, runs fakesrc ! fakesink, stopping on ::handoff via a message to the bus. (assert_live_count): New util function, uses alloc traces to check cleanup. * check/gst/gstghostpad.c (test_ghost_pads): More refcount checks. To be modified when unlink drops the internal pad.
2005-06-27 18:35:05 +00:00
*
* MT safe.
*/
gboolean
gst_pad_activate_pull (GstPad * pad, gboolean active)
{
GstActivateMode old, new;
GstPad *peer;
gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. Original commit message from CVS: 2005-06-27 Andy Wingo <wingo@pobox.com> * gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. * gst/gsttrace.c (gst_alloc_trace_list_sorted): New helper, returns a sorted copy of the trace list. (gst_alloc_trace_print_live): New API, only prints traces with live objects. Sort the list. (gst_alloc_trace_print_all): Sort the list. (gst_alloc_trace_print): Align columns. * gst/elements/gstttypefindelement.c: * gst/elements/gsttee.c: * gst/base/gstbasesrc.c: * gst/base/gstbasesink.c: * gst/base/gstbasetransform.c: * gst/gstqueue.c: Adapt for pad activation changes. * gst/gstpipeline.c (gst_pipeline_init): Unref after parenting sched. (gst_pipeline_dispose): Drop ref on sched. * gst/gstpad.c (gst_pad_init): Set the default activate func. (gst_pad_activate_default): Push mode by default. (pre_activate_switch, post_activate_switch): New stubs, things to do before and after switching activation modes on pads. (gst_pad_set_active): Take a boolean and not a mode, dispatch to the pad's activate function to choose which mode to activate. Shortcut on deactivation and call the right function directly. (gst_pad_activate_pull): New API, (de)activates a pad in pull mode. (gst_pad_activate_push): New API, same for push mode. (gst_pad_set_activate_function) (gst_pad_set_activatepull_function) (gst_pad_set_activatepush_function): Setters for new API. * gst/gstminiobject.c (gst_mini_object_new, gst_mini_object_free): Trace all miniobjects. (gst_mini_object_make_writable): Unref the arg if we copy, like gst_caps_make_writable. * gst/gstmessage.c (_gst_message_initialize): No trace init. * gst/gstghostpad.c (gst_proxy_pad_do_activate) (gst_proxy_pad_do_activatepull, gst_proxy_pad_do_activatepush): Adapt for new pad API. * gst/gstevent.c (_gst_event_initialize): Don't initialize trace. * gst/gstelement.h: * gst/gstelement.c (gst_element_iterate_src_pads) (gst_element_iterate_sink_pads): New API functions. * gst/gstelement.c (iterator_fold_with_resync): New utility, should fold into gstiterator.c in some form. (gst_element_pads_activate): Simplified via use of fold and delegation of decisions to gstpad->activate. * gst/gstbus.c (gst_bus_source_finalize): Set the bus to NULL, help in debugging. * gst/gstbuffer.c (_gst_buffer_initialize): Ref the buffer type class once in init, like gstmessage. Didn't run into this issue but it seems correct. Don't initialize a trace, gstminiobject does that. * check/pipelines/simple_launch_lines.c (test_stop_from_app): New test, runs fakesrc ! fakesink, stopping on ::handoff via a message to the bus. (assert_live_count): New util function, uses alloc traces to check cleanup. * check/gst/gstghostpad.c (test_ghost_pads): More refcount checks. To be modified when unlink drops the internal pad.
2005-06-27 18:35:05 +00:00
g_return_val_if_fail (GST_IS_PAD (pad), FALSE);
GST_OBJECT_LOCK (pad);
gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. Original commit message from CVS: 2005-06-27 Andy Wingo <wingo@pobox.com> * gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. * gst/gsttrace.c (gst_alloc_trace_list_sorted): New helper, returns a sorted copy of the trace list. (gst_alloc_trace_print_live): New API, only prints traces with live objects. Sort the list. (gst_alloc_trace_print_all): Sort the list. (gst_alloc_trace_print): Align columns. * gst/elements/gstttypefindelement.c: * gst/elements/gsttee.c: * gst/base/gstbasesrc.c: * gst/base/gstbasesink.c: * gst/base/gstbasetransform.c: * gst/gstqueue.c: Adapt for pad activation changes. * gst/gstpipeline.c (gst_pipeline_init): Unref after parenting sched. (gst_pipeline_dispose): Drop ref on sched. * gst/gstpad.c (gst_pad_init): Set the default activate func. (gst_pad_activate_default): Push mode by default. (pre_activate_switch, post_activate_switch): New stubs, things to do before and after switching activation modes on pads. (gst_pad_set_active): Take a boolean and not a mode, dispatch to the pad's activate function to choose which mode to activate. Shortcut on deactivation and call the right function directly. (gst_pad_activate_pull): New API, (de)activates a pad in pull mode. (gst_pad_activate_push): New API, same for push mode. (gst_pad_set_activate_function) (gst_pad_set_activatepull_function) (gst_pad_set_activatepush_function): Setters for new API. * gst/gstminiobject.c (gst_mini_object_new, gst_mini_object_free): Trace all miniobjects. (gst_mini_object_make_writable): Unref the arg if we copy, like gst_caps_make_writable. * gst/gstmessage.c (_gst_message_initialize): No trace init. * gst/gstghostpad.c (gst_proxy_pad_do_activate) (gst_proxy_pad_do_activatepull, gst_proxy_pad_do_activatepush): Adapt for new pad API. * gst/gstevent.c (_gst_event_initialize): Don't initialize trace. * gst/gstelement.h: * gst/gstelement.c (gst_element_iterate_src_pads) (gst_element_iterate_sink_pads): New API functions. * gst/gstelement.c (iterator_fold_with_resync): New utility, should fold into gstiterator.c in some form. (gst_element_pads_activate): Simplified via use of fold and delegation of decisions to gstpad->activate. * gst/gstbus.c (gst_bus_source_finalize): Set the bus to NULL, help in debugging. * gst/gstbuffer.c (_gst_buffer_initialize): Ref the buffer type class once in init, like gstmessage. Didn't run into this issue but it seems correct. Don't initialize a trace, gstminiobject does that. * check/pipelines/simple_launch_lines.c (test_stop_from_app): New test, runs fakesrc ! fakesink, stopping on ::handoff via a message to the bus. (assert_live_count): New util function, uses alloc traces to check cleanup. * check/gst/gstghostpad.c (test_ghost_pads): More refcount checks. To be modified when unlink drops the internal pad.
2005-06-27 18:35:05 +00:00
old = GST_PAD_ACTIVATE_MODE (pad);
GST_OBJECT_UNLOCK (pad);
gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. Original commit message from CVS: 2005-06-27 Andy Wingo <wingo@pobox.com> * gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. * gst/gsttrace.c (gst_alloc_trace_list_sorted): New helper, returns a sorted copy of the trace list. (gst_alloc_trace_print_live): New API, only prints traces with live objects. Sort the list. (gst_alloc_trace_print_all): Sort the list. (gst_alloc_trace_print): Align columns. * gst/elements/gstttypefindelement.c: * gst/elements/gsttee.c: * gst/base/gstbasesrc.c: * gst/base/gstbasesink.c: * gst/base/gstbasetransform.c: * gst/gstqueue.c: Adapt for pad activation changes. * gst/gstpipeline.c (gst_pipeline_init): Unref after parenting sched. (gst_pipeline_dispose): Drop ref on sched. * gst/gstpad.c (gst_pad_init): Set the default activate func. (gst_pad_activate_default): Push mode by default. (pre_activate_switch, post_activate_switch): New stubs, things to do before and after switching activation modes on pads. (gst_pad_set_active): Take a boolean and not a mode, dispatch to the pad's activate function to choose which mode to activate. Shortcut on deactivation and call the right function directly. (gst_pad_activate_pull): New API, (de)activates a pad in pull mode. (gst_pad_activate_push): New API, same for push mode. (gst_pad_set_activate_function) (gst_pad_set_activatepull_function) (gst_pad_set_activatepush_function): Setters for new API. * gst/gstminiobject.c (gst_mini_object_new, gst_mini_object_free): Trace all miniobjects. (gst_mini_object_make_writable): Unref the arg if we copy, like gst_caps_make_writable. * gst/gstmessage.c (_gst_message_initialize): No trace init. * gst/gstghostpad.c (gst_proxy_pad_do_activate) (gst_proxy_pad_do_activatepull, gst_proxy_pad_do_activatepush): Adapt for new pad API. * gst/gstevent.c (_gst_event_initialize): Don't initialize trace. * gst/gstelement.h: * gst/gstelement.c (gst_element_iterate_src_pads) (gst_element_iterate_sink_pads): New API functions. * gst/gstelement.c (iterator_fold_with_resync): New utility, should fold into gstiterator.c in some form. (gst_element_pads_activate): Simplified via use of fold and delegation of decisions to gstpad->activate. * gst/gstbus.c (gst_bus_source_finalize): Set the bus to NULL, help in debugging. * gst/gstbuffer.c (_gst_buffer_initialize): Ref the buffer type class once in init, like gstmessage. Didn't run into this issue but it seems correct. Don't initialize a trace, gstminiobject does that. * check/pipelines/simple_launch_lines.c (test_stop_from_app): New test, runs fakesrc ! fakesink, stopping on ::handoff via a message to the bus. (assert_live_count): New util function, uses alloc traces to check cleanup. * check/gst/gstghostpad.c (test_ghost_pads): More refcount checks. To be modified when unlink drops the internal pad.
2005-06-27 18:35:05 +00:00
if (active) {
switch (old) {
case GST_ACTIVATE_PULL:
goto was_ok;
case GST_ACTIVATE_PUSH:
/* pad was activate in the wrong direction, deactivate it
* and reactivate it in pull mode */
if (G_UNLIKELY (!gst_pad_activate_push (pad, FALSE)))
goto deactivate_failed;
/* fallthrough, pad is deactivated now. */
case GST_ACTIVATE_NONE:
break;
}
gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. Original commit message from CVS: 2005-06-27 Andy Wingo <wingo@pobox.com> * gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. * gst/gsttrace.c (gst_alloc_trace_list_sorted): New helper, returns a sorted copy of the trace list. (gst_alloc_trace_print_live): New API, only prints traces with live objects. Sort the list. (gst_alloc_trace_print_all): Sort the list. (gst_alloc_trace_print): Align columns. * gst/elements/gstttypefindelement.c: * gst/elements/gsttee.c: * gst/base/gstbasesrc.c: * gst/base/gstbasesink.c: * gst/base/gstbasetransform.c: * gst/gstqueue.c: Adapt for pad activation changes. * gst/gstpipeline.c (gst_pipeline_init): Unref after parenting sched. (gst_pipeline_dispose): Drop ref on sched. * gst/gstpad.c (gst_pad_init): Set the default activate func. (gst_pad_activate_default): Push mode by default. (pre_activate_switch, post_activate_switch): New stubs, things to do before and after switching activation modes on pads. (gst_pad_set_active): Take a boolean and not a mode, dispatch to the pad's activate function to choose which mode to activate. Shortcut on deactivation and call the right function directly. (gst_pad_activate_pull): New API, (de)activates a pad in pull mode. (gst_pad_activate_push): New API, same for push mode. (gst_pad_set_activate_function) (gst_pad_set_activatepull_function) (gst_pad_set_activatepush_function): Setters for new API. * gst/gstminiobject.c (gst_mini_object_new, gst_mini_object_free): Trace all miniobjects. (gst_mini_object_make_writable): Unref the arg if we copy, like gst_caps_make_writable. * gst/gstmessage.c (_gst_message_initialize): No trace init. * gst/gstghostpad.c (gst_proxy_pad_do_activate) (gst_proxy_pad_do_activatepull, gst_proxy_pad_do_activatepush): Adapt for new pad API. * gst/gstevent.c (_gst_event_initialize): Don't initialize trace. * gst/gstelement.h: * gst/gstelement.c (gst_element_iterate_src_pads) (gst_element_iterate_sink_pads): New API functions. * gst/gstelement.c (iterator_fold_with_resync): New utility, should fold into gstiterator.c in some form. (gst_element_pads_activate): Simplified via use of fold and delegation of decisions to gstpad->activate. * gst/gstbus.c (gst_bus_source_finalize): Set the bus to NULL, help in debugging. * gst/gstbuffer.c (_gst_buffer_initialize): Ref the buffer type class once in init, like gstmessage. Didn't run into this issue but it seems correct. Don't initialize a trace, gstminiobject does that. * check/pipelines/simple_launch_lines.c (test_stop_from_app): New test, runs fakesrc ! fakesink, stopping on ::handoff via a message to the bus. (assert_live_count): New util function, uses alloc traces to check cleanup. * check/gst/gstghostpad.c (test_ghost_pads): More refcount checks. To be modified when unlink drops the internal pad.
2005-06-27 18:35:05 +00:00
} else {
switch (old) {
case GST_ACTIVATE_NONE:
goto was_ok;
case GST_ACTIVATE_PUSH:
/* pad was activated in the other direction, deactivate it
* in push mode, this should not happen... */
if (G_UNLIKELY (!gst_pad_activate_push (pad, FALSE)))
goto deactivate_failed;
/* everything is fine now */
goto was_ok;
case GST_ACTIVATE_PULL:
break;
}
}
gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. Original commit message from CVS: 2005-06-27 Andy Wingo <wingo@pobox.com> * gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. * gst/gsttrace.c (gst_alloc_trace_list_sorted): New helper, returns a sorted copy of the trace list. (gst_alloc_trace_print_live): New API, only prints traces with live objects. Sort the list. (gst_alloc_trace_print_all): Sort the list. (gst_alloc_trace_print): Align columns. * gst/elements/gstttypefindelement.c: * gst/elements/gsttee.c: * gst/base/gstbasesrc.c: * gst/base/gstbasesink.c: * gst/base/gstbasetransform.c: * gst/gstqueue.c: Adapt for pad activation changes. * gst/gstpipeline.c (gst_pipeline_init): Unref after parenting sched. (gst_pipeline_dispose): Drop ref on sched. * gst/gstpad.c (gst_pad_init): Set the default activate func. (gst_pad_activate_default): Push mode by default. (pre_activate_switch, post_activate_switch): New stubs, things to do before and after switching activation modes on pads. (gst_pad_set_active): Take a boolean and not a mode, dispatch to the pad's activate function to choose which mode to activate. Shortcut on deactivation and call the right function directly. (gst_pad_activate_pull): New API, (de)activates a pad in pull mode. (gst_pad_activate_push): New API, same for push mode. (gst_pad_set_activate_function) (gst_pad_set_activatepull_function) (gst_pad_set_activatepush_function): Setters for new API. * gst/gstminiobject.c (gst_mini_object_new, gst_mini_object_free): Trace all miniobjects. (gst_mini_object_make_writable): Unref the arg if we copy, like gst_caps_make_writable. * gst/gstmessage.c (_gst_message_initialize): No trace init. * gst/gstghostpad.c (gst_proxy_pad_do_activate) (gst_proxy_pad_do_activatepull, gst_proxy_pad_do_activatepush): Adapt for new pad API. * gst/gstevent.c (_gst_event_initialize): Don't initialize trace. * gst/gstelement.h: * gst/gstelement.c (gst_element_iterate_src_pads) (gst_element_iterate_sink_pads): New API functions. * gst/gstelement.c (iterator_fold_with_resync): New utility, should fold into gstiterator.c in some form. (gst_element_pads_activate): Simplified via use of fold and delegation of decisions to gstpad->activate. * gst/gstbus.c (gst_bus_source_finalize): Set the bus to NULL, help in debugging. * gst/gstbuffer.c (_gst_buffer_initialize): Ref the buffer type class once in init, like gstmessage. Didn't run into this issue but it seems correct. Don't initialize a trace, gstminiobject does that. * check/pipelines/simple_launch_lines.c (test_stop_from_app): New test, runs fakesrc ! fakesink, stopping on ::handoff via a message to the bus. (assert_live_count): New util function, uses alloc traces to check cleanup. * check/gst/gstghostpad.c (test_ghost_pads): More refcount checks. To be modified when unlink drops the internal pad.
2005-06-27 18:35:05 +00:00
if (gst_pad_get_direction (pad) == GST_PAD_SINK) {
if ((peer = gst_pad_get_peer (pad))) {
if (G_UNLIKELY (!gst_pad_activate_pull (peer, active)))
goto peer_failed;
gst_object_unref (peer);
}
} else {
if (G_UNLIKELY (GST_PAD_GETRANGEFUNC (pad) == NULL))
goto failure; /* Can't activate pull on a src without a
getrange function */
}
new = active ? GST_ACTIVATE_PULL : GST_ACTIVATE_NONE;
pre_activate (pad, new);
gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. Original commit message from CVS: 2005-06-27 Andy Wingo <wingo@pobox.com> * gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. * gst/gsttrace.c (gst_alloc_trace_list_sorted): New helper, returns a sorted copy of the trace list. (gst_alloc_trace_print_live): New API, only prints traces with live objects. Sort the list. (gst_alloc_trace_print_all): Sort the list. (gst_alloc_trace_print): Align columns. * gst/elements/gstttypefindelement.c: * gst/elements/gsttee.c: * gst/base/gstbasesrc.c: * gst/base/gstbasesink.c: * gst/base/gstbasetransform.c: * gst/gstqueue.c: Adapt for pad activation changes. * gst/gstpipeline.c (gst_pipeline_init): Unref after parenting sched. (gst_pipeline_dispose): Drop ref on sched. * gst/gstpad.c (gst_pad_init): Set the default activate func. (gst_pad_activate_default): Push mode by default. (pre_activate_switch, post_activate_switch): New stubs, things to do before and after switching activation modes on pads. (gst_pad_set_active): Take a boolean and not a mode, dispatch to the pad's activate function to choose which mode to activate. Shortcut on deactivation and call the right function directly. (gst_pad_activate_pull): New API, (de)activates a pad in pull mode. (gst_pad_activate_push): New API, same for push mode. (gst_pad_set_activate_function) (gst_pad_set_activatepull_function) (gst_pad_set_activatepush_function): Setters for new API. * gst/gstminiobject.c (gst_mini_object_new, gst_mini_object_free): Trace all miniobjects. (gst_mini_object_make_writable): Unref the arg if we copy, like gst_caps_make_writable. * gst/gstmessage.c (_gst_message_initialize): No trace init. * gst/gstghostpad.c (gst_proxy_pad_do_activate) (gst_proxy_pad_do_activatepull, gst_proxy_pad_do_activatepush): Adapt for new pad API. * gst/gstevent.c (_gst_event_initialize): Don't initialize trace. * gst/gstelement.h: * gst/gstelement.c (gst_element_iterate_src_pads) (gst_element_iterate_sink_pads): New API functions. * gst/gstelement.c (iterator_fold_with_resync): New utility, should fold into gstiterator.c in some form. (gst_element_pads_activate): Simplified via use of fold and delegation of decisions to gstpad->activate. * gst/gstbus.c (gst_bus_source_finalize): Set the bus to NULL, help in debugging. * gst/gstbuffer.c (_gst_buffer_initialize): Ref the buffer type class once in init, like gstmessage. Didn't run into this issue but it seems correct. Don't initialize a trace, gstminiobject does that. * check/pipelines/simple_launch_lines.c (test_stop_from_app): New test, runs fakesrc ! fakesink, stopping on ::handoff via a message to the bus. (assert_live_count): New util function, uses alloc traces to check cleanup. * check/gst/gstghostpad.c (test_ghost_pads): More refcount checks. To be modified when unlink drops the internal pad.
2005-06-27 18:35:05 +00:00
if (GST_PAD_ACTIVATEPULLFUNC (pad)) {
if (G_UNLIKELY (!GST_PAD_ACTIVATEPULLFUNC (pad) (pad, active)))
gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. Original commit message from CVS: 2005-06-27 Andy Wingo <wingo@pobox.com> * gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. * gst/gsttrace.c (gst_alloc_trace_list_sorted): New helper, returns a sorted copy of the trace list. (gst_alloc_trace_print_live): New API, only prints traces with live objects. Sort the list. (gst_alloc_trace_print_all): Sort the list. (gst_alloc_trace_print): Align columns. * gst/elements/gstttypefindelement.c: * gst/elements/gsttee.c: * gst/base/gstbasesrc.c: * gst/base/gstbasesink.c: * gst/base/gstbasetransform.c: * gst/gstqueue.c: Adapt for pad activation changes. * gst/gstpipeline.c (gst_pipeline_init): Unref after parenting sched. (gst_pipeline_dispose): Drop ref on sched. * gst/gstpad.c (gst_pad_init): Set the default activate func. (gst_pad_activate_default): Push mode by default. (pre_activate_switch, post_activate_switch): New stubs, things to do before and after switching activation modes on pads. (gst_pad_set_active): Take a boolean and not a mode, dispatch to the pad's activate function to choose which mode to activate. Shortcut on deactivation and call the right function directly. (gst_pad_activate_pull): New API, (de)activates a pad in pull mode. (gst_pad_activate_push): New API, same for push mode. (gst_pad_set_activate_function) (gst_pad_set_activatepull_function) (gst_pad_set_activatepush_function): Setters for new API. * gst/gstminiobject.c (gst_mini_object_new, gst_mini_object_free): Trace all miniobjects. (gst_mini_object_make_writable): Unref the arg if we copy, like gst_caps_make_writable. * gst/gstmessage.c (_gst_message_initialize): No trace init. * gst/gstghostpad.c (gst_proxy_pad_do_activate) (gst_proxy_pad_do_activatepull, gst_proxy_pad_do_activatepush): Adapt for new pad API. * gst/gstevent.c (_gst_event_initialize): Don't initialize trace. * gst/gstelement.h: * gst/gstelement.c (gst_element_iterate_src_pads) (gst_element_iterate_sink_pads): New API functions. * gst/gstelement.c (iterator_fold_with_resync): New utility, should fold into gstiterator.c in some form. (gst_element_pads_activate): Simplified via use of fold and delegation of decisions to gstpad->activate. * gst/gstbus.c (gst_bus_source_finalize): Set the bus to NULL, help in debugging. * gst/gstbuffer.c (_gst_buffer_initialize): Ref the buffer type class once in init, like gstmessage. Didn't run into this issue but it seems correct. Don't initialize a trace, gstminiobject does that. * check/pipelines/simple_launch_lines.c (test_stop_from_app): New test, runs fakesrc ! fakesink, stopping on ::handoff via a message to the bus. (assert_live_count): New util function, uses alloc traces to check cleanup. * check/gst/gstghostpad.c (test_ghost_pads): More refcount checks. To be modified when unlink drops the internal pad.
2005-06-27 18:35:05 +00:00
goto failure;
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
} else {
gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. Original commit message from CVS: 2005-06-27 Andy Wingo <wingo@pobox.com> * gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. * gst/gsttrace.c (gst_alloc_trace_list_sorted): New helper, returns a sorted copy of the trace list. (gst_alloc_trace_print_live): New API, only prints traces with live objects. Sort the list. (gst_alloc_trace_print_all): Sort the list. (gst_alloc_trace_print): Align columns. * gst/elements/gstttypefindelement.c: * gst/elements/gsttee.c: * gst/base/gstbasesrc.c: * gst/base/gstbasesink.c: * gst/base/gstbasetransform.c: * gst/gstqueue.c: Adapt for pad activation changes. * gst/gstpipeline.c (gst_pipeline_init): Unref after parenting sched. (gst_pipeline_dispose): Drop ref on sched. * gst/gstpad.c (gst_pad_init): Set the default activate func. (gst_pad_activate_default): Push mode by default. (pre_activate_switch, post_activate_switch): New stubs, things to do before and after switching activation modes on pads. (gst_pad_set_active): Take a boolean and not a mode, dispatch to the pad's activate function to choose which mode to activate. Shortcut on deactivation and call the right function directly. (gst_pad_activate_pull): New API, (de)activates a pad in pull mode. (gst_pad_activate_push): New API, same for push mode. (gst_pad_set_activate_function) (gst_pad_set_activatepull_function) (gst_pad_set_activatepush_function): Setters for new API. * gst/gstminiobject.c (gst_mini_object_new, gst_mini_object_free): Trace all miniobjects. (gst_mini_object_make_writable): Unref the arg if we copy, like gst_caps_make_writable. * gst/gstmessage.c (_gst_message_initialize): No trace init. * gst/gstghostpad.c (gst_proxy_pad_do_activate) (gst_proxy_pad_do_activatepull, gst_proxy_pad_do_activatepush): Adapt for new pad API. * gst/gstevent.c (_gst_event_initialize): Don't initialize trace. * gst/gstelement.h: * gst/gstelement.c (gst_element_iterate_src_pads) (gst_element_iterate_sink_pads): New API functions. * gst/gstelement.c (iterator_fold_with_resync): New utility, should fold into gstiterator.c in some form. (gst_element_pads_activate): Simplified via use of fold and delegation of decisions to gstpad->activate. * gst/gstbus.c (gst_bus_source_finalize): Set the bus to NULL, help in debugging. * gst/gstbuffer.c (_gst_buffer_initialize): Ref the buffer type class once in init, like gstmessage. Didn't run into this issue but it seems correct. Don't initialize a trace, gstminiobject does that. * check/pipelines/simple_launch_lines.c (test_stop_from_app): New test, runs fakesrc ! fakesink, stopping on ::handoff via a message to the bus. (assert_live_count): New util function, uses alloc traces to check cleanup. * check/gst/gstghostpad.c (test_ghost_pads): More refcount checks. To be modified when unlink drops the internal pad.
2005-06-27 18:35:05 +00:00
/* can happen for sinks of passthrough elements */
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
}
post_activate (pad, new);
GST_CAT_DEBUG_OBJECT (GST_CAT_PADS, pad, "%s in pull mode",
active ? "activated" : "deactivated");
return TRUE;
gst/: Simplify pad activation. Original commit message from CVS: * gst/base/Makefile.am: * gst/base/README: * gst/base/gstbasesink.c: (gst_basesink_get_type), (gst_basesink_base_init), (gst_basesink_class_init), (gst_basesink_pad_getcaps), (gst_basesink_init), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesink.h: * gst/base/gstbasetransform.c: (gst_base_transform_get_type), (gst_base_transform_base_init), (gst_base_transform_finalize), (gst_base_transform_class_init), (gst_base_transform_init), (gst_base_transform_proxy_getcaps), (gst_base_transform_setcaps), (gst_base_transform_event), (gst_base_transform_getrange), (gst_base_transform_chain), (gst_base_transform_handle_buffer), (gst_base_transform_set_property), (gst_base_transform_get_property), (gst_base_transform_sink_activate), (gst_base_transform_src_activate), (gst_base_transform_change_state): * gst/base/gstbasetransform.h: * gst/elements/gstidentity.c: (gst_identity_finalize), (gst_identity_class_init), (gst_identity_init), (gst_identity_event), (gst_identity_check_perfect), (gst_identity_transform), (gst_identity_set_property), (gst_identity_get_property), (gst_identity_change_state): * gst/elements/gstidentity.h: * gst/gstelement.c: (gst_element_get_state_func), (gst_element_lost_state), (gst_element_pads_activate): * gst/gstpad.c: (gst_pad_set_active), (gst_pad_peer_set_active), (gst_pad_check_pull_range), (gst_pad_pull_range): * gst/gstpad.h: Simplify pad activation. Added function to check if pull_range can be performed. Error out when pulling inactive or flushing pads. Removed const from refcounted types as it does not make sense. Simplify pad templates in basesink Added base class for simple 1-to-1 transforms. Make identity subclass the base transform.
2005-03-29 16:18:12 +00:00
was_ok:
{
gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. Original commit message from CVS: 2005-06-27 Andy Wingo <wingo@pobox.com> * gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. * gst/gsttrace.c (gst_alloc_trace_list_sorted): New helper, returns a sorted copy of the trace list. (gst_alloc_trace_print_live): New API, only prints traces with live objects. Sort the list. (gst_alloc_trace_print_all): Sort the list. (gst_alloc_trace_print): Align columns. * gst/elements/gstttypefindelement.c: * gst/elements/gsttee.c: * gst/base/gstbasesrc.c: * gst/base/gstbasesink.c: * gst/base/gstbasetransform.c: * gst/gstqueue.c: Adapt for pad activation changes. * gst/gstpipeline.c (gst_pipeline_init): Unref after parenting sched. (gst_pipeline_dispose): Drop ref on sched. * gst/gstpad.c (gst_pad_init): Set the default activate func. (gst_pad_activate_default): Push mode by default. (pre_activate_switch, post_activate_switch): New stubs, things to do before and after switching activation modes on pads. (gst_pad_set_active): Take a boolean and not a mode, dispatch to the pad's activate function to choose which mode to activate. Shortcut on deactivation and call the right function directly. (gst_pad_activate_pull): New API, (de)activates a pad in pull mode. (gst_pad_activate_push): New API, same for push mode. (gst_pad_set_activate_function) (gst_pad_set_activatepull_function) (gst_pad_set_activatepush_function): Setters for new API. * gst/gstminiobject.c (gst_mini_object_new, gst_mini_object_free): Trace all miniobjects. (gst_mini_object_make_writable): Unref the arg if we copy, like gst_caps_make_writable. * gst/gstmessage.c (_gst_message_initialize): No trace init. * gst/gstghostpad.c (gst_proxy_pad_do_activate) (gst_proxy_pad_do_activatepull, gst_proxy_pad_do_activatepush): Adapt for new pad API. * gst/gstevent.c (_gst_event_initialize): Don't initialize trace. * gst/gstelement.h: * gst/gstelement.c (gst_element_iterate_src_pads) (gst_element_iterate_sink_pads): New API functions. * gst/gstelement.c (iterator_fold_with_resync): New utility, should fold into gstiterator.c in some form. (gst_element_pads_activate): Simplified via use of fold and delegation of decisions to gstpad->activate. * gst/gstbus.c (gst_bus_source_finalize): Set the bus to NULL, help in debugging. * gst/gstbuffer.c (_gst_buffer_initialize): Ref the buffer type class once in init, like gstmessage. Didn't run into this issue but it seems correct. Don't initialize a trace, gstminiobject does that. * check/pipelines/simple_launch_lines.c (test_stop_from_app): New test, runs fakesrc ! fakesink, stopping on ::handoff via a message to the bus. (assert_live_count): New util function, uses alloc traces to check cleanup. * check/gst/gstghostpad.c (test_ghost_pads): More refcount checks. To be modified when unlink drops the internal pad.
2005-06-27 18:35:05 +00:00
GST_CAT_DEBUG_OBJECT (GST_CAT_PADS, pad, "already %s in pull mode",
active ? "activated" : "deactivated");
gst/: Simplify pad activation. Original commit message from CVS: * gst/base/Makefile.am: * gst/base/README: * gst/base/gstbasesink.c: (gst_basesink_get_type), (gst_basesink_base_init), (gst_basesink_class_init), (gst_basesink_pad_getcaps), (gst_basesink_init), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesink.h: * gst/base/gstbasetransform.c: (gst_base_transform_get_type), (gst_base_transform_base_init), (gst_base_transform_finalize), (gst_base_transform_class_init), (gst_base_transform_init), (gst_base_transform_proxy_getcaps), (gst_base_transform_setcaps), (gst_base_transform_event), (gst_base_transform_getrange), (gst_base_transform_chain), (gst_base_transform_handle_buffer), (gst_base_transform_set_property), (gst_base_transform_get_property), (gst_base_transform_sink_activate), (gst_base_transform_src_activate), (gst_base_transform_change_state): * gst/base/gstbasetransform.h: * gst/elements/gstidentity.c: (gst_identity_finalize), (gst_identity_class_init), (gst_identity_init), (gst_identity_event), (gst_identity_check_perfect), (gst_identity_transform), (gst_identity_set_property), (gst_identity_get_property), (gst_identity_change_state): * gst/elements/gstidentity.h: * gst/gstelement.c: (gst_element_get_state_func), (gst_element_lost_state), (gst_element_pads_activate): * gst/gstpad.c: (gst_pad_set_active), (gst_pad_peer_set_active), (gst_pad_check_pull_range), (gst_pad_pull_range): * gst/gstpad.h: Simplify pad activation. Added function to check if pull_range can be performed. Error out when pulling inactive or flushing pads. Removed const from refcounted types as it does not make sense. Simplify pad templates in basesink Added base class for simple 1-to-1 transforms. Make identity subclass the base transform.
2005-03-29 16:18:12 +00:00
return TRUE;
}
deactivate_failed:
{
GST_CAT_DEBUG_OBJECT (GST_CAT_PADS, pad,
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
"failed to %s in switch to pull from mode %d",
(active ? "activate" : "deactivate"), old);
return FALSE;
}
peer_failed:
{
GST_OBJECT_LOCK (peer);
GST_CAT_DEBUG_OBJECT (GST_CAT_PADS, pad,
"activate_pull on peer (%s:%s) failed", GST_DEBUG_PAD_NAME (peer));
GST_OBJECT_UNLOCK (peer);
gst_object_unref (peer);
return FALSE;
}
gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. Original commit message from CVS: 2005-06-27 Andy Wingo <wingo@pobox.com> * gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. * gst/gsttrace.c (gst_alloc_trace_list_sorted): New helper, returns a sorted copy of the trace list. (gst_alloc_trace_print_live): New API, only prints traces with live objects. Sort the list. (gst_alloc_trace_print_all): Sort the list. (gst_alloc_trace_print): Align columns. * gst/elements/gstttypefindelement.c: * gst/elements/gsttee.c: * gst/base/gstbasesrc.c: * gst/base/gstbasesink.c: * gst/base/gstbasetransform.c: * gst/gstqueue.c: Adapt for pad activation changes. * gst/gstpipeline.c (gst_pipeline_init): Unref after parenting sched. (gst_pipeline_dispose): Drop ref on sched. * gst/gstpad.c (gst_pad_init): Set the default activate func. (gst_pad_activate_default): Push mode by default. (pre_activate_switch, post_activate_switch): New stubs, things to do before and after switching activation modes on pads. (gst_pad_set_active): Take a boolean and not a mode, dispatch to the pad's activate function to choose which mode to activate. Shortcut on deactivation and call the right function directly. (gst_pad_activate_pull): New API, (de)activates a pad in pull mode. (gst_pad_activate_push): New API, same for push mode. (gst_pad_set_activate_function) (gst_pad_set_activatepull_function) (gst_pad_set_activatepush_function): Setters for new API. * gst/gstminiobject.c (gst_mini_object_new, gst_mini_object_free): Trace all miniobjects. (gst_mini_object_make_writable): Unref the arg if we copy, like gst_caps_make_writable. * gst/gstmessage.c (_gst_message_initialize): No trace init. * gst/gstghostpad.c (gst_proxy_pad_do_activate) (gst_proxy_pad_do_activatepull, gst_proxy_pad_do_activatepush): Adapt for new pad API. * gst/gstevent.c (_gst_event_initialize): Don't initialize trace. * gst/gstelement.h: * gst/gstelement.c (gst_element_iterate_src_pads) (gst_element_iterate_sink_pads): New API functions. * gst/gstelement.c (iterator_fold_with_resync): New utility, should fold into gstiterator.c in some form. (gst_element_pads_activate): Simplified via use of fold and delegation of decisions to gstpad->activate. * gst/gstbus.c (gst_bus_source_finalize): Set the bus to NULL, help in debugging. * gst/gstbuffer.c (_gst_buffer_initialize): Ref the buffer type class once in init, like gstmessage. Didn't run into this issue but it seems correct. Don't initialize a trace, gstminiobject does that. * check/pipelines/simple_launch_lines.c (test_stop_from_app): New test, runs fakesrc ! fakesink, stopping on ::handoff via a message to the bus. (assert_live_count): New util function, uses alloc traces to check cleanup. * check/gst/gstghostpad.c (test_ghost_pads): More refcount checks. To be modified when unlink drops the internal pad.
2005-06-27 18:35:05 +00:00
failure:
{
GST_OBJECT_LOCK (pad);
gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. Original commit message from CVS: 2005-06-27 Andy Wingo <wingo@pobox.com> * gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. * gst/gsttrace.c (gst_alloc_trace_list_sorted): New helper, returns a sorted copy of the trace list. (gst_alloc_trace_print_live): New API, only prints traces with live objects. Sort the list. (gst_alloc_trace_print_all): Sort the list. (gst_alloc_trace_print): Align columns. * gst/elements/gstttypefindelement.c: * gst/elements/gsttee.c: * gst/base/gstbasesrc.c: * gst/base/gstbasesink.c: * gst/base/gstbasetransform.c: * gst/gstqueue.c: Adapt for pad activation changes. * gst/gstpipeline.c (gst_pipeline_init): Unref after parenting sched. (gst_pipeline_dispose): Drop ref on sched. * gst/gstpad.c (gst_pad_init): Set the default activate func. (gst_pad_activate_default): Push mode by default. (pre_activate_switch, post_activate_switch): New stubs, things to do before and after switching activation modes on pads. (gst_pad_set_active): Take a boolean and not a mode, dispatch to the pad's activate function to choose which mode to activate. Shortcut on deactivation and call the right function directly. (gst_pad_activate_pull): New API, (de)activates a pad in pull mode. (gst_pad_activate_push): New API, same for push mode. (gst_pad_set_activate_function) (gst_pad_set_activatepull_function) (gst_pad_set_activatepush_function): Setters for new API. * gst/gstminiobject.c (gst_mini_object_new, gst_mini_object_free): Trace all miniobjects. (gst_mini_object_make_writable): Unref the arg if we copy, like gst_caps_make_writable. * gst/gstmessage.c (_gst_message_initialize): No trace init. * gst/gstghostpad.c (gst_proxy_pad_do_activate) (gst_proxy_pad_do_activatepull, gst_proxy_pad_do_activatepush): Adapt for new pad API. * gst/gstevent.c (_gst_event_initialize): Don't initialize trace. * gst/gstelement.h: * gst/gstelement.c (gst_element_iterate_src_pads) (gst_element_iterate_sink_pads): New API functions. * gst/gstelement.c (iterator_fold_with_resync): New utility, should fold into gstiterator.c in some form. (gst_element_pads_activate): Simplified via use of fold and delegation of decisions to gstpad->activate. * gst/gstbus.c (gst_bus_source_finalize): Set the bus to NULL, help in debugging. * gst/gstbuffer.c (_gst_buffer_initialize): Ref the buffer type class once in init, like gstmessage. Didn't run into this issue but it seems correct. Don't initialize a trace, gstminiobject does that. * check/pipelines/simple_launch_lines.c (test_stop_from_app): New test, runs fakesrc ! fakesink, stopping on ::handoff via a message to the bus. (assert_live_count): New util function, uses alloc traces to check cleanup. * check/gst/gstghostpad.c (test_ghost_pads): More refcount checks. To be modified when unlink drops the internal pad.
2005-06-27 18:35:05 +00:00
GST_CAT_INFO_OBJECT (GST_CAT_PADS, pad, "failed to %s in pull mode",
active ? "activate" : "deactivate");
GST_PAD_SET_FLUSHING (pad);
GST_PAD_ACTIVATE_MODE (pad) = old;
GST_OBJECT_UNLOCK (pad);
return FALSE;
}
}
gst/: Simplify pad activation. Original commit message from CVS: * gst/base/Makefile.am: * gst/base/README: * gst/base/gstbasesink.c: (gst_basesink_get_type), (gst_basesink_base_init), (gst_basesink_class_init), (gst_basesink_pad_getcaps), (gst_basesink_init), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesink.h: * gst/base/gstbasetransform.c: (gst_base_transform_get_type), (gst_base_transform_base_init), (gst_base_transform_finalize), (gst_base_transform_class_init), (gst_base_transform_init), (gst_base_transform_proxy_getcaps), (gst_base_transform_setcaps), (gst_base_transform_event), (gst_base_transform_getrange), (gst_base_transform_chain), (gst_base_transform_handle_buffer), (gst_base_transform_set_property), (gst_base_transform_get_property), (gst_base_transform_sink_activate), (gst_base_transform_src_activate), (gst_base_transform_change_state): * gst/base/gstbasetransform.h: * gst/elements/gstidentity.c: (gst_identity_finalize), (gst_identity_class_init), (gst_identity_init), (gst_identity_event), (gst_identity_check_perfect), (gst_identity_transform), (gst_identity_set_property), (gst_identity_get_property), (gst_identity_change_state): * gst/elements/gstidentity.h: * gst/gstelement.c: (gst_element_get_state_func), (gst_element_lost_state), (gst_element_pads_activate): * gst/gstpad.c: (gst_pad_set_active), (gst_pad_peer_set_active), (gst_pad_check_pull_range), (gst_pad_pull_range): * gst/gstpad.h: Simplify pad activation. Added function to check if pull_range can be performed. Error out when pulling inactive or flushing pads. Removed const from refcounted types as it does not make sense. Simplify pad templates in basesink Added base class for simple 1-to-1 transforms. Make identity subclass the base transform.
2005-03-29 16:18:12 +00:00
/**
gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. Original commit message from CVS: 2005-06-27 Andy Wingo <wingo@pobox.com> * gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. * gst/gsttrace.c (gst_alloc_trace_list_sorted): New helper, returns a sorted copy of the trace list. (gst_alloc_trace_print_live): New API, only prints traces with live objects. Sort the list. (gst_alloc_trace_print_all): Sort the list. (gst_alloc_trace_print): Align columns. * gst/elements/gstttypefindelement.c: * gst/elements/gsttee.c: * gst/base/gstbasesrc.c: * gst/base/gstbasesink.c: * gst/base/gstbasetransform.c: * gst/gstqueue.c: Adapt for pad activation changes. * gst/gstpipeline.c (gst_pipeline_init): Unref after parenting sched. (gst_pipeline_dispose): Drop ref on sched. * gst/gstpad.c (gst_pad_init): Set the default activate func. (gst_pad_activate_default): Push mode by default. (pre_activate_switch, post_activate_switch): New stubs, things to do before and after switching activation modes on pads. (gst_pad_set_active): Take a boolean and not a mode, dispatch to the pad's activate function to choose which mode to activate. Shortcut on deactivation and call the right function directly. (gst_pad_activate_pull): New API, (de)activates a pad in pull mode. (gst_pad_activate_push): New API, same for push mode. (gst_pad_set_activate_function) (gst_pad_set_activatepull_function) (gst_pad_set_activatepush_function): Setters for new API. * gst/gstminiobject.c (gst_mini_object_new, gst_mini_object_free): Trace all miniobjects. (gst_mini_object_make_writable): Unref the arg if we copy, like gst_caps_make_writable. * gst/gstmessage.c (_gst_message_initialize): No trace init. * gst/gstghostpad.c (gst_proxy_pad_do_activate) (gst_proxy_pad_do_activatepull, gst_proxy_pad_do_activatepush): Adapt for new pad API. * gst/gstevent.c (_gst_event_initialize): Don't initialize trace. * gst/gstelement.h: * gst/gstelement.c (gst_element_iterate_src_pads) (gst_element_iterate_sink_pads): New API functions. * gst/gstelement.c (iterator_fold_with_resync): New utility, should fold into gstiterator.c in some form. (gst_element_pads_activate): Simplified via use of fold and delegation of decisions to gstpad->activate. * gst/gstbus.c (gst_bus_source_finalize): Set the bus to NULL, help in debugging. * gst/gstbuffer.c (_gst_buffer_initialize): Ref the buffer type class once in init, like gstmessage. Didn't run into this issue but it seems correct. Don't initialize a trace, gstminiobject does that. * check/pipelines/simple_launch_lines.c (test_stop_from_app): New test, runs fakesrc ! fakesink, stopping on ::handoff via a message to the bus. (assert_live_count): New util function, uses alloc traces to check cleanup. * check/gst/gstghostpad.c (test_ghost_pads): More refcount checks. To be modified when unlink drops the internal pad.
2005-06-27 18:35:05 +00:00
* gst_pad_activate_push:
* @pad: the #GstPad to activate or deactivate.
* @active: whether the pad should be active or not.
gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. Original commit message from CVS: 2005-06-27 Andy Wingo <wingo@pobox.com> * gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. * gst/gsttrace.c (gst_alloc_trace_list_sorted): New helper, returns a sorted copy of the trace list. (gst_alloc_trace_print_live): New API, only prints traces with live objects. Sort the list. (gst_alloc_trace_print_all): Sort the list. (gst_alloc_trace_print): Align columns. * gst/elements/gstttypefindelement.c: * gst/elements/gsttee.c: * gst/base/gstbasesrc.c: * gst/base/gstbasesink.c: * gst/base/gstbasetransform.c: * gst/gstqueue.c: Adapt for pad activation changes. * gst/gstpipeline.c (gst_pipeline_init): Unref after parenting sched. (gst_pipeline_dispose): Drop ref on sched. * gst/gstpad.c (gst_pad_init): Set the default activate func. (gst_pad_activate_default): Push mode by default. (pre_activate_switch, post_activate_switch): New stubs, things to do before and after switching activation modes on pads. (gst_pad_set_active): Take a boolean and not a mode, dispatch to the pad's activate function to choose which mode to activate. Shortcut on deactivation and call the right function directly. (gst_pad_activate_pull): New API, (de)activates a pad in pull mode. (gst_pad_activate_push): New API, same for push mode. (gst_pad_set_activate_function) (gst_pad_set_activatepull_function) (gst_pad_set_activatepush_function): Setters for new API. * gst/gstminiobject.c (gst_mini_object_new, gst_mini_object_free): Trace all miniobjects. (gst_mini_object_make_writable): Unref the arg if we copy, like gst_caps_make_writable. * gst/gstmessage.c (_gst_message_initialize): No trace init. * gst/gstghostpad.c (gst_proxy_pad_do_activate) (gst_proxy_pad_do_activatepull, gst_proxy_pad_do_activatepush): Adapt for new pad API. * gst/gstevent.c (_gst_event_initialize): Don't initialize trace. * gst/gstelement.h: * gst/gstelement.c (gst_element_iterate_src_pads) (gst_element_iterate_sink_pads): New API functions. * gst/gstelement.c (iterator_fold_with_resync): New utility, should fold into gstiterator.c in some form. (gst_element_pads_activate): Simplified via use of fold and delegation of decisions to gstpad->activate. * gst/gstbus.c (gst_bus_source_finalize): Set the bus to NULL, help in debugging. * gst/gstbuffer.c (_gst_buffer_initialize): Ref the buffer type class once in init, like gstmessage. Didn't run into this issue but it seems correct. Don't initialize a trace, gstminiobject does that. * check/pipelines/simple_launch_lines.c (test_stop_from_app): New test, runs fakesrc ! fakesink, stopping on ::handoff via a message to the bus. (assert_live_count): New util function, uses alloc traces to check cleanup. * check/gst/gstghostpad.c (test_ghost_pads): More refcount checks. To be modified when unlink drops the internal pad.
2005-06-27 18:35:05 +00:00
*
* Activates or deactivates the given pad in push mode via dispatching to the
* pad's activatepushfunc. For use from within pad activation functions only.
gst/: Simplify pad activation. Original commit message from CVS: * gst/base/Makefile.am: * gst/base/README: * gst/base/gstbasesink.c: (gst_basesink_get_type), (gst_basesink_base_init), (gst_basesink_class_init), (gst_basesink_pad_getcaps), (gst_basesink_init), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesink.h: * gst/base/gstbasetransform.c: (gst_base_transform_get_type), (gst_base_transform_base_init), (gst_base_transform_finalize), (gst_base_transform_class_init), (gst_base_transform_init), (gst_base_transform_proxy_getcaps), (gst_base_transform_setcaps), (gst_base_transform_event), (gst_base_transform_getrange), (gst_base_transform_chain), (gst_base_transform_handle_buffer), (gst_base_transform_set_property), (gst_base_transform_get_property), (gst_base_transform_sink_activate), (gst_base_transform_src_activate), (gst_base_transform_change_state): * gst/base/gstbasetransform.h: * gst/elements/gstidentity.c: (gst_identity_finalize), (gst_identity_class_init), (gst_identity_init), (gst_identity_event), (gst_identity_check_perfect), (gst_identity_transform), (gst_identity_set_property), (gst_identity_get_property), (gst_identity_change_state): * gst/elements/gstidentity.h: * gst/gstelement.c: (gst_element_get_state_func), (gst_element_lost_state), (gst_element_pads_activate): * gst/gstpad.c: (gst_pad_set_active), (gst_pad_peer_set_active), (gst_pad_check_pull_range), (gst_pad_pull_range): * gst/gstpad.h: Simplify pad activation. Added function to check if pull_range can be performed. Error out when pulling inactive or flushing pads. Removed const from refcounted types as it does not make sense. Simplify pad templates in basesink Added base class for simple 1-to-1 transforms. Make identity subclass the base transform.
2005-03-29 16:18:12 +00:00
*
gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. Original commit message from CVS: 2005-06-27 Andy Wingo <wingo@pobox.com> * gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. * gst/gsttrace.c (gst_alloc_trace_list_sorted): New helper, returns a sorted copy of the trace list. (gst_alloc_trace_print_live): New API, only prints traces with live objects. Sort the list. (gst_alloc_trace_print_all): Sort the list. (gst_alloc_trace_print): Align columns. * gst/elements/gstttypefindelement.c: * gst/elements/gsttee.c: * gst/base/gstbasesrc.c: * gst/base/gstbasesink.c: * gst/base/gstbasetransform.c: * gst/gstqueue.c: Adapt for pad activation changes. * gst/gstpipeline.c (gst_pipeline_init): Unref after parenting sched. (gst_pipeline_dispose): Drop ref on sched. * gst/gstpad.c (gst_pad_init): Set the default activate func. (gst_pad_activate_default): Push mode by default. (pre_activate_switch, post_activate_switch): New stubs, things to do before and after switching activation modes on pads. (gst_pad_set_active): Take a boolean and not a mode, dispatch to the pad's activate function to choose which mode to activate. Shortcut on deactivation and call the right function directly. (gst_pad_activate_pull): New API, (de)activates a pad in pull mode. (gst_pad_activate_push): New API, same for push mode. (gst_pad_set_activate_function) (gst_pad_set_activatepull_function) (gst_pad_set_activatepush_function): Setters for new API. * gst/gstminiobject.c (gst_mini_object_new, gst_mini_object_free): Trace all miniobjects. (gst_mini_object_make_writable): Unref the arg if we copy, like gst_caps_make_writable. * gst/gstmessage.c (_gst_message_initialize): No trace init. * gst/gstghostpad.c (gst_proxy_pad_do_activate) (gst_proxy_pad_do_activatepull, gst_proxy_pad_do_activatepush): Adapt for new pad API. * gst/gstevent.c (_gst_event_initialize): Don't initialize trace. * gst/gstelement.h: * gst/gstelement.c (gst_element_iterate_src_pads) (gst_element_iterate_sink_pads): New API functions. * gst/gstelement.c (iterator_fold_with_resync): New utility, should fold into gstiterator.c in some form. (gst_element_pads_activate): Simplified via use of fold and delegation of decisions to gstpad->activate. * gst/gstbus.c (gst_bus_source_finalize): Set the bus to NULL, help in debugging. * gst/gstbuffer.c (_gst_buffer_initialize): Ref the buffer type class once in init, like gstmessage. Didn't run into this issue but it seems correct. Don't initialize a trace, gstminiobject does that. * check/pipelines/simple_launch_lines.c (test_stop_from_app): New test, runs fakesrc ! fakesink, stopping on ::handoff via a message to the bus. (assert_live_count): New util function, uses alloc traces to check cleanup. * check/gst/gstghostpad.c (test_ghost_pads): More refcount checks. To be modified when unlink drops the internal pad.
2005-06-27 18:35:05 +00:00
* If you don't know what this is, you probably don't want to call it.
gst/: Simplify pad activation. Original commit message from CVS: * gst/base/Makefile.am: * gst/base/README: * gst/base/gstbasesink.c: (gst_basesink_get_type), (gst_basesink_base_init), (gst_basesink_class_init), (gst_basesink_pad_getcaps), (gst_basesink_init), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesink.h: * gst/base/gstbasetransform.c: (gst_base_transform_get_type), (gst_base_transform_base_init), (gst_base_transform_finalize), (gst_base_transform_class_init), (gst_base_transform_init), (gst_base_transform_proxy_getcaps), (gst_base_transform_setcaps), (gst_base_transform_event), (gst_base_transform_getrange), (gst_base_transform_chain), (gst_base_transform_handle_buffer), (gst_base_transform_set_property), (gst_base_transform_get_property), (gst_base_transform_sink_activate), (gst_base_transform_src_activate), (gst_base_transform_change_state): * gst/base/gstbasetransform.h: * gst/elements/gstidentity.c: (gst_identity_finalize), (gst_identity_class_init), (gst_identity_init), (gst_identity_event), (gst_identity_check_perfect), (gst_identity_transform), (gst_identity_set_property), (gst_identity_get_property), (gst_identity_change_state): * gst/elements/gstidentity.h: * gst/gstelement.c: (gst_element_get_state_func), (gst_element_lost_state), (gst_element_pads_activate): * gst/gstpad.c: (gst_pad_set_active), (gst_pad_peer_set_active), (gst_pad_check_pull_range), (gst_pad_pull_range): * gst/gstpad.h: Simplify pad activation. Added function to check if pull_range can be performed. Error out when pulling inactive or flushing pads. Removed const from refcounted types as it does not make sense. Simplify pad templates in basesink Added base class for simple 1-to-1 transforms. Make identity subclass the base transform.
2005-03-29 16:18:12 +00:00
*
* Returns: %TRUE if the operation was successful.
gst/: Simplify pad activation. Original commit message from CVS: * gst/base/Makefile.am: * gst/base/README: * gst/base/gstbasesink.c: (gst_basesink_get_type), (gst_basesink_base_init), (gst_basesink_class_init), (gst_basesink_pad_getcaps), (gst_basesink_init), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesink.h: * gst/base/gstbasetransform.c: (gst_base_transform_get_type), (gst_base_transform_base_init), (gst_base_transform_finalize), (gst_base_transform_class_init), (gst_base_transform_init), (gst_base_transform_proxy_getcaps), (gst_base_transform_setcaps), (gst_base_transform_event), (gst_base_transform_getrange), (gst_base_transform_chain), (gst_base_transform_handle_buffer), (gst_base_transform_set_property), (gst_base_transform_get_property), (gst_base_transform_sink_activate), (gst_base_transform_src_activate), (gst_base_transform_change_state): * gst/base/gstbasetransform.h: * gst/elements/gstidentity.c: (gst_identity_finalize), (gst_identity_class_init), (gst_identity_init), (gst_identity_event), (gst_identity_check_perfect), (gst_identity_transform), (gst_identity_set_property), (gst_identity_get_property), (gst_identity_change_state): * gst/elements/gstidentity.h: * gst/gstelement.c: (gst_element_get_state_func), (gst_element_lost_state), (gst_element_pads_activate): * gst/gstpad.c: (gst_pad_set_active), (gst_pad_peer_set_active), (gst_pad_check_pull_range), (gst_pad_pull_range): * gst/gstpad.h: Simplify pad activation. Added function to check if pull_range can be performed. Error out when pulling inactive or flushing pads. Removed const from refcounted types as it does not make sense. Simplify pad templates in basesink Added base class for simple 1-to-1 transforms. Make identity subclass the base transform.
2005-03-29 16:18:12 +00:00
*
* MT safe.
*/
gboolean
gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. Original commit message from CVS: 2005-06-27 Andy Wingo <wingo@pobox.com> * gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. * gst/gsttrace.c (gst_alloc_trace_list_sorted): New helper, returns a sorted copy of the trace list. (gst_alloc_trace_print_live): New API, only prints traces with live objects. Sort the list. (gst_alloc_trace_print_all): Sort the list. (gst_alloc_trace_print): Align columns. * gst/elements/gstttypefindelement.c: * gst/elements/gsttee.c: * gst/base/gstbasesrc.c: * gst/base/gstbasesink.c: * gst/base/gstbasetransform.c: * gst/gstqueue.c: Adapt for pad activation changes. * gst/gstpipeline.c (gst_pipeline_init): Unref after parenting sched. (gst_pipeline_dispose): Drop ref on sched. * gst/gstpad.c (gst_pad_init): Set the default activate func. (gst_pad_activate_default): Push mode by default. (pre_activate_switch, post_activate_switch): New stubs, things to do before and after switching activation modes on pads. (gst_pad_set_active): Take a boolean and not a mode, dispatch to the pad's activate function to choose which mode to activate. Shortcut on deactivation and call the right function directly. (gst_pad_activate_pull): New API, (de)activates a pad in pull mode. (gst_pad_activate_push): New API, same for push mode. (gst_pad_set_activate_function) (gst_pad_set_activatepull_function) (gst_pad_set_activatepush_function): Setters for new API. * gst/gstminiobject.c (gst_mini_object_new, gst_mini_object_free): Trace all miniobjects. (gst_mini_object_make_writable): Unref the arg if we copy, like gst_caps_make_writable. * gst/gstmessage.c (_gst_message_initialize): No trace init. * gst/gstghostpad.c (gst_proxy_pad_do_activate) (gst_proxy_pad_do_activatepull, gst_proxy_pad_do_activatepush): Adapt for new pad API. * gst/gstevent.c (_gst_event_initialize): Don't initialize trace. * gst/gstelement.h: * gst/gstelement.c (gst_element_iterate_src_pads) (gst_element_iterate_sink_pads): New API functions. * gst/gstelement.c (iterator_fold_with_resync): New utility, should fold into gstiterator.c in some form. (gst_element_pads_activate): Simplified via use of fold and delegation of decisions to gstpad->activate. * gst/gstbus.c (gst_bus_source_finalize): Set the bus to NULL, help in debugging. * gst/gstbuffer.c (_gst_buffer_initialize): Ref the buffer type class once in init, like gstmessage. Didn't run into this issue but it seems correct. Don't initialize a trace, gstminiobject does that. * check/pipelines/simple_launch_lines.c (test_stop_from_app): New test, runs fakesrc ! fakesink, stopping on ::handoff via a message to the bus. (assert_live_count): New util function, uses alloc traces to check cleanup. * check/gst/gstghostpad.c (test_ghost_pads): More refcount checks. To be modified when unlink drops the internal pad.
2005-06-27 18:35:05 +00:00
gst_pad_activate_push (GstPad * pad, gboolean active)
gst/: Simplify pad activation. Original commit message from CVS: * gst/base/Makefile.am: * gst/base/README: * gst/base/gstbasesink.c: (gst_basesink_get_type), (gst_basesink_base_init), (gst_basesink_class_init), (gst_basesink_pad_getcaps), (gst_basesink_init), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesink.h: * gst/base/gstbasetransform.c: (gst_base_transform_get_type), (gst_base_transform_base_init), (gst_base_transform_finalize), (gst_base_transform_class_init), (gst_base_transform_init), (gst_base_transform_proxy_getcaps), (gst_base_transform_setcaps), (gst_base_transform_event), (gst_base_transform_getrange), (gst_base_transform_chain), (gst_base_transform_handle_buffer), (gst_base_transform_set_property), (gst_base_transform_get_property), (gst_base_transform_sink_activate), (gst_base_transform_src_activate), (gst_base_transform_change_state): * gst/base/gstbasetransform.h: * gst/elements/gstidentity.c: (gst_identity_finalize), (gst_identity_class_init), (gst_identity_init), (gst_identity_event), (gst_identity_check_perfect), (gst_identity_transform), (gst_identity_set_property), (gst_identity_get_property), (gst_identity_change_state): * gst/elements/gstidentity.h: * gst/gstelement.c: (gst_element_get_state_func), (gst_element_lost_state), (gst_element_pads_activate): * gst/gstpad.c: (gst_pad_set_active), (gst_pad_peer_set_active), (gst_pad_check_pull_range), (gst_pad_pull_range): * gst/gstpad.h: Simplify pad activation. Added function to check if pull_range can be performed. Error out when pulling inactive or flushing pads. Removed const from refcounted types as it does not make sense. Simplify pad templates in basesink Added base class for simple 1-to-1 transforms. Make identity subclass the base transform.
2005-03-29 16:18:12 +00:00
{
GstActivateMode old, new;
gst/: Simplify pad activation. Original commit message from CVS: * gst/base/Makefile.am: * gst/base/README: * gst/base/gstbasesink.c: (gst_basesink_get_type), (gst_basesink_base_init), (gst_basesink_class_init), (gst_basesink_pad_getcaps), (gst_basesink_init), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesink.h: * gst/base/gstbasetransform.c: (gst_base_transform_get_type), (gst_base_transform_base_init), (gst_base_transform_finalize), (gst_base_transform_class_init), (gst_base_transform_init), (gst_base_transform_proxy_getcaps), (gst_base_transform_setcaps), (gst_base_transform_event), (gst_base_transform_getrange), (gst_base_transform_chain), (gst_base_transform_handle_buffer), (gst_base_transform_set_property), (gst_base_transform_get_property), (gst_base_transform_sink_activate), (gst_base_transform_src_activate), (gst_base_transform_change_state): * gst/base/gstbasetransform.h: * gst/elements/gstidentity.c: (gst_identity_finalize), (gst_identity_class_init), (gst_identity_init), (gst_identity_event), (gst_identity_check_perfect), (gst_identity_transform), (gst_identity_set_property), (gst_identity_get_property), (gst_identity_change_state): * gst/elements/gstidentity.h: * gst/gstelement.c: (gst_element_get_state_func), (gst_element_lost_state), (gst_element_pads_activate): * gst/gstpad.c: (gst_pad_set_active), (gst_pad_peer_set_active), (gst_pad_check_pull_range), (gst_pad_pull_range): * gst/gstpad.h: Simplify pad activation. Added function to check if pull_range can be performed. Error out when pulling inactive or flushing pads. Removed const from refcounted types as it does not make sense. Simplify pad templates in basesink Added base class for simple 1-to-1 transforms. Make identity subclass the base transform.
2005-03-29 16:18:12 +00:00
gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. Original commit message from CVS: 2005-06-27 Andy Wingo <wingo@pobox.com> * gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. * gst/gsttrace.c (gst_alloc_trace_list_sorted): New helper, returns a sorted copy of the trace list. (gst_alloc_trace_print_live): New API, only prints traces with live objects. Sort the list. (gst_alloc_trace_print_all): Sort the list. (gst_alloc_trace_print): Align columns. * gst/elements/gstttypefindelement.c: * gst/elements/gsttee.c: * gst/base/gstbasesrc.c: * gst/base/gstbasesink.c: * gst/base/gstbasetransform.c: * gst/gstqueue.c: Adapt for pad activation changes. * gst/gstpipeline.c (gst_pipeline_init): Unref after parenting sched. (gst_pipeline_dispose): Drop ref on sched. * gst/gstpad.c (gst_pad_init): Set the default activate func. (gst_pad_activate_default): Push mode by default. (pre_activate_switch, post_activate_switch): New stubs, things to do before and after switching activation modes on pads. (gst_pad_set_active): Take a boolean and not a mode, dispatch to the pad's activate function to choose which mode to activate. Shortcut on deactivation and call the right function directly. (gst_pad_activate_pull): New API, (de)activates a pad in pull mode. (gst_pad_activate_push): New API, same for push mode. (gst_pad_set_activate_function) (gst_pad_set_activatepull_function) (gst_pad_set_activatepush_function): Setters for new API. * gst/gstminiobject.c (gst_mini_object_new, gst_mini_object_free): Trace all miniobjects. (gst_mini_object_make_writable): Unref the arg if we copy, like gst_caps_make_writable. * gst/gstmessage.c (_gst_message_initialize): No trace init. * gst/gstghostpad.c (gst_proxy_pad_do_activate) (gst_proxy_pad_do_activatepull, gst_proxy_pad_do_activatepush): Adapt for new pad API. * gst/gstevent.c (_gst_event_initialize): Don't initialize trace. * gst/gstelement.h: * gst/gstelement.c (gst_element_iterate_src_pads) (gst_element_iterate_sink_pads): New API functions. * gst/gstelement.c (iterator_fold_with_resync): New utility, should fold into gstiterator.c in some form. (gst_element_pads_activate): Simplified via use of fold and delegation of decisions to gstpad->activate. * gst/gstbus.c (gst_bus_source_finalize): Set the bus to NULL, help in debugging. * gst/gstbuffer.c (_gst_buffer_initialize): Ref the buffer type class once in init, like gstmessage. Didn't run into this issue but it seems correct. Don't initialize a trace, gstminiobject does that. * check/pipelines/simple_launch_lines.c (test_stop_from_app): New test, runs fakesrc ! fakesink, stopping on ::handoff via a message to the bus. (assert_live_count): New util function, uses alloc traces to check cleanup. * check/gst/gstghostpad.c (test_ghost_pads): More refcount checks. To be modified when unlink drops the internal pad.
2005-06-27 18:35:05 +00:00
g_return_val_if_fail (GST_IS_PAD (pad), FALSE);
GST_CAT_DEBUG_OBJECT (GST_CAT_PADS, pad, "trying to set %s in push mode",
active ? "activated" : "deactivated");
gst/: Simplify pad activation. Original commit message from CVS: * gst/base/Makefile.am: * gst/base/README: * gst/base/gstbasesink.c: (gst_basesink_get_type), (gst_basesink_base_init), (gst_basesink_class_init), (gst_basesink_pad_getcaps), (gst_basesink_init), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesink.h: * gst/base/gstbasetransform.c: (gst_base_transform_get_type), (gst_base_transform_base_init), (gst_base_transform_finalize), (gst_base_transform_class_init), (gst_base_transform_init), (gst_base_transform_proxy_getcaps), (gst_base_transform_setcaps), (gst_base_transform_event), (gst_base_transform_getrange), (gst_base_transform_chain), (gst_base_transform_handle_buffer), (gst_base_transform_set_property), (gst_base_transform_get_property), (gst_base_transform_sink_activate), (gst_base_transform_src_activate), (gst_base_transform_change_state): * gst/base/gstbasetransform.h: * gst/elements/gstidentity.c: (gst_identity_finalize), (gst_identity_class_init), (gst_identity_init), (gst_identity_event), (gst_identity_check_perfect), (gst_identity_transform), (gst_identity_set_property), (gst_identity_get_property), (gst_identity_change_state): * gst/elements/gstidentity.h: * gst/gstelement.c: (gst_element_get_state_func), (gst_element_lost_state), (gst_element_pads_activate): * gst/gstpad.c: (gst_pad_set_active), (gst_pad_peer_set_active), (gst_pad_check_pull_range), (gst_pad_pull_range): * gst/gstpad.h: Simplify pad activation. Added function to check if pull_range can be performed. Error out when pulling inactive or flushing pads. Removed const from refcounted types as it does not make sense. Simplify pad templates in basesink Added base class for simple 1-to-1 transforms. Make identity subclass the base transform.
2005-03-29 16:18:12 +00:00
GST_OBJECT_LOCK (pad);
gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. Original commit message from CVS: 2005-06-27 Andy Wingo <wingo@pobox.com> * gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. * gst/gsttrace.c (gst_alloc_trace_list_sorted): New helper, returns a sorted copy of the trace list. (gst_alloc_trace_print_live): New API, only prints traces with live objects. Sort the list. (gst_alloc_trace_print_all): Sort the list. (gst_alloc_trace_print): Align columns. * gst/elements/gstttypefindelement.c: * gst/elements/gsttee.c: * gst/base/gstbasesrc.c: * gst/base/gstbasesink.c: * gst/base/gstbasetransform.c: * gst/gstqueue.c: Adapt for pad activation changes. * gst/gstpipeline.c (gst_pipeline_init): Unref after parenting sched. (gst_pipeline_dispose): Drop ref on sched. * gst/gstpad.c (gst_pad_init): Set the default activate func. (gst_pad_activate_default): Push mode by default. (pre_activate_switch, post_activate_switch): New stubs, things to do before and after switching activation modes on pads. (gst_pad_set_active): Take a boolean and not a mode, dispatch to the pad's activate function to choose which mode to activate. Shortcut on deactivation and call the right function directly. (gst_pad_activate_pull): New API, (de)activates a pad in pull mode. (gst_pad_activate_push): New API, same for push mode. (gst_pad_set_activate_function) (gst_pad_set_activatepull_function) (gst_pad_set_activatepush_function): Setters for new API. * gst/gstminiobject.c (gst_mini_object_new, gst_mini_object_free): Trace all miniobjects. (gst_mini_object_make_writable): Unref the arg if we copy, like gst_caps_make_writable. * gst/gstmessage.c (_gst_message_initialize): No trace init. * gst/gstghostpad.c (gst_proxy_pad_do_activate) (gst_proxy_pad_do_activatepull, gst_proxy_pad_do_activatepush): Adapt for new pad API. * gst/gstevent.c (_gst_event_initialize): Don't initialize trace. * gst/gstelement.h: * gst/gstelement.c (gst_element_iterate_src_pads) (gst_element_iterate_sink_pads): New API functions. * gst/gstelement.c (iterator_fold_with_resync): New utility, should fold into gstiterator.c in some form. (gst_element_pads_activate): Simplified via use of fold and delegation of decisions to gstpad->activate. * gst/gstbus.c (gst_bus_source_finalize): Set the bus to NULL, help in debugging. * gst/gstbuffer.c (_gst_buffer_initialize): Ref the buffer type class once in init, like gstmessage. Didn't run into this issue but it seems correct. Don't initialize a trace, gstminiobject does that. * check/pipelines/simple_launch_lines.c (test_stop_from_app): New test, runs fakesrc ! fakesink, stopping on ::handoff via a message to the bus. (assert_live_count): New util function, uses alloc traces to check cleanup. * check/gst/gstghostpad.c (test_ghost_pads): More refcount checks. To be modified when unlink drops the internal pad.
2005-06-27 18:35:05 +00:00
old = GST_PAD_ACTIVATE_MODE (pad);
GST_OBJECT_UNLOCK (pad);
gst/: Simplify pad activation. Original commit message from CVS: * gst/base/Makefile.am: * gst/base/README: * gst/base/gstbasesink.c: (gst_basesink_get_type), (gst_basesink_base_init), (gst_basesink_class_init), (gst_basesink_pad_getcaps), (gst_basesink_init), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesink.h: * gst/base/gstbasetransform.c: (gst_base_transform_get_type), (gst_base_transform_base_init), (gst_base_transform_finalize), (gst_base_transform_class_init), (gst_base_transform_init), (gst_base_transform_proxy_getcaps), (gst_base_transform_setcaps), (gst_base_transform_event), (gst_base_transform_getrange), (gst_base_transform_chain), (gst_base_transform_handle_buffer), (gst_base_transform_set_property), (gst_base_transform_get_property), (gst_base_transform_sink_activate), (gst_base_transform_src_activate), (gst_base_transform_change_state): * gst/base/gstbasetransform.h: * gst/elements/gstidentity.c: (gst_identity_finalize), (gst_identity_class_init), (gst_identity_init), (gst_identity_event), (gst_identity_check_perfect), (gst_identity_transform), (gst_identity_set_property), (gst_identity_get_property), (gst_identity_change_state): * gst/elements/gstidentity.h: * gst/gstelement.c: (gst_element_get_state_func), (gst_element_lost_state), (gst_element_pads_activate): * gst/gstpad.c: (gst_pad_set_active), (gst_pad_peer_set_active), (gst_pad_check_pull_range), (gst_pad_pull_range): * gst/gstpad.h: Simplify pad activation. Added function to check if pull_range can be performed. Error out when pulling inactive or flushing pads. Removed const from refcounted types as it does not make sense. Simplify pad templates in basesink Added base class for simple 1-to-1 transforms. Make identity subclass the base transform.
2005-03-29 16:18:12 +00:00
gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. Original commit message from CVS: 2005-06-27 Andy Wingo <wingo@pobox.com> * gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. * gst/gsttrace.c (gst_alloc_trace_list_sorted): New helper, returns a sorted copy of the trace list. (gst_alloc_trace_print_live): New API, only prints traces with live objects. Sort the list. (gst_alloc_trace_print_all): Sort the list. (gst_alloc_trace_print): Align columns. * gst/elements/gstttypefindelement.c: * gst/elements/gsttee.c: * gst/base/gstbasesrc.c: * gst/base/gstbasesink.c: * gst/base/gstbasetransform.c: * gst/gstqueue.c: Adapt for pad activation changes. * gst/gstpipeline.c (gst_pipeline_init): Unref after parenting sched. (gst_pipeline_dispose): Drop ref on sched. * gst/gstpad.c (gst_pad_init): Set the default activate func. (gst_pad_activate_default): Push mode by default. (pre_activate_switch, post_activate_switch): New stubs, things to do before and after switching activation modes on pads. (gst_pad_set_active): Take a boolean and not a mode, dispatch to the pad's activate function to choose which mode to activate. Shortcut on deactivation and call the right function directly. (gst_pad_activate_pull): New API, (de)activates a pad in pull mode. (gst_pad_activate_push): New API, same for push mode. (gst_pad_set_activate_function) (gst_pad_set_activatepull_function) (gst_pad_set_activatepush_function): Setters for new API. * gst/gstminiobject.c (gst_mini_object_new, gst_mini_object_free): Trace all miniobjects. (gst_mini_object_make_writable): Unref the arg if we copy, like gst_caps_make_writable. * gst/gstmessage.c (_gst_message_initialize): No trace init. * gst/gstghostpad.c (gst_proxy_pad_do_activate) (gst_proxy_pad_do_activatepull, gst_proxy_pad_do_activatepush): Adapt for new pad API. * gst/gstevent.c (_gst_event_initialize): Don't initialize trace. * gst/gstelement.h: * gst/gstelement.c (gst_element_iterate_src_pads) (gst_element_iterate_sink_pads): New API functions. * gst/gstelement.c (iterator_fold_with_resync): New utility, should fold into gstiterator.c in some form. (gst_element_pads_activate): Simplified via use of fold and delegation of decisions to gstpad->activate. * gst/gstbus.c (gst_bus_source_finalize): Set the bus to NULL, help in debugging. * gst/gstbuffer.c (_gst_buffer_initialize): Ref the buffer type class once in init, like gstmessage. Didn't run into this issue but it seems correct. Don't initialize a trace, gstminiobject does that. * check/pipelines/simple_launch_lines.c (test_stop_from_app): New test, runs fakesrc ! fakesink, stopping on ::handoff via a message to the bus. (assert_live_count): New util function, uses alloc traces to check cleanup. * check/gst/gstghostpad.c (test_ghost_pads): More refcount checks. To be modified when unlink drops the internal pad.
2005-06-27 18:35:05 +00:00
if (active) {
switch (old) {
case GST_ACTIVATE_PUSH:
goto was_ok;
case GST_ACTIVATE_PULL:
/* pad was activate in the wrong direction, deactivate it
* an reactivate it in push mode */
if (G_UNLIKELY (!gst_pad_activate_pull (pad, FALSE)))
goto deactivate_failed;
/* fallthrough, pad is deactivated now. */
case GST_ACTIVATE_NONE:
break;
}
gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. Original commit message from CVS: 2005-06-27 Andy Wingo <wingo@pobox.com> * gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. * gst/gsttrace.c (gst_alloc_trace_list_sorted): New helper, returns a sorted copy of the trace list. (gst_alloc_trace_print_live): New API, only prints traces with live objects. Sort the list. (gst_alloc_trace_print_all): Sort the list. (gst_alloc_trace_print): Align columns. * gst/elements/gstttypefindelement.c: * gst/elements/gsttee.c: * gst/base/gstbasesrc.c: * gst/base/gstbasesink.c: * gst/base/gstbasetransform.c: * gst/gstqueue.c: Adapt for pad activation changes. * gst/gstpipeline.c (gst_pipeline_init): Unref after parenting sched. (gst_pipeline_dispose): Drop ref on sched. * gst/gstpad.c (gst_pad_init): Set the default activate func. (gst_pad_activate_default): Push mode by default. (pre_activate_switch, post_activate_switch): New stubs, things to do before and after switching activation modes on pads. (gst_pad_set_active): Take a boolean and not a mode, dispatch to the pad's activate function to choose which mode to activate. Shortcut on deactivation and call the right function directly. (gst_pad_activate_pull): New API, (de)activates a pad in pull mode. (gst_pad_activate_push): New API, same for push mode. (gst_pad_set_activate_function) (gst_pad_set_activatepull_function) (gst_pad_set_activatepush_function): Setters for new API. * gst/gstminiobject.c (gst_mini_object_new, gst_mini_object_free): Trace all miniobjects. (gst_mini_object_make_writable): Unref the arg if we copy, like gst_caps_make_writable. * gst/gstmessage.c (_gst_message_initialize): No trace init. * gst/gstghostpad.c (gst_proxy_pad_do_activate) (gst_proxy_pad_do_activatepull, gst_proxy_pad_do_activatepush): Adapt for new pad API. * gst/gstevent.c (_gst_event_initialize): Don't initialize trace. * gst/gstelement.h: * gst/gstelement.c (gst_element_iterate_src_pads) (gst_element_iterate_sink_pads): New API functions. * gst/gstelement.c (iterator_fold_with_resync): New utility, should fold into gstiterator.c in some form. (gst_element_pads_activate): Simplified via use of fold and delegation of decisions to gstpad->activate. * gst/gstbus.c (gst_bus_source_finalize): Set the bus to NULL, help in debugging. * gst/gstbuffer.c (_gst_buffer_initialize): Ref the buffer type class once in init, like gstmessage. Didn't run into this issue but it seems correct. Don't initialize a trace, gstminiobject does that. * check/pipelines/simple_launch_lines.c (test_stop_from_app): New test, runs fakesrc ! fakesink, stopping on ::handoff via a message to the bus. (assert_live_count): New util function, uses alloc traces to check cleanup. * check/gst/gstghostpad.c (test_ghost_pads): More refcount checks. To be modified when unlink drops the internal pad.
2005-06-27 18:35:05 +00:00
} else {
switch (old) {
case GST_ACTIVATE_NONE:
goto was_ok;
case GST_ACTIVATE_PULL:
/* pad was activated in the other direction, deactivate it
* in pull mode, this should not happen... */
if (G_UNLIKELY (!gst_pad_activate_pull (pad, FALSE)))
goto deactivate_failed;
/* everything is fine now */
goto was_ok;
case GST_ACTIVATE_PUSH:
break;
}
gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. Original commit message from CVS: 2005-06-27 Andy Wingo <wingo@pobox.com> * gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. * gst/gsttrace.c (gst_alloc_trace_list_sorted): New helper, returns a sorted copy of the trace list. (gst_alloc_trace_print_live): New API, only prints traces with live objects. Sort the list. (gst_alloc_trace_print_all): Sort the list. (gst_alloc_trace_print): Align columns. * gst/elements/gstttypefindelement.c: * gst/elements/gsttee.c: * gst/base/gstbasesrc.c: * gst/base/gstbasesink.c: * gst/base/gstbasetransform.c: * gst/gstqueue.c: Adapt for pad activation changes. * gst/gstpipeline.c (gst_pipeline_init): Unref after parenting sched. (gst_pipeline_dispose): Drop ref on sched. * gst/gstpad.c (gst_pad_init): Set the default activate func. (gst_pad_activate_default): Push mode by default. (pre_activate_switch, post_activate_switch): New stubs, things to do before and after switching activation modes on pads. (gst_pad_set_active): Take a boolean and not a mode, dispatch to the pad's activate function to choose which mode to activate. Shortcut on deactivation and call the right function directly. (gst_pad_activate_pull): New API, (de)activates a pad in pull mode. (gst_pad_activate_push): New API, same for push mode. (gst_pad_set_activate_function) (gst_pad_set_activatepull_function) (gst_pad_set_activatepush_function): Setters for new API. * gst/gstminiobject.c (gst_mini_object_new, gst_mini_object_free): Trace all miniobjects. (gst_mini_object_make_writable): Unref the arg if we copy, like gst_caps_make_writable. * gst/gstmessage.c (_gst_message_initialize): No trace init. * gst/gstghostpad.c (gst_proxy_pad_do_activate) (gst_proxy_pad_do_activatepull, gst_proxy_pad_do_activatepush): Adapt for new pad API. * gst/gstevent.c (_gst_event_initialize): Don't initialize trace. * gst/gstelement.h: * gst/gstelement.c (gst_element_iterate_src_pads) (gst_element_iterate_sink_pads): New API functions. * gst/gstelement.c (iterator_fold_with_resync): New utility, should fold into gstiterator.c in some form. (gst_element_pads_activate): Simplified via use of fold and delegation of decisions to gstpad->activate. * gst/gstbus.c (gst_bus_source_finalize): Set the bus to NULL, help in debugging. * gst/gstbuffer.c (_gst_buffer_initialize): Ref the buffer type class once in init, like gstmessage. Didn't run into this issue but it seems correct. Don't initialize a trace, gstminiobject does that. * check/pipelines/simple_launch_lines.c (test_stop_from_app): New test, runs fakesrc ! fakesink, stopping on ::handoff via a message to the bus. (assert_live_count): New util function, uses alloc traces to check cleanup. * check/gst/gstghostpad.c (test_ghost_pads): More refcount checks. To be modified when unlink drops the internal pad.
2005-06-27 18:35:05 +00:00
}
gst/: Simplify pad activation. Original commit message from CVS: * gst/base/Makefile.am: * gst/base/README: * gst/base/gstbasesink.c: (gst_basesink_get_type), (gst_basesink_base_init), (gst_basesink_class_init), (gst_basesink_pad_getcaps), (gst_basesink_init), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesink.h: * gst/base/gstbasetransform.c: (gst_base_transform_get_type), (gst_base_transform_base_init), (gst_base_transform_finalize), (gst_base_transform_class_init), (gst_base_transform_init), (gst_base_transform_proxy_getcaps), (gst_base_transform_setcaps), (gst_base_transform_event), (gst_base_transform_getrange), (gst_base_transform_chain), (gst_base_transform_handle_buffer), (gst_base_transform_set_property), (gst_base_transform_get_property), (gst_base_transform_sink_activate), (gst_base_transform_src_activate), (gst_base_transform_change_state): * gst/base/gstbasetransform.h: * gst/elements/gstidentity.c: (gst_identity_finalize), (gst_identity_class_init), (gst_identity_init), (gst_identity_event), (gst_identity_check_perfect), (gst_identity_transform), (gst_identity_set_property), (gst_identity_get_property), (gst_identity_change_state): * gst/elements/gstidentity.h: * gst/gstelement.c: (gst_element_get_state_func), (gst_element_lost_state), (gst_element_pads_activate): * gst/gstpad.c: (gst_pad_set_active), (gst_pad_peer_set_active), (gst_pad_check_pull_range), (gst_pad_pull_range): * gst/gstpad.h: Simplify pad activation. Added function to check if pull_range can be performed. Error out when pulling inactive or flushing pads. Removed const from refcounted types as it does not make sense. Simplify pad templates in basesink Added base class for simple 1-to-1 transforms. Make identity subclass the base transform.
2005-03-29 16:18:12 +00:00
new = active ? GST_ACTIVATE_PUSH : GST_ACTIVATE_NONE;
pre_activate (pad, new);
gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. Original commit message from CVS: 2005-06-27 Andy Wingo <wingo@pobox.com> * gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. * gst/gsttrace.c (gst_alloc_trace_list_sorted): New helper, returns a sorted copy of the trace list. (gst_alloc_trace_print_live): New API, only prints traces with live objects. Sort the list. (gst_alloc_trace_print_all): Sort the list. (gst_alloc_trace_print): Align columns. * gst/elements/gstttypefindelement.c: * gst/elements/gsttee.c: * gst/base/gstbasesrc.c: * gst/base/gstbasesink.c: * gst/base/gstbasetransform.c: * gst/gstqueue.c: Adapt for pad activation changes. * gst/gstpipeline.c (gst_pipeline_init): Unref after parenting sched. (gst_pipeline_dispose): Drop ref on sched. * gst/gstpad.c (gst_pad_init): Set the default activate func. (gst_pad_activate_default): Push mode by default. (pre_activate_switch, post_activate_switch): New stubs, things to do before and after switching activation modes on pads. (gst_pad_set_active): Take a boolean and not a mode, dispatch to the pad's activate function to choose which mode to activate. Shortcut on deactivation and call the right function directly. (gst_pad_activate_pull): New API, (de)activates a pad in pull mode. (gst_pad_activate_push): New API, same for push mode. (gst_pad_set_activate_function) (gst_pad_set_activatepull_function) (gst_pad_set_activatepush_function): Setters for new API. * gst/gstminiobject.c (gst_mini_object_new, gst_mini_object_free): Trace all miniobjects. (gst_mini_object_make_writable): Unref the arg if we copy, like gst_caps_make_writable. * gst/gstmessage.c (_gst_message_initialize): No trace init. * gst/gstghostpad.c (gst_proxy_pad_do_activate) (gst_proxy_pad_do_activatepull, gst_proxy_pad_do_activatepush): Adapt for new pad API. * gst/gstevent.c (_gst_event_initialize): Don't initialize trace. * gst/gstelement.h: * gst/gstelement.c (gst_element_iterate_src_pads) (gst_element_iterate_sink_pads): New API functions. * gst/gstelement.c (iterator_fold_with_resync): New utility, should fold into gstiterator.c in some form. (gst_element_pads_activate): Simplified via use of fold and delegation of decisions to gstpad->activate. * gst/gstbus.c (gst_bus_source_finalize): Set the bus to NULL, help in debugging. * gst/gstbuffer.c (_gst_buffer_initialize): Ref the buffer type class once in init, like gstmessage. Didn't run into this issue but it seems correct. Don't initialize a trace, gstminiobject does that. * check/pipelines/simple_launch_lines.c (test_stop_from_app): New test, runs fakesrc ! fakesink, stopping on ::handoff via a message to the bus. (assert_live_count): New util function, uses alloc traces to check cleanup. * check/gst/gstghostpad.c (test_ghost_pads): More refcount checks. To be modified when unlink drops the internal pad.
2005-06-27 18:35:05 +00:00
if (GST_PAD_ACTIVATEPUSHFUNC (pad)) {
if (G_UNLIKELY (!GST_PAD_ACTIVATEPUSHFUNC (pad) (pad, active))) {
gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. Original commit message from CVS: 2005-06-27 Andy Wingo <wingo@pobox.com> * gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. * gst/gsttrace.c (gst_alloc_trace_list_sorted): New helper, returns a sorted copy of the trace list. (gst_alloc_trace_print_live): New API, only prints traces with live objects. Sort the list. (gst_alloc_trace_print_all): Sort the list. (gst_alloc_trace_print): Align columns. * gst/elements/gstttypefindelement.c: * gst/elements/gsttee.c: * gst/base/gstbasesrc.c: * gst/base/gstbasesink.c: * gst/base/gstbasetransform.c: * gst/gstqueue.c: Adapt for pad activation changes. * gst/gstpipeline.c (gst_pipeline_init): Unref after parenting sched. (gst_pipeline_dispose): Drop ref on sched. * gst/gstpad.c (gst_pad_init): Set the default activate func. (gst_pad_activate_default): Push mode by default. (pre_activate_switch, post_activate_switch): New stubs, things to do before and after switching activation modes on pads. (gst_pad_set_active): Take a boolean and not a mode, dispatch to the pad's activate function to choose which mode to activate. Shortcut on deactivation and call the right function directly. (gst_pad_activate_pull): New API, (de)activates a pad in pull mode. (gst_pad_activate_push): New API, same for push mode. (gst_pad_set_activate_function) (gst_pad_set_activatepull_function) (gst_pad_set_activatepush_function): Setters for new API. * gst/gstminiobject.c (gst_mini_object_new, gst_mini_object_free): Trace all miniobjects. (gst_mini_object_make_writable): Unref the arg if we copy, like gst_caps_make_writable. * gst/gstmessage.c (_gst_message_initialize): No trace init. * gst/gstghostpad.c (gst_proxy_pad_do_activate) (gst_proxy_pad_do_activatepull, gst_proxy_pad_do_activatepush): Adapt for new pad API. * gst/gstevent.c (_gst_event_initialize): Don't initialize trace. * gst/gstelement.h: * gst/gstelement.c (gst_element_iterate_src_pads) (gst_element_iterate_sink_pads): New API functions. * gst/gstelement.c (iterator_fold_with_resync): New utility, should fold into gstiterator.c in some form. (gst_element_pads_activate): Simplified via use of fold and delegation of decisions to gstpad->activate. * gst/gstbus.c (gst_bus_source_finalize): Set the bus to NULL, help in debugging. * gst/gstbuffer.c (_gst_buffer_initialize): Ref the buffer type class once in init, like gstmessage. Didn't run into this issue but it seems correct. Don't initialize a trace, gstminiobject does that. * check/pipelines/simple_launch_lines.c (test_stop_from_app): New test, runs fakesrc ! fakesink, stopping on ::handoff via a message to the bus. (assert_live_count): New util function, uses alloc traces to check cleanup. * check/gst/gstghostpad.c (test_ghost_pads): More refcount checks. To be modified when unlink drops the internal pad.
2005-06-27 18:35:05 +00:00
goto failure;
}
} else {
/* quite ok, element relies on state change func to prepare itself */
}
post_activate (pad, new);
GST_CAT_DEBUG_OBJECT (GST_CAT_PADS, pad, "%s in push mode",
active ? "activated" : "deactivated");
return TRUE;
gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. Original commit message from CVS: 2005-06-27 Andy Wingo <wingo@pobox.com> * gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. * gst/gsttrace.c (gst_alloc_trace_list_sorted): New helper, returns a sorted copy of the trace list. (gst_alloc_trace_print_live): New API, only prints traces with live objects. Sort the list. (gst_alloc_trace_print_all): Sort the list. (gst_alloc_trace_print): Align columns. * gst/elements/gstttypefindelement.c: * gst/elements/gsttee.c: * gst/base/gstbasesrc.c: * gst/base/gstbasesink.c: * gst/base/gstbasetransform.c: * gst/gstqueue.c: Adapt for pad activation changes. * gst/gstpipeline.c (gst_pipeline_init): Unref after parenting sched. (gst_pipeline_dispose): Drop ref on sched. * gst/gstpad.c (gst_pad_init): Set the default activate func. (gst_pad_activate_default): Push mode by default. (pre_activate_switch, post_activate_switch): New stubs, things to do before and after switching activation modes on pads. (gst_pad_set_active): Take a boolean and not a mode, dispatch to the pad's activate function to choose which mode to activate. Shortcut on deactivation and call the right function directly. (gst_pad_activate_pull): New API, (de)activates a pad in pull mode. (gst_pad_activate_push): New API, same for push mode. (gst_pad_set_activate_function) (gst_pad_set_activatepull_function) (gst_pad_set_activatepush_function): Setters for new API. * gst/gstminiobject.c (gst_mini_object_new, gst_mini_object_free): Trace all miniobjects. (gst_mini_object_make_writable): Unref the arg if we copy, like gst_caps_make_writable. * gst/gstmessage.c (_gst_message_initialize): No trace init. * gst/gstghostpad.c (gst_proxy_pad_do_activate) (gst_proxy_pad_do_activatepull, gst_proxy_pad_do_activatepush): Adapt for new pad API. * gst/gstevent.c (_gst_event_initialize): Don't initialize trace. * gst/gstelement.h: * gst/gstelement.c (gst_element_iterate_src_pads) (gst_element_iterate_sink_pads): New API functions. * gst/gstelement.c (iterator_fold_with_resync): New utility, should fold into gstiterator.c in some form. (gst_element_pads_activate): Simplified via use of fold and delegation of decisions to gstpad->activate. * gst/gstbus.c (gst_bus_source_finalize): Set the bus to NULL, help in debugging. * gst/gstbuffer.c (_gst_buffer_initialize): Ref the buffer type class once in init, like gstmessage. Didn't run into this issue but it seems correct. Don't initialize a trace, gstminiobject does that. * check/pipelines/simple_launch_lines.c (test_stop_from_app): New test, runs fakesrc ! fakesink, stopping on ::handoff via a message to the bus. (assert_live_count): New util function, uses alloc traces to check cleanup. * check/gst/gstghostpad.c (test_ghost_pads): More refcount checks. To be modified when unlink drops the internal pad.
2005-06-27 18:35:05 +00:00
was_ok:
{
GST_CAT_DEBUG_OBJECT (GST_CAT_PADS, pad, "already %s in push mode",
active ? "activated" : "deactivated");
return TRUE;
}
deactivate_failed:
{
GST_CAT_DEBUG_OBJECT (GST_CAT_PADS, pad,
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
"failed to %s in switch to push from mode %d",
(active ? "activate" : "deactivate"), old);
return FALSE;
}
gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. Original commit message from CVS: 2005-06-27 Andy Wingo <wingo@pobox.com> * gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. * gst/gsttrace.c (gst_alloc_trace_list_sorted): New helper, returns a sorted copy of the trace list. (gst_alloc_trace_print_live): New API, only prints traces with live objects. Sort the list. (gst_alloc_trace_print_all): Sort the list. (gst_alloc_trace_print): Align columns. * gst/elements/gstttypefindelement.c: * gst/elements/gsttee.c: * gst/base/gstbasesrc.c: * gst/base/gstbasesink.c: * gst/base/gstbasetransform.c: * gst/gstqueue.c: Adapt for pad activation changes. * gst/gstpipeline.c (gst_pipeline_init): Unref after parenting sched. (gst_pipeline_dispose): Drop ref on sched. * gst/gstpad.c (gst_pad_init): Set the default activate func. (gst_pad_activate_default): Push mode by default. (pre_activate_switch, post_activate_switch): New stubs, things to do before and after switching activation modes on pads. (gst_pad_set_active): Take a boolean and not a mode, dispatch to the pad's activate function to choose which mode to activate. Shortcut on deactivation and call the right function directly. (gst_pad_activate_pull): New API, (de)activates a pad in pull mode. (gst_pad_activate_push): New API, same for push mode. (gst_pad_set_activate_function) (gst_pad_set_activatepull_function) (gst_pad_set_activatepush_function): Setters for new API. * gst/gstminiobject.c (gst_mini_object_new, gst_mini_object_free): Trace all miniobjects. (gst_mini_object_make_writable): Unref the arg if we copy, like gst_caps_make_writable. * gst/gstmessage.c (_gst_message_initialize): No trace init. * gst/gstghostpad.c (gst_proxy_pad_do_activate) (gst_proxy_pad_do_activatepull, gst_proxy_pad_do_activatepush): Adapt for new pad API. * gst/gstevent.c (_gst_event_initialize): Don't initialize trace. * gst/gstelement.h: * gst/gstelement.c (gst_element_iterate_src_pads) (gst_element_iterate_sink_pads): New API functions. * gst/gstelement.c (iterator_fold_with_resync): New utility, should fold into gstiterator.c in some form. (gst_element_pads_activate): Simplified via use of fold and delegation of decisions to gstpad->activate. * gst/gstbus.c (gst_bus_source_finalize): Set the bus to NULL, help in debugging. * gst/gstbuffer.c (_gst_buffer_initialize): Ref the buffer type class once in init, like gstmessage. Didn't run into this issue but it seems correct. Don't initialize a trace, gstminiobject does that. * check/pipelines/simple_launch_lines.c (test_stop_from_app): New test, runs fakesrc ! fakesink, stopping on ::handoff via a message to the bus. (assert_live_count): New util function, uses alloc traces to check cleanup. * check/gst/gstghostpad.c (test_ghost_pads): More refcount checks. To be modified when unlink drops the internal pad.
2005-06-27 18:35:05 +00:00
failure:
gst/: Simplify pad activation. Original commit message from CVS: * gst/base/Makefile.am: * gst/base/README: * gst/base/gstbasesink.c: (gst_basesink_get_type), (gst_basesink_base_init), (gst_basesink_class_init), (gst_basesink_pad_getcaps), (gst_basesink_init), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesink.h: * gst/base/gstbasetransform.c: (gst_base_transform_get_type), (gst_base_transform_base_init), (gst_base_transform_finalize), (gst_base_transform_class_init), (gst_base_transform_init), (gst_base_transform_proxy_getcaps), (gst_base_transform_setcaps), (gst_base_transform_event), (gst_base_transform_getrange), (gst_base_transform_chain), (gst_base_transform_handle_buffer), (gst_base_transform_set_property), (gst_base_transform_get_property), (gst_base_transform_sink_activate), (gst_base_transform_src_activate), (gst_base_transform_change_state): * gst/base/gstbasetransform.h: * gst/elements/gstidentity.c: (gst_identity_finalize), (gst_identity_class_init), (gst_identity_init), (gst_identity_event), (gst_identity_check_perfect), (gst_identity_transform), (gst_identity_set_property), (gst_identity_get_property), (gst_identity_change_state): * gst/elements/gstidentity.h: * gst/gstelement.c: (gst_element_get_state_func), (gst_element_lost_state), (gst_element_pads_activate): * gst/gstpad.c: (gst_pad_set_active), (gst_pad_peer_set_active), (gst_pad_check_pull_range), (gst_pad_pull_range): * gst/gstpad.h: Simplify pad activation. Added function to check if pull_range can be performed. Error out when pulling inactive or flushing pads. Removed const from refcounted types as it does not make sense. Simplify pad templates in basesink Added base class for simple 1-to-1 transforms. Make identity subclass the base transform.
2005-03-29 16:18:12 +00:00
{
GST_OBJECT_LOCK (pad);
gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. Original commit message from CVS: 2005-06-27 Andy Wingo <wingo@pobox.com> * gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. * gst/gsttrace.c (gst_alloc_trace_list_sorted): New helper, returns a sorted copy of the trace list. (gst_alloc_trace_print_live): New API, only prints traces with live objects. Sort the list. (gst_alloc_trace_print_all): Sort the list. (gst_alloc_trace_print): Align columns. * gst/elements/gstttypefindelement.c: * gst/elements/gsttee.c: * gst/base/gstbasesrc.c: * gst/base/gstbasesink.c: * gst/base/gstbasetransform.c: * gst/gstqueue.c: Adapt for pad activation changes. * gst/gstpipeline.c (gst_pipeline_init): Unref after parenting sched. (gst_pipeline_dispose): Drop ref on sched. * gst/gstpad.c (gst_pad_init): Set the default activate func. (gst_pad_activate_default): Push mode by default. (pre_activate_switch, post_activate_switch): New stubs, things to do before and after switching activation modes on pads. (gst_pad_set_active): Take a boolean and not a mode, dispatch to the pad's activate function to choose which mode to activate. Shortcut on deactivation and call the right function directly. (gst_pad_activate_pull): New API, (de)activates a pad in pull mode. (gst_pad_activate_push): New API, same for push mode. (gst_pad_set_activate_function) (gst_pad_set_activatepull_function) (gst_pad_set_activatepush_function): Setters for new API. * gst/gstminiobject.c (gst_mini_object_new, gst_mini_object_free): Trace all miniobjects. (gst_mini_object_make_writable): Unref the arg if we copy, like gst_caps_make_writable. * gst/gstmessage.c (_gst_message_initialize): No trace init. * gst/gstghostpad.c (gst_proxy_pad_do_activate) (gst_proxy_pad_do_activatepull, gst_proxy_pad_do_activatepush): Adapt for new pad API. * gst/gstevent.c (_gst_event_initialize): Don't initialize trace. * gst/gstelement.h: * gst/gstelement.c (gst_element_iterate_src_pads) (gst_element_iterate_sink_pads): New API functions. * gst/gstelement.c (iterator_fold_with_resync): New utility, should fold into gstiterator.c in some form. (gst_element_pads_activate): Simplified via use of fold and delegation of decisions to gstpad->activate. * gst/gstbus.c (gst_bus_source_finalize): Set the bus to NULL, help in debugging. * gst/gstbuffer.c (_gst_buffer_initialize): Ref the buffer type class once in init, like gstmessage. Didn't run into this issue but it seems correct. Don't initialize a trace, gstminiobject does that. * check/pipelines/simple_launch_lines.c (test_stop_from_app): New test, runs fakesrc ! fakesink, stopping on ::handoff via a message to the bus. (assert_live_count): New util function, uses alloc traces to check cleanup. * check/gst/gstghostpad.c (test_ghost_pads): More refcount checks. To be modified when unlink drops the internal pad.
2005-06-27 18:35:05 +00:00
GST_CAT_INFO_OBJECT (GST_CAT_PADS, pad, "failed to %s in push mode",
active ? "activate" : "deactivate");
GST_PAD_SET_FLUSHING (pad);
GST_PAD_ACTIVATE_MODE (pad) = old;
GST_OBJECT_UNLOCK (pad);
gst/: Simplify pad activation. Original commit message from CVS: * gst/base/Makefile.am: * gst/base/README: * gst/base/gstbasesink.c: (gst_basesink_get_type), (gst_basesink_base_init), (gst_basesink_class_init), (gst_basesink_pad_getcaps), (gst_basesink_init), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesink.h: * gst/base/gstbasetransform.c: (gst_base_transform_get_type), (gst_base_transform_base_init), (gst_base_transform_finalize), (gst_base_transform_class_init), (gst_base_transform_init), (gst_base_transform_proxy_getcaps), (gst_base_transform_setcaps), (gst_base_transform_event), (gst_base_transform_getrange), (gst_base_transform_chain), (gst_base_transform_handle_buffer), (gst_base_transform_set_property), (gst_base_transform_get_property), (gst_base_transform_sink_activate), (gst_base_transform_src_activate), (gst_base_transform_change_state): * gst/base/gstbasetransform.h: * gst/elements/gstidentity.c: (gst_identity_finalize), (gst_identity_class_init), (gst_identity_init), (gst_identity_event), (gst_identity_check_perfect), (gst_identity_transform), (gst_identity_set_property), (gst_identity_get_property), (gst_identity_change_state): * gst/elements/gstidentity.h: * gst/gstelement.c: (gst_element_get_state_func), (gst_element_lost_state), (gst_element_pads_activate): * gst/gstpad.c: (gst_pad_set_active), (gst_pad_peer_set_active), (gst_pad_check_pull_range), (gst_pad_pull_range): * gst/gstpad.h: Simplify pad activation. Added function to check if pull_range can be performed. Error out when pulling inactive or flushing pads. Removed const from refcounted types as it does not make sense. Simplify pad templates in basesink Added base class for simple 1-to-1 transforms. Make identity subclass the base transform.
2005-03-29 16:18:12 +00:00
return FALSE;
}
}
/**
* gst_pad_is_active:
* @pad: the #GstPad to query
*
* Query if a pad is active
*
* Returns: TRUE if the pad is active.
*
* MT safe.
*/
gboolean
gst_pad_is_active (GstPad * pad)
{
gboolean result = FALSE;
g_return_val_if_fail (GST_IS_PAD (pad), FALSE);
GST_OBJECT_LOCK (pad);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
result = GST_PAD_MODE_ACTIVATE (GST_PAD_ACTIVATE_MODE (pad));
GST_OBJECT_UNLOCK (pad);
return result;
}
/**
* gst_pad_set_blocked_async:
* @pad: the #GstPad to block or unblock
* @blocked: boolean indicating whether the pad should be blocked or unblocked
* @callback: #GstPadBlockCallback that will be called when the
* operation succeeds
* @user_data: user data passed to the callback
*
* Blocks or unblocks the dataflow on a pad. The provided callback
* is called when the operation succeeds; this happens right before the next
* attempt at pushing a buffer on the pad.
*
* This can take a while as the pad can only become blocked when real dataflow
* is happening.
* When the pipeline is stalled, for example in PAUSED, this can
* take an indeterminate amount of time.
* You can pass NULL as the callback to make this call block. Be careful with
* this blocking call as it might not return for reasons stated above.
*
* Returns: TRUE if the pad could be blocked. This function can fail
* if wrong parameters were passed or the pad was already in the
* requested state.
*
* MT safe.
*/
gboolean
gst_pad_set_blocked_async (GstPad * pad, gboolean blocked,
GstPadBlockCallback callback, gpointer user_data)
{
gst/gstbin.c: (de)activate src pads before calling state_change on the childs. Original commit message from CVS: * gst/gstbin.c: (activate_pads), (iterator_activate_fold_with_resync), (gst_bin_src_pads_activate), (gst_bin_change_state_func): (de)activate src pads before calling state_change on the childs. This is to avoid the case where a src ghostpad is blocked (holding the stream lock), which would block the deactivation of the ghostpad's target pad. * gst/gstghostpad.c: (gst_proxy_pad_do_query_type), (gst_proxy_pad_do_event), (gst_proxy_pad_do_query), (gst_proxy_pad_do_internal_link), (gst_proxy_pad_do_bufferalloc), (gst_proxy_pad_do_chain), (gst_proxy_pad_do_getrange), (gst_proxy_pad_do_checkgetrange), (gst_proxy_pad_do_getcaps), (gst_proxy_pad_do_acceptcaps), (gst_proxy_pad_do_fixatecaps), (gst_proxy_pad_do_setcaps), (gst_proxy_pad_set_target_unlocked), (gst_proxy_pad_set_target), (gst_proxy_pad_get_internal), (gst_proxy_pad_dispose), (gst_proxy_pad_init), (gst_ghost_pad_parent_set), (gst_ghost_pad_parent_unset), (gst_ghost_pad_class_init), (gst_ghost_pad_internal_do_activate_push), (gst_ghost_pad_internal_do_activate_pull), (gst_ghost_pad_do_activate_push), (gst_ghost_pad_do_activate_pull), (gst_ghost_pad_do_link), (gst_ghost_pad_do_unlink), (gst_ghost_pad_dispose), (gst_ghost_pad_new_no_target), (gst_ghost_pad_new), (gst_ghost_pad_set_target): GhostPads now create their internal GstProxyPad at creation (and not when they're linked, as it was being done previously). The internal and target pads are linked straight away. The data will also travel through the other pad in order to make pad blocking and probes non-hackish (the probe/block now really happens on the GhostPad and not on the target). * gst/gstpad.c: (gst_pad_set_blocked_async), (gst_pad_link_prepare), (gst_pad_push_event): Remove previous ghostpad cruft. * gst/gstutils.c: (gst_pad_add_data_probe), (gst_pad_add_event_probe), (gst_pad_add_buffer_probe), (gst_pad_remove_data_probe), (gst_pad_remove_event_probe), (gst_pad_remove_buffer_probe): Remove previous ghost pad cruft. Added more detailed debug statements. * tests/check/gst/gstghostpad.c: (GST_START_TEST): Fix the testsuite for refcounting changes. The comments about who has references were correct, but the refcount being checked wasn't the same (!?!).
2006-07-11 16:20:09 +00:00
gboolean was_blocked = FALSE;
g_return_val_if_fail (GST_IS_PAD (pad), FALSE);
GST_OBJECT_LOCK (pad);
was_blocked = GST_PAD_IS_BLOCKED (pad);
if (G_UNLIKELY (was_blocked == blocked))
goto had_right_state;
if (blocked) {
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_LOG_OBJECT (GST_CAT_SCHEDULING, pad, "blocking pad");
renamed GST_FLAGS macros to GST_OBJECT_FLAGS moved bitshift from macro to enum definition Original commit message from CVS: * check/gst/gstbin.c: (GST_START_TEST): * docs/gst/gstreamer-sections.txt: * gst/base/gstbasesink.c: (gst_base_sink_init): * gst/base/gstbasesrc.c: (gst_base_src_init), (gst_base_src_get_range), (gst_base_src_check_get_range), (gst_base_src_start), (gst_base_src_stop): * gst/base/gstbasesrc.h: * gst/elements/gstfakesrc.c: (gst_fake_src_set_property): * gst/gstbin.c: (gst_bin_add_func), (gst_bin_remove_func), (bin_element_is_sink), (reset_degree), (gst_bin_element_set_state), (bin_bus_handler): * gst/gstbin.h: * gst/gstbuffer.h: * gst/gstbus.c: (gst_bus_post), (gst_bus_set_flushing): * gst/gstbus.h: * gst/gstelement.c: (gst_element_is_locked_state), (gst_element_set_locked_state), (gst_element_commit_state), (gst_element_set_state): * gst/gstelement.h: * gst/gstindex.c: (gst_index_init): * gst/gstindex.h: * gst/gstminiobject.h: * gst/gstobject.c: (gst_object_init), (gst_object_sink), (gst_object_set_parent): * gst/gstobject.h: * gst/gstpad.c: (gst_pad_set_blocked_async), (gst_pad_is_blocked), (gst_pad_get_caps_unlocked), (gst_pad_set_caps): * gst/gstpad.h: * gst/gstpadtemplate.h: * gst/gstpipeline.c: (gst_pipeline_provide_clock_func), (gst_pipeline_use_clock), (gst_pipeline_auto_clock): * gst/gstpipeline.h: * gst/indexers/gstfileindex.c: (gst_file_index_load), (gst_file_index_commit): * testsuite/bytestream/filepadsink.c: (gst_fp_sink_init): * testsuite/pad/link.c: (gst_test_src_init), (gst_test_filter_init), (gst_test_sink_init): * testsuite/states/locked.c: (main): renamed GST_FLAGS macros to GST_OBJECT_FLAGS moved bitshift from macro to enum definition
2005-10-12 14:28:39 +00:00
GST_OBJECT_FLAG_SET (pad, GST_PAD_BLOCKED);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
pad->block_callback = callback;
pad->block_data = user_data;
if (!callback) {
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
GST_CAT_LOG_OBJECT (GST_CAT_SCHEDULING, pad, "waiting for block");
GST_PAD_BLOCK_WAIT (pad);
GST_CAT_LOG_OBJECT (GST_CAT_SCHEDULING, pad, "blocked");
}
} else {
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_LOG_OBJECT (GST_CAT_SCHEDULING, pad, "unblocking pad");
renamed GST_FLAGS macros to GST_OBJECT_FLAGS moved bitshift from macro to enum definition Original commit message from CVS: * check/gst/gstbin.c: (GST_START_TEST): * docs/gst/gstreamer-sections.txt: * gst/base/gstbasesink.c: (gst_base_sink_init): * gst/base/gstbasesrc.c: (gst_base_src_init), (gst_base_src_get_range), (gst_base_src_check_get_range), (gst_base_src_start), (gst_base_src_stop): * gst/base/gstbasesrc.h: * gst/elements/gstfakesrc.c: (gst_fake_src_set_property): * gst/gstbin.c: (gst_bin_add_func), (gst_bin_remove_func), (bin_element_is_sink), (reset_degree), (gst_bin_element_set_state), (bin_bus_handler): * gst/gstbin.h: * gst/gstbuffer.h: * gst/gstbus.c: (gst_bus_post), (gst_bus_set_flushing): * gst/gstbus.h: * gst/gstelement.c: (gst_element_is_locked_state), (gst_element_set_locked_state), (gst_element_commit_state), (gst_element_set_state): * gst/gstelement.h: * gst/gstindex.c: (gst_index_init): * gst/gstindex.h: * gst/gstminiobject.h: * gst/gstobject.c: (gst_object_init), (gst_object_sink), (gst_object_set_parent): * gst/gstobject.h: * gst/gstpad.c: (gst_pad_set_blocked_async), (gst_pad_is_blocked), (gst_pad_get_caps_unlocked), (gst_pad_set_caps): * gst/gstpad.h: * gst/gstpadtemplate.h: * gst/gstpipeline.c: (gst_pipeline_provide_clock_func), (gst_pipeline_use_clock), (gst_pipeline_auto_clock): * gst/gstpipeline.h: * gst/indexers/gstfileindex.c: (gst_file_index_load), (gst_file_index_commit): * testsuite/bytestream/filepadsink.c: (gst_fp_sink_init): * testsuite/pad/link.c: (gst_test_src_init), (gst_test_filter_init), (gst_test_sink_init): * testsuite/states/locked.c: (main): renamed GST_FLAGS macros to GST_OBJECT_FLAGS moved bitshift from macro to enum definition
2005-10-12 14:28:39 +00:00
GST_OBJECT_FLAG_UNSET (pad, GST_PAD_BLOCKED);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
pad->block_callback = callback;
pad->block_data = user_data;
GST_PAD_BLOCK_BROADCAST (pad);
if (!callback) {
/* no callback, wait for the unblock to happen */
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
GST_CAT_LOG_OBJECT (GST_CAT_SCHEDULING, pad, "waiting for unblock");
GST_PAD_BLOCK_WAIT (pad);
GST_CAT_LOG_OBJECT (GST_CAT_SCHEDULING, pad, "unblocked");
}
}
GST_OBJECT_UNLOCK (pad);
return TRUE;
had_right_state:
{
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
GST_CAT_LOG_OBJECT (GST_CAT_SCHEDULING, pad,
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
"pad was in right state (%d)", was_blocked);
GST_OBJECT_UNLOCK (pad);
return FALSE;
}
}
/**
* gst_pad_set_blocked:
* @pad: the #GstPad to block or unblock
* @blocked: boolean indicating we should block or unblock
*
* Blocks or unblocks the dataflow on a pad. This function is
* a shortcut for gst_pad_set_blocked_async() with a NULL
* callback.
*
* Returns: TRUE if the pad could be blocked. This function can fail
* wrong parameters were passed or the pad was already in the
* requested state.
*
* MT safe.
*/
gboolean
gst_pad_set_blocked (GstPad * pad, gboolean blocked)
{
return gst_pad_set_blocked_async (pad, blocked, NULL, NULL);
}
/**
* gst_pad_is_blocked:
* @pad: the #GstPad to query
*
* Checks if the pad is blocked or not. This function returns the
* last requested state of the pad. It is not certain that the pad
* is actually blocking at this point (see gst_pad_is_blocking()).
*
* Returns: TRUE if the pad is blocked.
*
* MT safe.
*/
gboolean
gst_pad_is_blocked (GstPad * pad)
{
gboolean result = FALSE;
g_return_val_if_fail (GST_IS_PAD (pad), result);
GST_OBJECT_LOCK (pad);
renamed GST_FLAGS macros to GST_OBJECT_FLAGS moved bitshift from macro to enum definition Original commit message from CVS: * check/gst/gstbin.c: (GST_START_TEST): * docs/gst/gstreamer-sections.txt: * gst/base/gstbasesink.c: (gst_base_sink_init): * gst/base/gstbasesrc.c: (gst_base_src_init), (gst_base_src_get_range), (gst_base_src_check_get_range), (gst_base_src_start), (gst_base_src_stop): * gst/base/gstbasesrc.h: * gst/elements/gstfakesrc.c: (gst_fake_src_set_property): * gst/gstbin.c: (gst_bin_add_func), (gst_bin_remove_func), (bin_element_is_sink), (reset_degree), (gst_bin_element_set_state), (bin_bus_handler): * gst/gstbin.h: * gst/gstbuffer.h: * gst/gstbus.c: (gst_bus_post), (gst_bus_set_flushing): * gst/gstbus.h: * gst/gstelement.c: (gst_element_is_locked_state), (gst_element_set_locked_state), (gst_element_commit_state), (gst_element_set_state): * gst/gstelement.h: * gst/gstindex.c: (gst_index_init): * gst/gstindex.h: * gst/gstminiobject.h: * gst/gstobject.c: (gst_object_init), (gst_object_sink), (gst_object_set_parent): * gst/gstobject.h: * gst/gstpad.c: (gst_pad_set_blocked_async), (gst_pad_is_blocked), (gst_pad_get_caps_unlocked), (gst_pad_set_caps): * gst/gstpad.h: * gst/gstpadtemplate.h: * gst/gstpipeline.c: (gst_pipeline_provide_clock_func), (gst_pipeline_use_clock), (gst_pipeline_auto_clock): * gst/gstpipeline.h: * gst/indexers/gstfileindex.c: (gst_file_index_load), (gst_file_index_commit): * testsuite/bytestream/filepadsink.c: (gst_fp_sink_init): * testsuite/pad/link.c: (gst_test_src_init), (gst_test_filter_init), (gst_test_sink_init): * testsuite/states/locked.c: (main): renamed GST_FLAGS macros to GST_OBJECT_FLAGS moved bitshift from macro to enum definition
2005-10-12 14:28:39 +00:00
result = GST_OBJECT_FLAG_IS_SET (pad, GST_PAD_BLOCKED);
GST_OBJECT_UNLOCK (pad);
return result;
}
/**
* gst_pad_is_blocking:
* @pad: the #GstPad to query
*
* Checks if the pad is blocking or not. This is a guaranteed state
* of whether the pad is actually blocking on a #GstBuffer or a #GstEvent.
*
* Returns: TRUE if the pad is blocking.
*
* MT safe.
*
* Since: 0.10.11
*/
gboolean
gst_pad_is_blocking (GstPad * pad)
{
gboolean result = FALSE;
g_return_val_if_fail (GST_IS_PAD (pad), result);
GST_OBJECT_LOCK (pad);
/* the blocking flag is only valid if the pad is not flushing */
result = GST_OBJECT_FLAG_IS_SET (pad, GST_PAD_BLOCKING) &&
!GST_OBJECT_FLAG_IS_SET (pad, GST_PAD_FLUSHING);
GST_OBJECT_UNLOCK (pad);
return result;
}
/**
* gst_pad_set_activate_function:
* @pad: a #GstPad.
* @activate: the #GstPadActivateFunction to set.
*
* Sets the given activate function for @pad. The activate function will
* dispatch to gst_pad_activate_push() or gst_pad_activate_pull() to perform
* the actual activation. Only makes sense to set on sink pads.
gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. Original commit message from CVS: 2005-06-27 Andy Wingo <wingo@pobox.com> * gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. * gst/gsttrace.c (gst_alloc_trace_list_sorted): New helper, returns a sorted copy of the trace list. (gst_alloc_trace_print_live): New API, only prints traces with live objects. Sort the list. (gst_alloc_trace_print_all): Sort the list. (gst_alloc_trace_print): Align columns. * gst/elements/gstttypefindelement.c: * gst/elements/gsttee.c: * gst/base/gstbasesrc.c: * gst/base/gstbasesink.c: * gst/base/gstbasetransform.c: * gst/gstqueue.c: Adapt for pad activation changes. * gst/gstpipeline.c (gst_pipeline_init): Unref after parenting sched. (gst_pipeline_dispose): Drop ref on sched. * gst/gstpad.c (gst_pad_init): Set the default activate func. (gst_pad_activate_default): Push mode by default. (pre_activate_switch, post_activate_switch): New stubs, things to do before and after switching activation modes on pads. (gst_pad_set_active): Take a boolean and not a mode, dispatch to the pad's activate function to choose which mode to activate. Shortcut on deactivation and call the right function directly. (gst_pad_activate_pull): New API, (de)activates a pad in pull mode. (gst_pad_activate_push): New API, same for push mode. (gst_pad_set_activate_function) (gst_pad_set_activatepull_function) (gst_pad_set_activatepush_function): Setters for new API. * gst/gstminiobject.c (gst_mini_object_new, gst_mini_object_free): Trace all miniobjects. (gst_mini_object_make_writable): Unref the arg if we copy, like gst_caps_make_writable. * gst/gstmessage.c (_gst_message_initialize): No trace init. * gst/gstghostpad.c (gst_proxy_pad_do_activate) (gst_proxy_pad_do_activatepull, gst_proxy_pad_do_activatepush): Adapt for new pad API. * gst/gstevent.c (_gst_event_initialize): Don't initialize trace. * gst/gstelement.h: * gst/gstelement.c (gst_element_iterate_src_pads) (gst_element_iterate_sink_pads): New API functions. * gst/gstelement.c (iterator_fold_with_resync): New utility, should fold into gstiterator.c in some form. (gst_element_pads_activate): Simplified via use of fold and delegation of decisions to gstpad->activate. * gst/gstbus.c (gst_bus_source_finalize): Set the bus to NULL, help in debugging. * gst/gstbuffer.c (_gst_buffer_initialize): Ref the buffer type class once in init, like gstmessage. Didn't run into this issue but it seems correct. Don't initialize a trace, gstminiobject does that. * check/pipelines/simple_launch_lines.c (test_stop_from_app): New test, runs fakesrc ! fakesink, stopping on ::handoff via a message to the bus. (assert_live_count): New util function, uses alloc traces to check cleanup. * check/gst/gstghostpad.c (test_ghost_pads): More refcount checks. To be modified when unlink drops the internal pad.
2005-06-27 18:35:05 +00:00
*
* Call this function if your sink pad can start a pull-based task.
*/
void
gst_pad_set_activate_function (GstPad * pad, GstPadActivateFunction activate)
{
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
g_return_if_fail (GST_IS_PAD (pad));
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
GST_PAD_ACTIVATEFUNC (pad) = activate;
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_DEBUG_OBJECT (GST_CAT_PADS, pad, "activatefunc set to %s",
GST_DEBUG_FUNCPTR_NAME (activate));
}
gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. Original commit message from CVS: 2005-06-27 Andy Wingo <wingo@pobox.com> * gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. * gst/gsttrace.c (gst_alloc_trace_list_sorted): New helper, returns a sorted copy of the trace list. (gst_alloc_trace_print_live): New API, only prints traces with live objects. Sort the list. (gst_alloc_trace_print_all): Sort the list. (gst_alloc_trace_print): Align columns. * gst/elements/gstttypefindelement.c: * gst/elements/gsttee.c: * gst/base/gstbasesrc.c: * gst/base/gstbasesink.c: * gst/base/gstbasetransform.c: * gst/gstqueue.c: Adapt for pad activation changes. * gst/gstpipeline.c (gst_pipeline_init): Unref after parenting sched. (gst_pipeline_dispose): Drop ref on sched. * gst/gstpad.c (gst_pad_init): Set the default activate func. (gst_pad_activate_default): Push mode by default. (pre_activate_switch, post_activate_switch): New stubs, things to do before and after switching activation modes on pads. (gst_pad_set_active): Take a boolean and not a mode, dispatch to the pad's activate function to choose which mode to activate. Shortcut on deactivation and call the right function directly. (gst_pad_activate_pull): New API, (de)activates a pad in pull mode. (gst_pad_activate_push): New API, same for push mode. (gst_pad_set_activate_function) (gst_pad_set_activatepull_function) (gst_pad_set_activatepush_function): Setters for new API. * gst/gstminiobject.c (gst_mini_object_new, gst_mini_object_free): Trace all miniobjects. (gst_mini_object_make_writable): Unref the arg if we copy, like gst_caps_make_writable. * gst/gstmessage.c (_gst_message_initialize): No trace init. * gst/gstghostpad.c (gst_proxy_pad_do_activate) (gst_proxy_pad_do_activatepull, gst_proxy_pad_do_activatepush): Adapt for new pad API. * gst/gstevent.c (_gst_event_initialize): Don't initialize trace. * gst/gstelement.h: * gst/gstelement.c (gst_element_iterate_src_pads) (gst_element_iterate_sink_pads): New API functions. * gst/gstelement.c (iterator_fold_with_resync): New utility, should fold into gstiterator.c in some form. (gst_element_pads_activate): Simplified via use of fold and delegation of decisions to gstpad->activate. * gst/gstbus.c (gst_bus_source_finalize): Set the bus to NULL, help in debugging. * gst/gstbuffer.c (_gst_buffer_initialize): Ref the buffer type class once in init, like gstmessage. Didn't run into this issue but it seems correct. Don't initialize a trace, gstminiobject does that. * check/pipelines/simple_launch_lines.c (test_stop_from_app): New test, runs fakesrc ! fakesink, stopping on ::handoff via a message to the bus. (assert_live_count): New util function, uses alloc traces to check cleanup. * check/gst/gstghostpad.c (test_ghost_pads): More refcount checks. To be modified when unlink drops the internal pad.
2005-06-27 18:35:05 +00:00
/**
* gst_pad_set_activatepull_function:
* @pad: a #GstPad.
* @activatepull: the #GstPadActivateModeFunction to set.
gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. Original commit message from CVS: 2005-06-27 Andy Wingo <wingo@pobox.com> * gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. * gst/gsttrace.c (gst_alloc_trace_list_sorted): New helper, returns a sorted copy of the trace list. (gst_alloc_trace_print_live): New API, only prints traces with live objects. Sort the list. (gst_alloc_trace_print_all): Sort the list. (gst_alloc_trace_print): Align columns. * gst/elements/gstttypefindelement.c: * gst/elements/gsttee.c: * gst/base/gstbasesrc.c: * gst/base/gstbasesink.c: * gst/base/gstbasetransform.c: * gst/gstqueue.c: Adapt for pad activation changes. * gst/gstpipeline.c (gst_pipeline_init): Unref after parenting sched. (gst_pipeline_dispose): Drop ref on sched. * gst/gstpad.c (gst_pad_init): Set the default activate func. (gst_pad_activate_default): Push mode by default. (pre_activate_switch, post_activate_switch): New stubs, things to do before and after switching activation modes on pads. (gst_pad_set_active): Take a boolean and not a mode, dispatch to the pad's activate function to choose which mode to activate. Shortcut on deactivation and call the right function directly. (gst_pad_activate_pull): New API, (de)activates a pad in pull mode. (gst_pad_activate_push): New API, same for push mode. (gst_pad_set_activate_function) (gst_pad_set_activatepull_function) (gst_pad_set_activatepush_function): Setters for new API. * gst/gstminiobject.c (gst_mini_object_new, gst_mini_object_free): Trace all miniobjects. (gst_mini_object_make_writable): Unref the arg if we copy, like gst_caps_make_writable. * gst/gstmessage.c (_gst_message_initialize): No trace init. * gst/gstghostpad.c (gst_proxy_pad_do_activate) (gst_proxy_pad_do_activatepull, gst_proxy_pad_do_activatepush): Adapt for new pad API. * gst/gstevent.c (_gst_event_initialize): Don't initialize trace. * gst/gstelement.h: * gst/gstelement.c (gst_element_iterate_src_pads) (gst_element_iterate_sink_pads): New API functions. * gst/gstelement.c (iterator_fold_with_resync): New utility, should fold into gstiterator.c in some form. (gst_element_pads_activate): Simplified via use of fold and delegation of decisions to gstpad->activate. * gst/gstbus.c (gst_bus_source_finalize): Set the bus to NULL, help in debugging. * gst/gstbuffer.c (_gst_buffer_initialize): Ref the buffer type class once in init, like gstmessage. Didn't run into this issue but it seems correct. Don't initialize a trace, gstminiobject does that. * check/pipelines/simple_launch_lines.c (test_stop_from_app): New test, runs fakesrc ! fakesink, stopping on ::handoff via a message to the bus. (assert_live_count): New util function, uses alloc traces to check cleanup. * check/gst/gstghostpad.c (test_ghost_pads): More refcount checks. To be modified when unlink drops the internal pad.
2005-06-27 18:35:05 +00:00
*
* Sets the given activate_pull function for the pad. An activate_pull function
* prepares the element and any upstream connections for pulling. See XXX
* part-activation.txt for details.
*/
void
gst_pad_set_activatepull_function (GstPad * pad,
GstPadActivateModeFunction activatepull)
{
g_return_if_fail (GST_IS_PAD (pad));
GST_PAD_ACTIVATEPULLFUNC (pad) = activatepull;
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_DEBUG_OBJECT (GST_CAT_PADS, pad, "activatepullfunc set to %s",
GST_DEBUG_FUNCPTR_NAME (activatepull));
gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. Original commit message from CVS: 2005-06-27 Andy Wingo <wingo@pobox.com> * gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. * gst/gsttrace.c (gst_alloc_trace_list_sorted): New helper, returns a sorted copy of the trace list. (gst_alloc_trace_print_live): New API, only prints traces with live objects. Sort the list. (gst_alloc_trace_print_all): Sort the list. (gst_alloc_trace_print): Align columns. * gst/elements/gstttypefindelement.c: * gst/elements/gsttee.c: * gst/base/gstbasesrc.c: * gst/base/gstbasesink.c: * gst/base/gstbasetransform.c: * gst/gstqueue.c: Adapt for pad activation changes. * gst/gstpipeline.c (gst_pipeline_init): Unref after parenting sched. (gst_pipeline_dispose): Drop ref on sched. * gst/gstpad.c (gst_pad_init): Set the default activate func. (gst_pad_activate_default): Push mode by default. (pre_activate_switch, post_activate_switch): New stubs, things to do before and after switching activation modes on pads. (gst_pad_set_active): Take a boolean and not a mode, dispatch to the pad's activate function to choose which mode to activate. Shortcut on deactivation and call the right function directly. (gst_pad_activate_pull): New API, (de)activates a pad in pull mode. (gst_pad_activate_push): New API, same for push mode. (gst_pad_set_activate_function) (gst_pad_set_activatepull_function) (gst_pad_set_activatepush_function): Setters for new API. * gst/gstminiobject.c (gst_mini_object_new, gst_mini_object_free): Trace all miniobjects. (gst_mini_object_make_writable): Unref the arg if we copy, like gst_caps_make_writable. * gst/gstmessage.c (_gst_message_initialize): No trace init. * gst/gstghostpad.c (gst_proxy_pad_do_activate) (gst_proxy_pad_do_activatepull, gst_proxy_pad_do_activatepush): Adapt for new pad API. * gst/gstevent.c (_gst_event_initialize): Don't initialize trace. * gst/gstelement.h: * gst/gstelement.c (gst_element_iterate_src_pads) (gst_element_iterate_sink_pads): New API functions. * gst/gstelement.c (iterator_fold_with_resync): New utility, should fold into gstiterator.c in some form. (gst_element_pads_activate): Simplified via use of fold and delegation of decisions to gstpad->activate. * gst/gstbus.c (gst_bus_source_finalize): Set the bus to NULL, help in debugging. * gst/gstbuffer.c (_gst_buffer_initialize): Ref the buffer type class once in init, like gstmessage. Didn't run into this issue but it seems correct. Don't initialize a trace, gstminiobject does that. * check/pipelines/simple_launch_lines.c (test_stop_from_app): New test, runs fakesrc ! fakesink, stopping on ::handoff via a message to the bus. (assert_live_count): New util function, uses alloc traces to check cleanup. * check/gst/gstghostpad.c (test_ghost_pads): More refcount checks. To be modified when unlink drops the internal pad.
2005-06-27 18:35:05 +00:00
}
/**
* gst_pad_set_activatepush_function:
* @pad: a #GstPad.
* @activatepush: the #GstPadActivateModeFunction to set.
gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. Original commit message from CVS: 2005-06-27 Andy Wingo <wingo@pobox.com> * gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. * gst/gsttrace.c (gst_alloc_trace_list_sorted): New helper, returns a sorted copy of the trace list. (gst_alloc_trace_print_live): New API, only prints traces with live objects. Sort the list. (gst_alloc_trace_print_all): Sort the list. (gst_alloc_trace_print): Align columns. * gst/elements/gstttypefindelement.c: * gst/elements/gsttee.c: * gst/base/gstbasesrc.c: * gst/base/gstbasesink.c: * gst/base/gstbasetransform.c: * gst/gstqueue.c: Adapt for pad activation changes. * gst/gstpipeline.c (gst_pipeline_init): Unref after parenting sched. (gst_pipeline_dispose): Drop ref on sched. * gst/gstpad.c (gst_pad_init): Set the default activate func. (gst_pad_activate_default): Push mode by default. (pre_activate_switch, post_activate_switch): New stubs, things to do before and after switching activation modes on pads. (gst_pad_set_active): Take a boolean and not a mode, dispatch to the pad's activate function to choose which mode to activate. Shortcut on deactivation and call the right function directly. (gst_pad_activate_pull): New API, (de)activates a pad in pull mode. (gst_pad_activate_push): New API, same for push mode. (gst_pad_set_activate_function) (gst_pad_set_activatepull_function) (gst_pad_set_activatepush_function): Setters for new API. * gst/gstminiobject.c (gst_mini_object_new, gst_mini_object_free): Trace all miniobjects. (gst_mini_object_make_writable): Unref the arg if we copy, like gst_caps_make_writable. * gst/gstmessage.c (_gst_message_initialize): No trace init. * gst/gstghostpad.c (gst_proxy_pad_do_activate) (gst_proxy_pad_do_activatepull, gst_proxy_pad_do_activatepush): Adapt for new pad API. * gst/gstevent.c (_gst_event_initialize): Don't initialize trace. * gst/gstelement.h: * gst/gstelement.c (gst_element_iterate_src_pads) (gst_element_iterate_sink_pads): New API functions. * gst/gstelement.c (iterator_fold_with_resync): New utility, should fold into gstiterator.c in some form. (gst_element_pads_activate): Simplified via use of fold and delegation of decisions to gstpad->activate. * gst/gstbus.c (gst_bus_source_finalize): Set the bus to NULL, help in debugging. * gst/gstbuffer.c (_gst_buffer_initialize): Ref the buffer type class once in init, like gstmessage. Didn't run into this issue but it seems correct. Don't initialize a trace, gstminiobject does that. * check/pipelines/simple_launch_lines.c (test_stop_from_app): New test, runs fakesrc ! fakesink, stopping on ::handoff via a message to the bus. (assert_live_count): New util function, uses alloc traces to check cleanup. * check/gst/gstghostpad.c (test_ghost_pads): More refcount checks. To be modified when unlink drops the internal pad.
2005-06-27 18:35:05 +00:00
*
* Sets the given activate_push function for the pad. An activate_push function
* prepares the element for pushing. See XXX part-activation.txt for details.
*/
void
gst_pad_set_activatepush_function (GstPad * pad,
GstPadActivateModeFunction activatepush)
{
g_return_if_fail (GST_IS_PAD (pad));
GST_PAD_ACTIVATEPUSHFUNC (pad) = activatepush;
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_DEBUG_OBJECT (GST_CAT_PADS, pad, "activatepushfunc set to %s",
GST_DEBUG_FUNCPTR_NAME (activatepush));
gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. Original commit message from CVS: 2005-06-27 Andy Wingo <wingo@pobox.com> * gst/base/gsttypefindhelper.c (gst_type_find_helper): Unref any remaining buffer. * gst/gsttrace.c (gst_alloc_trace_list_sorted): New helper, returns a sorted copy of the trace list. (gst_alloc_trace_print_live): New API, only prints traces with live objects. Sort the list. (gst_alloc_trace_print_all): Sort the list. (gst_alloc_trace_print): Align columns. * gst/elements/gstttypefindelement.c: * gst/elements/gsttee.c: * gst/base/gstbasesrc.c: * gst/base/gstbasesink.c: * gst/base/gstbasetransform.c: * gst/gstqueue.c: Adapt for pad activation changes. * gst/gstpipeline.c (gst_pipeline_init): Unref after parenting sched. (gst_pipeline_dispose): Drop ref on sched. * gst/gstpad.c (gst_pad_init): Set the default activate func. (gst_pad_activate_default): Push mode by default. (pre_activate_switch, post_activate_switch): New stubs, things to do before and after switching activation modes on pads. (gst_pad_set_active): Take a boolean and not a mode, dispatch to the pad's activate function to choose which mode to activate. Shortcut on deactivation and call the right function directly. (gst_pad_activate_pull): New API, (de)activates a pad in pull mode. (gst_pad_activate_push): New API, same for push mode. (gst_pad_set_activate_function) (gst_pad_set_activatepull_function) (gst_pad_set_activatepush_function): Setters for new API. * gst/gstminiobject.c (gst_mini_object_new, gst_mini_object_free): Trace all miniobjects. (gst_mini_object_make_writable): Unref the arg if we copy, like gst_caps_make_writable. * gst/gstmessage.c (_gst_message_initialize): No trace init. * gst/gstghostpad.c (gst_proxy_pad_do_activate) (gst_proxy_pad_do_activatepull, gst_proxy_pad_do_activatepush): Adapt for new pad API. * gst/gstevent.c (_gst_event_initialize): Don't initialize trace. * gst/gstelement.h: * gst/gstelement.c (gst_element_iterate_src_pads) (gst_element_iterate_sink_pads): New API functions. * gst/gstelement.c (iterator_fold_with_resync): New utility, should fold into gstiterator.c in some form. (gst_element_pads_activate): Simplified via use of fold and delegation of decisions to gstpad->activate. * gst/gstbus.c (gst_bus_source_finalize): Set the bus to NULL, help in debugging. * gst/gstbuffer.c (_gst_buffer_initialize): Ref the buffer type class once in init, like gstmessage. Didn't run into this issue but it seems correct. Don't initialize a trace, gstminiobject does that. * check/pipelines/simple_launch_lines.c (test_stop_from_app): New test, runs fakesrc ! fakesink, stopping on ::handoff via a message to the bus. (assert_live_count): New util function, uses alloc traces to check cleanup. * check/gst/gstghostpad.c (test_ghost_pads): More refcount checks. To be modified when unlink drops the internal pad.
2005-06-27 18:35:05 +00:00
}
/**
WARNING: Don't grab this updated unless you're really, REALLY sure. Original commit message from CVS: WARNING: Don't grab this updated unless you're really, REALLY sure. WARNING: Wait for the next one. Whole lotta changes here, including a few random bits: examples/*/Makefile: updated to use `libtool gcc`, not just `gcc` gst/ gstbuffer.h: updated to new flag style gst.c, gstdebug.h: added new debugging for function ptrs gstpipeline.c: set type of parent_class to the class, not the object gstthread.c: ditto plugins/ cdparanoia/cdparanoia.c: added an argument type, updated some defaults cobin/spindentity.c: updated to new do/while loopfunction style mp3encode/lame/gstlame.c: argument types, whole lotta lame options tests/: various changes Now, for the big changes: Once again, the scheduling system has changed. And once again, it broke a whole bunch of things. The gist of the change is that there is now a function pointer for gst_pad_push and gst_pad_pull, instead of a hard-wired function. Well, currently they are functions, but that's for debugging purposes only, they just call the function pointer after spewing lots of DEBUG(). This changed the GstPad structure a bit, and the GstPad API as well. Where elements used to provide chain() and pull() functions, they provide chain() and get() functions. gst_pad_set_pull[region]_function has been changed to get_pad_set_get[region]_function. This means all the elements out there that used to have pull functions need to be updated. The calls to that function have been changed in the normal elements, but the names of the functions passed is still _pull[region](), which is an aesthetic issue more than anything. As for what doesn't work yet, just about anything dealing with Connections is hosed, meaning threaded stuff won't work. This will be fixed about 12 hours from now, after I've slept, etc. The simplefake.c test works in both cothreaded and chained cases, but not much else will work due to the Connection problem. Needless to say, don't grab this unless you *need* these features *now*, else wait to update this stuff until tomorrow. I'm going to sleep now.
2000-12-16 10:18:09 +00:00
* gst_pad_set_chain_function:
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
* @pad: a sink #GstPad.
* @chain: the #GstPadChainFunction to set.
*
gst/gstelement.c (gst_element_dispose): Protect against multiple invocations. Original commit message from CVS: 2004-02-24 Andy Wingo <wingo@pobox.com> * gst/gstelement.c (gst_element_dispose): Protect against multiple invocations. * gst/schedulers/gstoptimalscheduler.c I added a mess of prototypes at the top of the file by way of documentation. Some of the operations on chains and groups were re-organized. (create_group): Added a type argument so if the group is enabled, the setup_group_scheduler knows what to do. (group_elements): Added a type argument here, too, to be passed on to create_group. (group_element_set_enabled): If an unlinked PLAYING element is added to a bin, we have to create a new group to hold the element, and this function will be called before the group is added to the chain. Thus we have a valid case for group->chain==NULL. Instead of calling chain_group_set_enabled, just set the flag on the group (the chain's status will be set when the group is added to it). (gst_opt_scheduler_state_transition, chain_group_set_enabled): Setup the group scheduler when the group is enabled, not specifically when an element goes PAUSED->PLAYING. This means PLAYING elements can be added, linked, and scheduled into a PLAYING pipeline, as was intended. (add_to_group): Don't ref the group twice. I don't know when this double-ref got in here. Removing it has the potential to cause segfaults if other parts of the scheduler are buggy. If you find that the scheduler is segfaulting for you, put in an extra ref here and see if that hacks over the underlying issue. Of course, then find out what code is unreffing a group it doesn't own... (create_group): Make the extra refcount floating, and remove it after adding the element. This means that... (unref_group): Destroy when the refcount reaches 0, not 1, like every other refcounted object in the known universe. (remove_from_group): When a group becomes empty, set it to be not active, and remove it from its chain. Don't unref it again, there's no floating reference any more. (destroy_group): We have to remove the group from the chain in remove_from_group (rather than here) to break refcounting cycles (the chain always has a ref on the group). So assert that group->chain==NULL. (ref_group_by_count): Removed, it was commented out anyway. (merge_chains): Use the remove_from_chain and add_to_chain primitives to do the reparenting, instead of rolling our own implementation. (add_to_chain): The first non-disabled group in the chain's group list will be the entry point for the chain. Because buffers can accumulate in loop elements' peer bufpens, we preferentially schedule loop groups before get groups to avoid unnecessary execution of get-based groups when the bufpens are already full. (gst_opt_scheduler_schedule_run_queue): Debug fixes. (get_group_schedule_function): Ditto. (loop_group_schedule_function): Ditto. (gst_opt_scheduler_loop_wrapper): Ditto. (gst_opt_scheduler_iterate): Ditto. I understand the opt scheduler now, yippee! * gst/gstpad.c: All throughout, added FIXMEs to look at for 0.9. (gst_pad_get_name, gst_pad_set_chain_function) (gst_pad_set_get_function, gst_pad_set_event_function) (gst_pad_set_event_mask_function, gst_pad_get_event_masks) (gst_pad_get_event_masks_default, gst_pad_set_convert_function) (gst_pad_set_query_function, gst_pad_get_query_types) (gst_pad_get_query_types_default) (gst_pad_set_internal_link_function) (gst_pad_set_formats_function, gst_pad_set_link_function) (gst_pad_set_fixate_function, gst_pad_set_getcaps_function) (gst_pad_set_bufferalloc_function, gst_pad_unlink) (gst_pad_renegotiate, gst_pad_set_parent, gst_pad_get_parent) (gst_pad_add_ghost_pad, gst_pad_proxy_getcaps) (gst_pad_proxy_pad_link, gst_pad_proxy_fixate) (gst_pad_get_pad_template_caps, gst_pad_check_compatibility) (gst_pad_get_peer, gst_pad_get_allowed_caps) (gst_pad_alloc_buffer, gst_pad_push, gst_pad_pull) (gst_pad_selectv, gst_pad_select, gst_pad_template_get_caps) (gst_pad_event_default_dispatch, gst_pad_event_default) (gst_pad_dispatcher, gst_pad_send_event, gst_pad_convert_default) (gst_pad_convert, gst_pad_query_default, gst_pad_query) (gst_pad_get_formats_default, gst_pad_get_formats): Better argument checks, and some doc fixes. (gst_pad_custom_new_from_template): Um, does anyone use these functions? Actually make a custom pad instead of a normal one. (gst_pad_try_set_caps): Transpose some checks. (gst_pad_try_set_caps_nonfixed): Same, and use a macro to check if the pad is in negotiation. (gst_pad_try_relink_filtered): Use pad_link_prepare. * gst/gstelement.c: Remove prototypes also defined in gstclock.h. * gst/gstelement.h: * gst/gstclock.h: Un-deprecate the old clocking API, as discussed on the list.
2004-02-25 13:16:12 +00:00
* Sets the given chain function for the pad. The chain function is called to
* process a #GstBuffer input buffer. see #GstPadChainFunction for more details.
*/
void
gst_pad_set_chain_function (GstPad * pad, GstPadChainFunction chain)
{
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
g_return_if_fail (GST_IS_PAD (pad));
g_return_if_fail (GST_PAD_DIRECTION (pad) == GST_PAD_SINK);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
GST_PAD_CHAINFUNC (pad) = chain;
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_DEBUG_OBJECT (GST_CAT_PADS, pad, "chainfunc set to %s",
GST_DEBUG_FUNCPTR_NAME (chain));
}
/**
* gst_pad_set_getrange_function:
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
* @pad: a source #GstPad.
* @get: the #GstPadGetRangeFunction to set.
*
* Sets the given getrange function for the pad. The getrange function is called to
* produce a new #GstBuffer to start the processing pipeline. see
* #GstPadGetRangeFunction for a description of the getrange function.
*/
void
gst_pad_set_getrange_function (GstPad * pad, GstPadGetRangeFunction get)
{
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
g_return_if_fail (GST_IS_PAD (pad));
g_return_if_fail (GST_PAD_DIRECTION (pad) == GST_PAD_SRC);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
GST_PAD_GETRANGEFUNC (pad) = get;
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_DEBUG_OBJECT (GST_CAT_PADS, pad, "getrangefunc set to %s",
GST_DEBUG_FUNCPTR_NAME (get));
}
gst/: More work on the generic source base class, implement seeking, query. Original commit message from CVS: * gst/base/README: * gst/base/gstbasesrc.c: (gst_basesrc_get_type), (gst_basesrc_init), (gst_basesrc_get_formats), (gst_basesrc_query), (gst_basesrc_get_event_mask), (gst_basesrc_do_seek), (gst_basesrc_event_handler), (gst_basesrc_get_range_unlocked), (gst_basesrc_check_get_range), (gst_basesrc_loop), (gst_basesrc_unlock), (gst_basesrc_get_size), (gst_basesrc_start), (gst_basesrc_stop), (gst_basesrc_activate), (gst_basesrc_change_state), (basesrc_find_peek), (basesrc_find_suggest), (gst_basesrc_type_find): * gst/base/gstbasesrc.h: * gst/elements/gstfilesrc.c: (gst_filesrc_base_init), (gst_filesrc_class_init), (gst_filesrc_init), (gst_filesrc_finalize), (gst_filesrc_set_location), (gst_filesrc_set_property), (gst_filesrc_get_property), (gst_filesrc_free_parent_mmap), (gst_filesrc_map_region), (gst_filesrc_map_small_region), (gst_filesrc_create_mmap), (gst_filesrc_create_read), (gst_filesrc_create), (gst_filesrc_get_size), (gst_filesrc_start), (gst_filesrc_stop): * gst/elements/gstfilesrc.h: * gst/gstelement.c: (gst_element_get_state_func), (gst_element_lost_state), (gst_element_pads_activate): * gst/gstpad.c: (gst_pad_set_active), (gst_pad_peer_set_active), (gst_pad_set_checkgetrange_function), (gst_pad_check_pull_range), (gst_pad_pull_range): * gst/gstpad.h: More work on the generic source base class, implement seeking, query. Make filesrc extend the base source class. Added gst_pad_set_checkgetrange_function to GstPad.
2005-04-06 17:30:48 +00:00
/**
* gst_pad_set_checkgetrange_function:
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
* @pad: a source #GstPad.
gst/: More work on the generic source base class, implement seeking, query. Original commit message from CVS: * gst/base/README: * gst/base/gstbasesrc.c: (gst_basesrc_get_type), (gst_basesrc_init), (gst_basesrc_get_formats), (gst_basesrc_query), (gst_basesrc_get_event_mask), (gst_basesrc_do_seek), (gst_basesrc_event_handler), (gst_basesrc_get_range_unlocked), (gst_basesrc_check_get_range), (gst_basesrc_loop), (gst_basesrc_unlock), (gst_basesrc_get_size), (gst_basesrc_start), (gst_basesrc_stop), (gst_basesrc_activate), (gst_basesrc_change_state), (basesrc_find_peek), (basesrc_find_suggest), (gst_basesrc_type_find): * gst/base/gstbasesrc.h: * gst/elements/gstfilesrc.c: (gst_filesrc_base_init), (gst_filesrc_class_init), (gst_filesrc_init), (gst_filesrc_finalize), (gst_filesrc_set_location), (gst_filesrc_set_property), (gst_filesrc_get_property), (gst_filesrc_free_parent_mmap), (gst_filesrc_map_region), (gst_filesrc_map_small_region), (gst_filesrc_create_mmap), (gst_filesrc_create_read), (gst_filesrc_create), (gst_filesrc_get_size), (gst_filesrc_start), (gst_filesrc_stop): * gst/elements/gstfilesrc.h: * gst/gstelement.c: (gst_element_get_state_func), (gst_element_lost_state), (gst_element_pads_activate): * gst/gstpad.c: (gst_pad_set_active), (gst_pad_peer_set_active), (gst_pad_set_checkgetrange_function), (gst_pad_check_pull_range), (gst_pad_pull_range): * gst/gstpad.h: More work on the generic source base class, implement seeking, query. Make filesrc extend the base source class. Added gst_pad_set_checkgetrange_function to GstPad.
2005-04-06 17:30:48 +00:00
* @check: the #GstPadCheckGetRangeFunction to set.
*
* Sets the given checkgetrange function for the pad. Implement this function on
* a pad if you dynamically support getrange based scheduling on the pad.
*/
void
gst_pad_set_checkgetrange_function (GstPad * pad,
GstPadCheckGetRangeFunction check)
{
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
g_return_if_fail (GST_IS_PAD (pad));
g_return_if_fail (GST_PAD_DIRECTION (pad) == GST_PAD_SRC);
gst/: More work on the generic source base class, implement seeking, query. Original commit message from CVS: * gst/base/README: * gst/base/gstbasesrc.c: (gst_basesrc_get_type), (gst_basesrc_init), (gst_basesrc_get_formats), (gst_basesrc_query), (gst_basesrc_get_event_mask), (gst_basesrc_do_seek), (gst_basesrc_event_handler), (gst_basesrc_get_range_unlocked), (gst_basesrc_check_get_range), (gst_basesrc_loop), (gst_basesrc_unlock), (gst_basesrc_get_size), (gst_basesrc_start), (gst_basesrc_stop), (gst_basesrc_activate), (gst_basesrc_change_state), (basesrc_find_peek), (basesrc_find_suggest), (gst_basesrc_type_find): * gst/base/gstbasesrc.h: * gst/elements/gstfilesrc.c: (gst_filesrc_base_init), (gst_filesrc_class_init), (gst_filesrc_init), (gst_filesrc_finalize), (gst_filesrc_set_location), (gst_filesrc_set_property), (gst_filesrc_get_property), (gst_filesrc_free_parent_mmap), (gst_filesrc_map_region), (gst_filesrc_map_small_region), (gst_filesrc_create_mmap), (gst_filesrc_create_read), (gst_filesrc_create), (gst_filesrc_get_size), (gst_filesrc_start), (gst_filesrc_stop): * gst/elements/gstfilesrc.h: * gst/gstelement.c: (gst_element_get_state_func), (gst_element_lost_state), (gst_element_pads_activate): * gst/gstpad.c: (gst_pad_set_active), (gst_pad_peer_set_active), (gst_pad_set_checkgetrange_function), (gst_pad_check_pull_range), (gst_pad_pull_range): * gst/gstpad.h: More work on the generic source base class, implement seeking, query. Make filesrc extend the base source class. Added gst_pad_set_checkgetrange_function to GstPad.
2005-04-06 17:30:48 +00:00
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
GST_PAD_CHECKGETRANGEFUNC (pad) = check;
gst/: More work on the generic source base class, implement seeking, query. Original commit message from CVS: * gst/base/README: * gst/base/gstbasesrc.c: (gst_basesrc_get_type), (gst_basesrc_init), (gst_basesrc_get_formats), (gst_basesrc_query), (gst_basesrc_get_event_mask), (gst_basesrc_do_seek), (gst_basesrc_event_handler), (gst_basesrc_get_range_unlocked), (gst_basesrc_check_get_range), (gst_basesrc_loop), (gst_basesrc_unlock), (gst_basesrc_get_size), (gst_basesrc_start), (gst_basesrc_stop), (gst_basesrc_activate), (gst_basesrc_change_state), (basesrc_find_peek), (basesrc_find_suggest), (gst_basesrc_type_find): * gst/base/gstbasesrc.h: * gst/elements/gstfilesrc.c: (gst_filesrc_base_init), (gst_filesrc_class_init), (gst_filesrc_init), (gst_filesrc_finalize), (gst_filesrc_set_location), (gst_filesrc_set_property), (gst_filesrc_get_property), (gst_filesrc_free_parent_mmap), (gst_filesrc_map_region), (gst_filesrc_map_small_region), (gst_filesrc_create_mmap), (gst_filesrc_create_read), (gst_filesrc_create), (gst_filesrc_get_size), (gst_filesrc_start), (gst_filesrc_stop): * gst/elements/gstfilesrc.h: * gst/gstelement.c: (gst_element_get_state_func), (gst_element_lost_state), (gst_element_pads_activate): * gst/gstpad.c: (gst_pad_set_active), (gst_pad_peer_set_active), (gst_pad_set_checkgetrange_function), (gst_pad_check_pull_range), (gst_pad_pull_range): * gst/gstpad.h: More work on the generic source base class, implement seeking, query. Make filesrc extend the base source class. Added gst_pad_set_checkgetrange_function to GstPad.
2005-04-06 17:30:48 +00:00
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_DEBUG_OBJECT (GST_CAT_PADS, pad, "checkgetrangefunc set to %s",
GST_DEBUG_FUNCPTR_NAME (check));
gst/: More work on the generic source base class, implement seeking, query. Original commit message from CVS: * gst/base/README: * gst/base/gstbasesrc.c: (gst_basesrc_get_type), (gst_basesrc_init), (gst_basesrc_get_formats), (gst_basesrc_query), (gst_basesrc_get_event_mask), (gst_basesrc_do_seek), (gst_basesrc_event_handler), (gst_basesrc_get_range_unlocked), (gst_basesrc_check_get_range), (gst_basesrc_loop), (gst_basesrc_unlock), (gst_basesrc_get_size), (gst_basesrc_start), (gst_basesrc_stop), (gst_basesrc_activate), (gst_basesrc_change_state), (basesrc_find_peek), (basesrc_find_suggest), (gst_basesrc_type_find): * gst/base/gstbasesrc.h: * gst/elements/gstfilesrc.c: (gst_filesrc_base_init), (gst_filesrc_class_init), (gst_filesrc_init), (gst_filesrc_finalize), (gst_filesrc_set_location), (gst_filesrc_set_property), (gst_filesrc_get_property), (gst_filesrc_free_parent_mmap), (gst_filesrc_map_region), (gst_filesrc_map_small_region), (gst_filesrc_create_mmap), (gst_filesrc_create_read), (gst_filesrc_create), (gst_filesrc_get_size), (gst_filesrc_start), (gst_filesrc_stop): * gst/elements/gstfilesrc.h: * gst/gstelement.c: (gst_element_get_state_func), (gst_element_lost_state), (gst_element_pads_activate): * gst/gstpad.c: (gst_pad_set_active), (gst_pad_peer_set_active), (gst_pad_set_checkgetrange_function), (gst_pad_check_pull_range), (gst_pad_pull_range): * gst/gstpad.h: More work on the generic source base class, implement seeking, query. Make filesrc extend the base source class. Added gst_pad_set_checkgetrange_function to GstPad.
2005-04-06 17:30:48 +00:00
}
/**
* gst_pad_set_event_function:
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
* @pad: a source #GstPad.
* @event: the #GstPadEventFunction to set.
*
* Sets the given event handler for the pad.
*/
void
gst_pad_set_event_function (GstPad * pad, GstPadEventFunction event)
{
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
g_return_if_fail (GST_IS_PAD (pad));
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
GST_PAD_EVENTFUNC (pad) = event;
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_DEBUG_OBJECT (GST_CAT_PADS, pad, "eventfunc for set to %s",
GST_DEBUG_FUNCPTR_NAME (event));
}
/**
* gst_pad_set_query_function:
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
* @pad: a #GstPad of either direction.
* @query: the #GstPadQueryFunction to set.
*
* Set the given query function for the pad.
*/
void
gst_pad_set_query_function (GstPad * pad, GstPadQueryFunction query)
{
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
g_return_if_fail (GST_IS_PAD (pad));
WARNING: Don't grab this updated unless you're really, REALLY sure. Original commit message from CVS: WARNING: Don't grab this updated unless you're really, REALLY sure. WARNING: Wait for the next one. Whole lotta changes here, including a few random bits: examples/*/Makefile: updated to use `libtool gcc`, not just `gcc` gst/ gstbuffer.h: updated to new flag style gst.c, gstdebug.h: added new debugging for function ptrs gstpipeline.c: set type of parent_class to the class, not the object gstthread.c: ditto plugins/ cdparanoia/cdparanoia.c: added an argument type, updated some defaults cobin/spindentity.c: updated to new do/while loopfunction style mp3encode/lame/gstlame.c: argument types, whole lotta lame options tests/: various changes Now, for the big changes: Once again, the scheduling system has changed. And once again, it broke a whole bunch of things. The gist of the change is that there is now a function pointer for gst_pad_push and gst_pad_pull, instead of a hard-wired function. Well, currently they are functions, but that's for debugging purposes only, they just call the function pointer after spewing lots of DEBUG(). This changed the GstPad structure a bit, and the GstPad API as well. Where elements used to provide chain() and pull() functions, they provide chain() and get() functions. gst_pad_set_pull[region]_function has been changed to get_pad_set_get[region]_function. This means all the elements out there that used to have pull functions need to be updated. The calls to that function have been changed in the normal elements, but the names of the functions passed is still _pull[region](), which is an aesthetic issue more than anything. As for what doesn't work yet, just about anything dealing with Connections is hosed, meaning threaded stuff won't work. This will be fixed about 12 hours from now, after I've slept, etc. The simplefake.c test works in both cothreaded and chained cases, but not much else will work due to the Connection problem. Needless to say, don't grab this unless you *need* these features *now*, else wait to update this stuff until tomorrow. I'm going to sleep now.
2000-12-16 10:18:09 +00:00
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
GST_PAD_QUERYFUNC (pad) = query;
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_DEBUG_OBJECT (GST_CAT_PADS, pad, "queryfunc set to %s",
GST_DEBUG_FUNCPTR_NAME (query));
}
/**
* gst_pad_set_query_type_function:
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
* @pad: a #GstPad of either direction.
* @type_func: the #GstPadQueryTypeFunction to set.
*
* Set the given query type function for the pad.
*/
void
gst_pad_set_query_type_function (GstPad * pad,
GstPadQueryTypeFunction type_func)
{
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
g_return_if_fail (GST_IS_PAD (pad));
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
GST_PAD_QUERYTYPEFUNC (pad) = type_func;
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_DEBUG_OBJECT (GST_CAT_PADS, pad, "querytypefunc set to %s",
GST_DEBUG_FUNCPTR_NAME (type_func));
}
/**
* gst_pad_get_query_types:
gst/gstelement.c (gst_element_dispose): Protect against multiple invocations. Original commit message from CVS: 2004-02-24 Andy Wingo <wingo@pobox.com> * gst/gstelement.c (gst_element_dispose): Protect against multiple invocations. * gst/schedulers/gstoptimalscheduler.c I added a mess of prototypes at the top of the file by way of documentation. Some of the operations on chains and groups were re-organized. (create_group): Added a type argument so if the group is enabled, the setup_group_scheduler knows what to do. (group_elements): Added a type argument here, too, to be passed on to create_group. (group_element_set_enabled): If an unlinked PLAYING element is added to a bin, we have to create a new group to hold the element, and this function will be called before the group is added to the chain. Thus we have a valid case for group->chain==NULL. Instead of calling chain_group_set_enabled, just set the flag on the group (the chain's status will be set when the group is added to it). (gst_opt_scheduler_state_transition, chain_group_set_enabled): Setup the group scheduler when the group is enabled, not specifically when an element goes PAUSED->PLAYING. This means PLAYING elements can be added, linked, and scheduled into a PLAYING pipeline, as was intended. (add_to_group): Don't ref the group twice. I don't know when this double-ref got in here. Removing it has the potential to cause segfaults if other parts of the scheduler are buggy. If you find that the scheduler is segfaulting for you, put in an extra ref here and see if that hacks over the underlying issue. Of course, then find out what code is unreffing a group it doesn't own... (create_group): Make the extra refcount floating, and remove it after adding the element. This means that... (unref_group): Destroy when the refcount reaches 0, not 1, like every other refcounted object in the known universe. (remove_from_group): When a group becomes empty, set it to be not active, and remove it from its chain. Don't unref it again, there's no floating reference any more. (destroy_group): We have to remove the group from the chain in remove_from_group (rather than here) to break refcounting cycles (the chain always has a ref on the group). So assert that group->chain==NULL. (ref_group_by_count): Removed, it was commented out anyway. (merge_chains): Use the remove_from_chain and add_to_chain primitives to do the reparenting, instead of rolling our own implementation. (add_to_chain): The first non-disabled group in the chain's group list will be the entry point for the chain. Because buffers can accumulate in loop elements' peer bufpens, we preferentially schedule loop groups before get groups to avoid unnecessary execution of get-based groups when the bufpens are already full. (gst_opt_scheduler_schedule_run_queue): Debug fixes. (get_group_schedule_function): Ditto. (loop_group_schedule_function): Ditto. (gst_opt_scheduler_loop_wrapper): Ditto. (gst_opt_scheduler_iterate): Ditto. I understand the opt scheduler now, yippee! * gst/gstpad.c: All throughout, added FIXMEs to look at for 0.9. (gst_pad_get_name, gst_pad_set_chain_function) (gst_pad_set_get_function, gst_pad_set_event_function) (gst_pad_set_event_mask_function, gst_pad_get_event_masks) (gst_pad_get_event_masks_default, gst_pad_set_convert_function) (gst_pad_set_query_function, gst_pad_get_query_types) (gst_pad_get_query_types_default) (gst_pad_set_internal_link_function) (gst_pad_set_formats_function, gst_pad_set_link_function) (gst_pad_set_fixate_function, gst_pad_set_getcaps_function) (gst_pad_set_bufferalloc_function, gst_pad_unlink) (gst_pad_renegotiate, gst_pad_set_parent, gst_pad_get_parent) (gst_pad_add_ghost_pad, gst_pad_proxy_getcaps) (gst_pad_proxy_pad_link, gst_pad_proxy_fixate) (gst_pad_get_pad_template_caps, gst_pad_check_compatibility) (gst_pad_get_peer, gst_pad_get_allowed_caps) (gst_pad_alloc_buffer, gst_pad_push, gst_pad_pull) (gst_pad_selectv, gst_pad_select, gst_pad_template_get_caps) (gst_pad_event_default_dispatch, gst_pad_event_default) (gst_pad_dispatcher, gst_pad_send_event, gst_pad_convert_default) (gst_pad_convert, gst_pad_query_default, gst_pad_query) (gst_pad_get_formats_default, gst_pad_get_formats): Better argument checks, and some doc fixes. (gst_pad_custom_new_from_template): Um, does anyone use these functions? Actually make a custom pad instead of a normal one. (gst_pad_try_set_caps): Transpose some checks. (gst_pad_try_set_caps_nonfixed): Same, and use a macro to check if the pad is in negotiation. (gst_pad_try_relink_filtered): Use pad_link_prepare. * gst/gstelement.c: Remove prototypes also defined in gstclock.h. * gst/gstelement.h: * gst/gstclock.h: Un-deprecate the old clocking API, as discussed on the list.
2004-02-25 13:16:12 +00:00
* @pad: a #GstPad.
*
* Get an array of supported queries that can be performed
* on this pad.
*
gst/gstpad.c (gst_pad_custom_new): Add a FIXME, this is a hacky way to do inheritance. Original commit message from CVS: 2004-02-10 Andy Wingo <wingo@pobox.com> * gst/gstpad.c (gst_pad_custom_new): Add a FIXME, this is a hacky way to do inheritance. (gst_pad_get_event_masks, gst_pad_get_event_masks_default) (gst_pad_get_query_types, gst_pad_get_query_types_default): Routine docs. (gst_pad_set_link_function, gst_pad_set_fixate_function) (gst_pad_set_getcaps_function): Doc from Dave's negotation random doc. (gst_pad_unlink, gst_pad_is_linked): Docs. (gst_pad_renegotiate): A brief description of capsnego. (gst_pad_try_set_caps): Document. (gst_pad_try_set_caps_nonfixed): Document. (gst_pad_can_link_filtered, gst_pad_link_filtered): Doc fixes. (gst_pad_set_parent): Deprecated (although not out of the API). (gst_pad_get_parent): Deprecated, although many plugins use this. (gst_pad_add_ghost_pad, gst_pad_remove_ghost_pad): Doc that these are private and will go away in 0.9. (gst_pad_perform_negotiate): Doc. (gst_pad_link_unnegotiate): I think this is meant to be static. (gst_pad_get_negotiated_caps, gst_pad_get_pad_template_caps) (gst_pad_template_get_caps_by_name, gst_pad_check_compatibility) (gst_pad_get_peer): Doc updates. (gst_pad_caps_change_notify): Doc. (gst_pad_alloc_buffer, gst_pad_push, gst_static_pad_template_get) (gst_ghost_pad_new): Doc fixes. * gst/gstobject.c (gst_object_get_parent, gst_object_unparent) (gst_object_check_uniqueness): * gst/gstelement.c (gst_element_add_pad) (gst_element_add_ghost_pad, gst_element_remove_pad) (gst_element_remove_ghost_pad, gst_element_get_pad) (gst_element_get_static_pad, gst_element_get_pad_list) (gst_element_class_get_pad_template_list) (gst_element_class_get_pad_template): Work on the docs. (gst_element_get_pad_template_list): Uses the class method. (gst_element_get_compatible_pad_template): Docs, and consolidate some test conditions. (gst_element_get_pad_from_template): New static function. (gst_element_request_compatible_pad): Docs, and work with non-request compatible templates. (gst_element_get_compatible_pad_filtered): Docs and remove redundant checks. (gst_element_get_compatible_pad, gst_element_link_pads_filtered) (gst_element_link_filtered, gst_element_link_many) (gst_element_link, gst_element_link_pads) (gst_element_unlink_many): Docs. 2004-02-05 Andy Wingo <wingo@pobox.com> * gst/gstpad.c (_gst_real_pad_fixate_accumulator): s/pointer/boxed/. * gst/gstmarshal.list (VOID:BOXED, BOXED:BOXED): New marshallers. * gst/gstpad.c (gst_real_pad_class_init): Use a BOXED:BOXED marshaller for ::fixate, and VOID:BOXED for ::caps-nego-failed, with the type=GST_TYPE_CAPS. This allows language bindings to know what kind of data they're dealing with. * gst/gstcaps.c (_gst_caps_value_init): GBoxed values initialize to NULL when g_value_init is called. GstCaps, which rolls its own type implementation, now does the same instead of allocating empty caps. (_gst_caps_initialize, _gst_caps_collect_value, _gst_caps_lcopy_value): Provide collect_value and lcopy_value type table methods. This allows G_VALUE_COLLECT to work. 2004-02-05 Andy Wingo <wingo@pobox.com> * configure.ac: * testsuite/Makefile.am (SUBDIRS): * testsuite/ghostpads/Makefile.am: * testsuite/ghostpads/ghostpads.c: A new test for ghost pads. * gst/gstpad.c (gst_pad_add_ghost_pad, gst_pad_remove_ghost_pad): These two routines are the only ones that set GST_GPAD_REALPAD(gpad), the ghost pad list, and the ghost pad's pad template. They should be made static, depending on ABI needs. (gst_real_pad_dispose): Handle the case of ghost pads without a parent. Assert after dealing with ghost pads that the ghost pad list is empty. (gst_ghost_pad_class_init): New property added, ::real-pad. Can be set after creation. (gst_ghost_pad_dispose): Set ::real-pad to NULL. (gst_ghost_pad_set_property, gst_ghost_pad_get_property): New functions. set_property will call add_ghost_pad/remove_ghost_pad as appropriate. (gst_ghost_pad_new): All the work is offloaded to g_object_new. * gst/gstelement.c (gst_element_add_pad): Handle ghost pads as well. (gst_element_add_ghost_pad): Remove code duplicated from _add_pad. (gst_element_remove_pad): Handle ghost pads as well. (gst_element_remove_ghost_pad): Deprecated (could be removed, depending on API-stability needs). 2004-02-05 Andy Wingo <wingo@pobox.com> * gst/gstbin.[ch]: (gst_bin_get_by_interface): GTypes are scalars, of course they're const
2004-02-11 13:26:04 +00:00
* Returns: a zero-terminated array of #GstQueryType.
*/
const GstQueryType *
gst_pad_get_query_types (GstPad * pad)
{
GstPadQueryTypeFunction func;
gst/gstelement.c (gst_element_dispose): Protect against multiple invocations. Original commit message from CVS: 2004-02-24 Andy Wingo <wingo@pobox.com> * gst/gstelement.c (gst_element_dispose): Protect against multiple invocations. * gst/schedulers/gstoptimalscheduler.c I added a mess of prototypes at the top of the file by way of documentation. Some of the operations on chains and groups were re-organized. (create_group): Added a type argument so if the group is enabled, the setup_group_scheduler knows what to do. (group_elements): Added a type argument here, too, to be passed on to create_group. (group_element_set_enabled): If an unlinked PLAYING element is added to a bin, we have to create a new group to hold the element, and this function will be called before the group is added to the chain. Thus we have a valid case for group->chain==NULL. Instead of calling chain_group_set_enabled, just set the flag on the group (the chain's status will be set when the group is added to it). (gst_opt_scheduler_state_transition, chain_group_set_enabled): Setup the group scheduler when the group is enabled, not specifically when an element goes PAUSED->PLAYING. This means PLAYING elements can be added, linked, and scheduled into a PLAYING pipeline, as was intended. (add_to_group): Don't ref the group twice. I don't know when this double-ref got in here. Removing it has the potential to cause segfaults if other parts of the scheduler are buggy. If you find that the scheduler is segfaulting for you, put in an extra ref here and see if that hacks over the underlying issue. Of course, then find out what code is unreffing a group it doesn't own... (create_group): Make the extra refcount floating, and remove it after adding the element. This means that... (unref_group): Destroy when the refcount reaches 0, not 1, like every other refcounted object in the known universe. (remove_from_group): When a group becomes empty, set it to be not active, and remove it from its chain. Don't unref it again, there's no floating reference any more. (destroy_group): We have to remove the group from the chain in remove_from_group (rather than here) to break refcounting cycles (the chain always has a ref on the group). So assert that group->chain==NULL. (ref_group_by_count): Removed, it was commented out anyway. (merge_chains): Use the remove_from_chain and add_to_chain primitives to do the reparenting, instead of rolling our own implementation. (add_to_chain): The first non-disabled group in the chain's group list will be the entry point for the chain. Because buffers can accumulate in loop elements' peer bufpens, we preferentially schedule loop groups before get groups to avoid unnecessary execution of get-based groups when the bufpens are already full. (gst_opt_scheduler_schedule_run_queue): Debug fixes. (get_group_schedule_function): Ditto. (loop_group_schedule_function): Ditto. (gst_opt_scheduler_loop_wrapper): Ditto. (gst_opt_scheduler_iterate): Ditto. I understand the opt scheduler now, yippee! * gst/gstpad.c: All throughout, added FIXMEs to look at for 0.9. (gst_pad_get_name, gst_pad_set_chain_function) (gst_pad_set_get_function, gst_pad_set_event_function) (gst_pad_set_event_mask_function, gst_pad_get_event_masks) (gst_pad_get_event_masks_default, gst_pad_set_convert_function) (gst_pad_set_query_function, gst_pad_get_query_types) (gst_pad_get_query_types_default) (gst_pad_set_internal_link_function) (gst_pad_set_formats_function, gst_pad_set_link_function) (gst_pad_set_fixate_function, gst_pad_set_getcaps_function) (gst_pad_set_bufferalloc_function, gst_pad_unlink) (gst_pad_renegotiate, gst_pad_set_parent, gst_pad_get_parent) (gst_pad_add_ghost_pad, gst_pad_proxy_getcaps) (gst_pad_proxy_pad_link, gst_pad_proxy_fixate) (gst_pad_get_pad_template_caps, gst_pad_check_compatibility) (gst_pad_get_peer, gst_pad_get_allowed_caps) (gst_pad_alloc_buffer, gst_pad_push, gst_pad_pull) (gst_pad_selectv, gst_pad_select, gst_pad_template_get_caps) (gst_pad_event_default_dispatch, gst_pad_event_default) (gst_pad_dispatcher, gst_pad_send_event, gst_pad_convert_default) (gst_pad_convert, gst_pad_query_default, gst_pad_query) (gst_pad_get_formats_default, gst_pad_get_formats): Better argument checks, and some doc fixes. (gst_pad_custom_new_from_template): Um, does anyone use these functions? Actually make a custom pad instead of a normal one. (gst_pad_try_set_caps): Transpose some checks. (gst_pad_try_set_caps_nonfixed): Same, and use a macro to check if the pad is in negotiation. (gst_pad_try_relink_filtered): Use pad_link_prepare. * gst/gstelement.c: Remove prototypes also defined in gstclock.h. * gst/gstelement.h: * gst/gstclock.h: Un-deprecate the old clocking API, as discussed on the list.
2004-02-25 13:16:12 +00:00
g_return_val_if_fail (GST_IS_PAD (pad), NULL);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
if (G_UNLIKELY ((func = GST_PAD_QUERYTYPEFUNC (pad)) == NULL))
goto no_func;
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
return func (pad);
no_func:
{
return NULL;
}
}
static gboolean
gst_pad_get_query_types_dispatcher (GstPad * pad, const GstQueryType ** data)
{
*data = gst_pad_get_query_types (pad);
return TRUE;
}
/**
* gst_pad_get_query_types_default:
gst/gstelement.c (gst_element_dispose): Protect against multiple invocations. Original commit message from CVS: 2004-02-24 Andy Wingo <wingo@pobox.com> * gst/gstelement.c (gst_element_dispose): Protect against multiple invocations. * gst/schedulers/gstoptimalscheduler.c I added a mess of prototypes at the top of the file by way of documentation. Some of the operations on chains and groups were re-organized. (create_group): Added a type argument so if the group is enabled, the setup_group_scheduler knows what to do. (group_elements): Added a type argument here, too, to be passed on to create_group. (group_element_set_enabled): If an unlinked PLAYING element is added to a bin, we have to create a new group to hold the element, and this function will be called before the group is added to the chain. Thus we have a valid case for group->chain==NULL. Instead of calling chain_group_set_enabled, just set the flag on the group (the chain's status will be set when the group is added to it). (gst_opt_scheduler_state_transition, chain_group_set_enabled): Setup the group scheduler when the group is enabled, not specifically when an element goes PAUSED->PLAYING. This means PLAYING elements can be added, linked, and scheduled into a PLAYING pipeline, as was intended. (add_to_group): Don't ref the group twice. I don't know when this double-ref got in here. Removing it has the potential to cause segfaults if other parts of the scheduler are buggy. If you find that the scheduler is segfaulting for you, put in an extra ref here and see if that hacks over the underlying issue. Of course, then find out what code is unreffing a group it doesn't own... (create_group): Make the extra refcount floating, and remove it after adding the element. This means that... (unref_group): Destroy when the refcount reaches 0, not 1, like every other refcounted object in the known universe. (remove_from_group): When a group becomes empty, set it to be not active, and remove it from its chain. Don't unref it again, there's no floating reference any more. (destroy_group): We have to remove the group from the chain in remove_from_group (rather than here) to break refcounting cycles (the chain always has a ref on the group). So assert that group->chain==NULL. (ref_group_by_count): Removed, it was commented out anyway. (merge_chains): Use the remove_from_chain and add_to_chain primitives to do the reparenting, instead of rolling our own implementation. (add_to_chain): The first non-disabled group in the chain's group list will be the entry point for the chain. Because buffers can accumulate in loop elements' peer bufpens, we preferentially schedule loop groups before get groups to avoid unnecessary execution of get-based groups when the bufpens are already full. (gst_opt_scheduler_schedule_run_queue): Debug fixes. (get_group_schedule_function): Ditto. (loop_group_schedule_function): Ditto. (gst_opt_scheduler_loop_wrapper): Ditto. (gst_opt_scheduler_iterate): Ditto. I understand the opt scheduler now, yippee! * gst/gstpad.c: All throughout, added FIXMEs to look at for 0.9. (gst_pad_get_name, gst_pad_set_chain_function) (gst_pad_set_get_function, gst_pad_set_event_function) (gst_pad_set_event_mask_function, gst_pad_get_event_masks) (gst_pad_get_event_masks_default, gst_pad_set_convert_function) (gst_pad_set_query_function, gst_pad_get_query_types) (gst_pad_get_query_types_default) (gst_pad_set_internal_link_function) (gst_pad_set_formats_function, gst_pad_set_link_function) (gst_pad_set_fixate_function, gst_pad_set_getcaps_function) (gst_pad_set_bufferalloc_function, gst_pad_unlink) (gst_pad_renegotiate, gst_pad_set_parent, gst_pad_get_parent) (gst_pad_add_ghost_pad, gst_pad_proxy_getcaps) (gst_pad_proxy_pad_link, gst_pad_proxy_fixate) (gst_pad_get_pad_template_caps, gst_pad_check_compatibility) (gst_pad_get_peer, gst_pad_get_allowed_caps) (gst_pad_alloc_buffer, gst_pad_push, gst_pad_pull) (gst_pad_selectv, gst_pad_select, gst_pad_template_get_caps) (gst_pad_event_default_dispatch, gst_pad_event_default) (gst_pad_dispatcher, gst_pad_send_event, gst_pad_convert_default) (gst_pad_convert, gst_pad_query_default, gst_pad_query) (gst_pad_get_formats_default, gst_pad_get_formats): Better argument checks, and some doc fixes. (gst_pad_custom_new_from_template): Um, does anyone use these functions? Actually make a custom pad instead of a normal one. (gst_pad_try_set_caps): Transpose some checks. (gst_pad_try_set_caps_nonfixed): Same, and use a macro to check if the pad is in negotiation. (gst_pad_try_relink_filtered): Use pad_link_prepare. * gst/gstelement.c: Remove prototypes also defined in gstclock.h. * gst/gstelement.h: * gst/gstclock.h: Un-deprecate the old clocking API, as discussed on the list.
2004-02-25 13:16:12 +00:00
* @pad: a #GstPad.
*
* Invoke the default dispatcher for the query types on
* the pad.
*
gst/gstpad.c (gst_pad_custom_new): Add a FIXME, this is a hacky way to do inheritance. Original commit message from CVS: 2004-02-10 Andy Wingo <wingo@pobox.com> * gst/gstpad.c (gst_pad_custom_new): Add a FIXME, this is a hacky way to do inheritance. (gst_pad_get_event_masks, gst_pad_get_event_masks_default) (gst_pad_get_query_types, gst_pad_get_query_types_default): Routine docs. (gst_pad_set_link_function, gst_pad_set_fixate_function) (gst_pad_set_getcaps_function): Doc from Dave's negotation random doc. (gst_pad_unlink, gst_pad_is_linked): Docs. (gst_pad_renegotiate): A brief description of capsnego. (gst_pad_try_set_caps): Document. (gst_pad_try_set_caps_nonfixed): Document. (gst_pad_can_link_filtered, gst_pad_link_filtered): Doc fixes. (gst_pad_set_parent): Deprecated (although not out of the API). (gst_pad_get_parent): Deprecated, although many plugins use this. (gst_pad_add_ghost_pad, gst_pad_remove_ghost_pad): Doc that these are private and will go away in 0.9. (gst_pad_perform_negotiate): Doc. (gst_pad_link_unnegotiate): I think this is meant to be static. (gst_pad_get_negotiated_caps, gst_pad_get_pad_template_caps) (gst_pad_template_get_caps_by_name, gst_pad_check_compatibility) (gst_pad_get_peer): Doc updates. (gst_pad_caps_change_notify): Doc. (gst_pad_alloc_buffer, gst_pad_push, gst_static_pad_template_get) (gst_ghost_pad_new): Doc fixes. * gst/gstobject.c (gst_object_get_parent, gst_object_unparent) (gst_object_check_uniqueness): * gst/gstelement.c (gst_element_add_pad) (gst_element_add_ghost_pad, gst_element_remove_pad) (gst_element_remove_ghost_pad, gst_element_get_pad) (gst_element_get_static_pad, gst_element_get_pad_list) (gst_element_class_get_pad_template_list) (gst_element_class_get_pad_template): Work on the docs. (gst_element_get_pad_template_list): Uses the class method. (gst_element_get_compatible_pad_template): Docs, and consolidate some test conditions. (gst_element_get_pad_from_template): New static function. (gst_element_request_compatible_pad): Docs, and work with non-request compatible templates. (gst_element_get_compatible_pad_filtered): Docs and remove redundant checks. (gst_element_get_compatible_pad, gst_element_link_pads_filtered) (gst_element_link_filtered, gst_element_link_many) (gst_element_link, gst_element_link_pads) (gst_element_unlink_many): Docs. 2004-02-05 Andy Wingo <wingo@pobox.com> * gst/gstpad.c (_gst_real_pad_fixate_accumulator): s/pointer/boxed/. * gst/gstmarshal.list (VOID:BOXED, BOXED:BOXED): New marshallers. * gst/gstpad.c (gst_real_pad_class_init): Use a BOXED:BOXED marshaller for ::fixate, and VOID:BOXED for ::caps-nego-failed, with the type=GST_TYPE_CAPS. This allows language bindings to know what kind of data they're dealing with. * gst/gstcaps.c (_gst_caps_value_init): GBoxed values initialize to NULL when g_value_init is called. GstCaps, which rolls its own type implementation, now does the same instead of allocating empty caps. (_gst_caps_initialize, _gst_caps_collect_value, _gst_caps_lcopy_value): Provide collect_value and lcopy_value type table methods. This allows G_VALUE_COLLECT to work. 2004-02-05 Andy Wingo <wingo@pobox.com> * configure.ac: * testsuite/Makefile.am (SUBDIRS): * testsuite/ghostpads/Makefile.am: * testsuite/ghostpads/ghostpads.c: A new test for ghost pads. * gst/gstpad.c (gst_pad_add_ghost_pad, gst_pad_remove_ghost_pad): These two routines are the only ones that set GST_GPAD_REALPAD(gpad), the ghost pad list, and the ghost pad's pad template. They should be made static, depending on ABI needs. (gst_real_pad_dispose): Handle the case of ghost pads without a parent. Assert after dealing with ghost pads that the ghost pad list is empty. (gst_ghost_pad_class_init): New property added, ::real-pad. Can be set after creation. (gst_ghost_pad_dispose): Set ::real-pad to NULL. (gst_ghost_pad_set_property, gst_ghost_pad_get_property): New functions. set_property will call add_ghost_pad/remove_ghost_pad as appropriate. (gst_ghost_pad_new): All the work is offloaded to g_object_new. * gst/gstelement.c (gst_element_add_pad): Handle ghost pads as well. (gst_element_add_ghost_pad): Remove code duplicated from _add_pad. (gst_element_remove_pad): Handle ghost pads as well. (gst_element_remove_ghost_pad): Deprecated (could be removed, depending on API-stability needs). 2004-02-05 Andy Wingo <wingo@pobox.com> * gst/gstbin.[ch]: (gst_bin_get_by_interface): GTypes are scalars, of course they're const
2004-02-11 13:26:04 +00:00
* Returns: an zero-terminated array of #GstQueryType, or NULL if none of the
* internally-linked pads has a query types function.
*/
const GstQueryType *
gst_pad_get_query_types_default (GstPad * pad)
{
GstQueryType *result = NULL;
gst/gstelement.c (gst_element_dispose): Protect against multiple invocations. Original commit message from CVS: 2004-02-24 Andy Wingo <wingo@pobox.com> * gst/gstelement.c (gst_element_dispose): Protect against multiple invocations. * gst/schedulers/gstoptimalscheduler.c I added a mess of prototypes at the top of the file by way of documentation. Some of the operations on chains and groups were re-organized. (create_group): Added a type argument so if the group is enabled, the setup_group_scheduler knows what to do. (group_elements): Added a type argument here, too, to be passed on to create_group. (group_element_set_enabled): If an unlinked PLAYING element is added to a bin, we have to create a new group to hold the element, and this function will be called before the group is added to the chain. Thus we have a valid case for group->chain==NULL. Instead of calling chain_group_set_enabled, just set the flag on the group (the chain's status will be set when the group is added to it). (gst_opt_scheduler_state_transition, chain_group_set_enabled): Setup the group scheduler when the group is enabled, not specifically when an element goes PAUSED->PLAYING. This means PLAYING elements can be added, linked, and scheduled into a PLAYING pipeline, as was intended. (add_to_group): Don't ref the group twice. I don't know when this double-ref got in here. Removing it has the potential to cause segfaults if other parts of the scheduler are buggy. If you find that the scheduler is segfaulting for you, put in an extra ref here and see if that hacks over the underlying issue. Of course, then find out what code is unreffing a group it doesn't own... (create_group): Make the extra refcount floating, and remove it after adding the element. This means that... (unref_group): Destroy when the refcount reaches 0, not 1, like every other refcounted object in the known universe. (remove_from_group): When a group becomes empty, set it to be not active, and remove it from its chain. Don't unref it again, there's no floating reference any more. (destroy_group): We have to remove the group from the chain in remove_from_group (rather than here) to break refcounting cycles (the chain always has a ref on the group). So assert that group->chain==NULL. (ref_group_by_count): Removed, it was commented out anyway. (merge_chains): Use the remove_from_chain and add_to_chain primitives to do the reparenting, instead of rolling our own implementation. (add_to_chain): The first non-disabled group in the chain's group list will be the entry point for the chain. Because buffers can accumulate in loop elements' peer bufpens, we preferentially schedule loop groups before get groups to avoid unnecessary execution of get-based groups when the bufpens are already full. (gst_opt_scheduler_schedule_run_queue): Debug fixes. (get_group_schedule_function): Ditto. (loop_group_schedule_function): Ditto. (gst_opt_scheduler_loop_wrapper): Ditto. (gst_opt_scheduler_iterate): Ditto. I understand the opt scheduler now, yippee! * gst/gstpad.c: All throughout, added FIXMEs to look at for 0.9. (gst_pad_get_name, gst_pad_set_chain_function) (gst_pad_set_get_function, gst_pad_set_event_function) (gst_pad_set_event_mask_function, gst_pad_get_event_masks) (gst_pad_get_event_masks_default, gst_pad_set_convert_function) (gst_pad_set_query_function, gst_pad_get_query_types) (gst_pad_get_query_types_default) (gst_pad_set_internal_link_function) (gst_pad_set_formats_function, gst_pad_set_link_function) (gst_pad_set_fixate_function, gst_pad_set_getcaps_function) (gst_pad_set_bufferalloc_function, gst_pad_unlink) (gst_pad_renegotiate, gst_pad_set_parent, gst_pad_get_parent) (gst_pad_add_ghost_pad, gst_pad_proxy_getcaps) (gst_pad_proxy_pad_link, gst_pad_proxy_fixate) (gst_pad_get_pad_template_caps, gst_pad_check_compatibility) (gst_pad_get_peer, gst_pad_get_allowed_caps) (gst_pad_alloc_buffer, gst_pad_push, gst_pad_pull) (gst_pad_selectv, gst_pad_select, gst_pad_template_get_caps) (gst_pad_event_default_dispatch, gst_pad_event_default) (gst_pad_dispatcher, gst_pad_send_event, gst_pad_convert_default) (gst_pad_convert, gst_pad_query_default, gst_pad_query) (gst_pad_get_formats_default, gst_pad_get_formats): Better argument checks, and some doc fixes. (gst_pad_custom_new_from_template): Um, does anyone use these functions? Actually make a custom pad instead of a normal one. (gst_pad_try_set_caps): Transpose some checks. (gst_pad_try_set_caps_nonfixed): Same, and use a macro to check if the pad is in negotiation. (gst_pad_try_relink_filtered): Use pad_link_prepare. * gst/gstelement.c: Remove prototypes also defined in gstclock.h. * gst/gstelement.h: * gst/gstclock.h: Un-deprecate the old clocking API, as discussed on the list.
2004-02-25 13:16:12 +00:00
g_return_val_if_fail (GST_IS_PAD (pad), NULL);
gst_pad_dispatcher (pad, (GstPadDispatcherFunction)
gst_pad_get_query_types_dispatcher, &result);
return result;
}
/**
* gst_pad_set_internal_link_function:
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
* @pad: a #GstPad of either direction.
* @intlink: the #GstPadIntLinkFunction to set.
*
* Sets the given internal link function for the pad.
*/
void
gst_pad_set_internal_link_function (GstPad * pad, GstPadIntLinkFunction intlink)
{
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
g_return_if_fail (GST_IS_PAD (pad));
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
GST_PAD_INTLINKFUNC (pad) = intlink;
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_DEBUG_OBJECT (GST_CAT_PADS, pad, "internal link set to %s",
GST_DEBUG_FUNCPTR_NAME (intlink));
}
/**
* gst_pad_set_link_function:
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
* @pad: a #GstPad.
* @link: the #GstPadLinkFunction to set.
*
* Sets the given link function for the pad. It will be called when
* the pad is linked with another pad.
*
* The return value #GST_PAD_LINK_OK should be used when the connection can be
* made.
*
* The return value #GST_PAD_LINK_REFUSED should be used when the connection
* cannot be made for some reason.
*
* If @link is installed on a source pad, it should call the #GstPadLinkFunction
* of the peer sink pad, if present.
*/
void
gst_pad_set_link_function (GstPad * pad, GstPadLinkFunction link)
{
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
g_return_if_fail (GST_IS_PAD (pad));
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
GST_PAD_LINKFUNC (pad) = link;
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_DEBUG_OBJECT (GST_CAT_PADS, pad, "linkfunc set to %s",
GST_DEBUG_FUNCPTR_NAME (link));
}
/**
* gst_pad_set_unlink_function:
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
* @pad: a #GstPad.
* @unlink: the #GstPadUnlinkFunction to set.
*
* Sets the given unlink function for the pad. It will be called
* when the pad is unlinked.
*/
void
gst_pad_set_unlink_function (GstPad * pad, GstPadUnlinkFunction unlink)
{
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
g_return_if_fail (GST_IS_PAD (pad));
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
GST_PAD_UNLINKFUNC (pad) = unlink;
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_DEBUG_OBJECT (GST_CAT_PADS, pad, "unlinkfunc set to %s",
GST_DEBUG_FUNCPTR_NAME (unlink));
}
/**
* gst_pad_set_getcaps_function:
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
* @pad: a #GstPad.
* @getcaps: the #GstPadGetCapsFunction to set.
*
gst/gstpad.c (gst_pad_custom_new): Add a FIXME, this is a hacky way to do inheritance. Original commit message from CVS: 2004-02-10 Andy Wingo <wingo@pobox.com> * gst/gstpad.c (gst_pad_custom_new): Add a FIXME, this is a hacky way to do inheritance. (gst_pad_get_event_masks, gst_pad_get_event_masks_default) (gst_pad_get_query_types, gst_pad_get_query_types_default): Routine docs. (gst_pad_set_link_function, gst_pad_set_fixate_function) (gst_pad_set_getcaps_function): Doc from Dave's negotation random doc. (gst_pad_unlink, gst_pad_is_linked): Docs. (gst_pad_renegotiate): A brief description of capsnego. (gst_pad_try_set_caps): Document. (gst_pad_try_set_caps_nonfixed): Document. (gst_pad_can_link_filtered, gst_pad_link_filtered): Doc fixes. (gst_pad_set_parent): Deprecated (although not out of the API). (gst_pad_get_parent): Deprecated, although many plugins use this. (gst_pad_add_ghost_pad, gst_pad_remove_ghost_pad): Doc that these are private and will go away in 0.9. (gst_pad_perform_negotiate): Doc. (gst_pad_link_unnegotiate): I think this is meant to be static. (gst_pad_get_negotiated_caps, gst_pad_get_pad_template_caps) (gst_pad_template_get_caps_by_name, gst_pad_check_compatibility) (gst_pad_get_peer): Doc updates. (gst_pad_caps_change_notify): Doc. (gst_pad_alloc_buffer, gst_pad_push, gst_static_pad_template_get) (gst_ghost_pad_new): Doc fixes. * gst/gstobject.c (gst_object_get_parent, gst_object_unparent) (gst_object_check_uniqueness): * gst/gstelement.c (gst_element_add_pad) (gst_element_add_ghost_pad, gst_element_remove_pad) (gst_element_remove_ghost_pad, gst_element_get_pad) (gst_element_get_static_pad, gst_element_get_pad_list) (gst_element_class_get_pad_template_list) (gst_element_class_get_pad_template): Work on the docs. (gst_element_get_pad_template_list): Uses the class method. (gst_element_get_compatible_pad_template): Docs, and consolidate some test conditions. (gst_element_get_pad_from_template): New static function. (gst_element_request_compatible_pad): Docs, and work with non-request compatible templates. (gst_element_get_compatible_pad_filtered): Docs and remove redundant checks. (gst_element_get_compatible_pad, gst_element_link_pads_filtered) (gst_element_link_filtered, gst_element_link_many) (gst_element_link, gst_element_link_pads) (gst_element_unlink_many): Docs. 2004-02-05 Andy Wingo <wingo@pobox.com> * gst/gstpad.c (_gst_real_pad_fixate_accumulator): s/pointer/boxed/. * gst/gstmarshal.list (VOID:BOXED, BOXED:BOXED): New marshallers. * gst/gstpad.c (gst_real_pad_class_init): Use a BOXED:BOXED marshaller for ::fixate, and VOID:BOXED for ::caps-nego-failed, with the type=GST_TYPE_CAPS. This allows language bindings to know what kind of data they're dealing with. * gst/gstcaps.c (_gst_caps_value_init): GBoxed values initialize to NULL when g_value_init is called. GstCaps, which rolls its own type implementation, now does the same instead of allocating empty caps. (_gst_caps_initialize, _gst_caps_collect_value, _gst_caps_lcopy_value): Provide collect_value and lcopy_value type table methods. This allows G_VALUE_COLLECT to work. 2004-02-05 Andy Wingo <wingo@pobox.com> * configure.ac: * testsuite/Makefile.am (SUBDIRS): * testsuite/ghostpads/Makefile.am: * testsuite/ghostpads/ghostpads.c: A new test for ghost pads. * gst/gstpad.c (gst_pad_add_ghost_pad, gst_pad_remove_ghost_pad): These two routines are the only ones that set GST_GPAD_REALPAD(gpad), the ghost pad list, and the ghost pad's pad template. They should be made static, depending on ABI needs. (gst_real_pad_dispose): Handle the case of ghost pads without a parent. Assert after dealing with ghost pads that the ghost pad list is empty. (gst_ghost_pad_class_init): New property added, ::real-pad. Can be set after creation. (gst_ghost_pad_dispose): Set ::real-pad to NULL. (gst_ghost_pad_set_property, gst_ghost_pad_get_property): New functions. set_property will call add_ghost_pad/remove_ghost_pad as appropriate. (gst_ghost_pad_new): All the work is offloaded to g_object_new. * gst/gstelement.c (gst_element_add_pad): Handle ghost pads as well. (gst_element_add_ghost_pad): Remove code duplicated from _add_pad. (gst_element_remove_pad): Handle ghost pads as well. (gst_element_remove_ghost_pad): Deprecated (could be removed, depending on API-stability needs). 2004-02-05 Andy Wingo <wingo@pobox.com> * gst/gstbin.[ch]: (gst_bin_get_by_interface): GTypes are scalars, of course they're const
2004-02-11 13:26:04 +00:00
* Sets the given getcaps function for the pad. @getcaps should return the
* allowable caps for a pad in the context of the element's state, its link to
* other elements, and the devices or files it has opened. These caps must be a
* subset of the pad template caps. In the NULL state with no links, @getcaps
* should ideally return the same caps as the pad template. In rare
* circumstances, an object property can affect the caps returned by @getcaps,
* but this is discouraged.
*
* You do not need to call this function if @pad's allowed caps are always the
* same as the pad template caps. This can only be true if the padtemplate
* has fixed simple caps.
gst/gstpad.c (gst_pad_custom_new): Add a FIXME, this is a hacky way to do inheritance. Original commit message from CVS: 2004-02-10 Andy Wingo <wingo@pobox.com> * gst/gstpad.c (gst_pad_custom_new): Add a FIXME, this is a hacky way to do inheritance. (gst_pad_get_event_masks, gst_pad_get_event_masks_default) (gst_pad_get_query_types, gst_pad_get_query_types_default): Routine docs. (gst_pad_set_link_function, gst_pad_set_fixate_function) (gst_pad_set_getcaps_function): Doc from Dave's negotation random doc. (gst_pad_unlink, gst_pad_is_linked): Docs. (gst_pad_renegotiate): A brief description of capsnego. (gst_pad_try_set_caps): Document. (gst_pad_try_set_caps_nonfixed): Document. (gst_pad_can_link_filtered, gst_pad_link_filtered): Doc fixes. (gst_pad_set_parent): Deprecated (although not out of the API). (gst_pad_get_parent): Deprecated, although many plugins use this. (gst_pad_add_ghost_pad, gst_pad_remove_ghost_pad): Doc that these are private and will go away in 0.9. (gst_pad_perform_negotiate): Doc. (gst_pad_link_unnegotiate): I think this is meant to be static. (gst_pad_get_negotiated_caps, gst_pad_get_pad_template_caps) (gst_pad_template_get_caps_by_name, gst_pad_check_compatibility) (gst_pad_get_peer): Doc updates. (gst_pad_caps_change_notify): Doc. (gst_pad_alloc_buffer, gst_pad_push, gst_static_pad_template_get) (gst_ghost_pad_new): Doc fixes. * gst/gstobject.c (gst_object_get_parent, gst_object_unparent) (gst_object_check_uniqueness): * gst/gstelement.c (gst_element_add_pad) (gst_element_add_ghost_pad, gst_element_remove_pad) (gst_element_remove_ghost_pad, gst_element_get_pad) (gst_element_get_static_pad, gst_element_get_pad_list) (gst_element_class_get_pad_template_list) (gst_element_class_get_pad_template): Work on the docs. (gst_element_get_pad_template_list): Uses the class method. (gst_element_get_compatible_pad_template): Docs, and consolidate some test conditions. (gst_element_get_pad_from_template): New static function. (gst_element_request_compatible_pad): Docs, and work with non-request compatible templates. (gst_element_get_compatible_pad_filtered): Docs and remove redundant checks. (gst_element_get_compatible_pad, gst_element_link_pads_filtered) (gst_element_link_filtered, gst_element_link_many) (gst_element_link, gst_element_link_pads) (gst_element_unlink_many): Docs. 2004-02-05 Andy Wingo <wingo@pobox.com> * gst/gstpad.c (_gst_real_pad_fixate_accumulator): s/pointer/boxed/. * gst/gstmarshal.list (VOID:BOXED, BOXED:BOXED): New marshallers. * gst/gstpad.c (gst_real_pad_class_init): Use a BOXED:BOXED marshaller for ::fixate, and VOID:BOXED for ::caps-nego-failed, with the type=GST_TYPE_CAPS. This allows language bindings to know what kind of data they're dealing with. * gst/gstcaps.c (_gst_caps_value_init): GBoxed values initialize to NULL when g_value_init is called. GstCaps, which rolls its own type implementation, now does the same instead of allocating empty caps. (_gst_caps_initialize, _gst_caps_collect_value, _gst_caps_lcopy_value): Provide collect_value and lcopy_value type table methods. This allows G_VALUE_COLLECT to work. 2004-02-05 Andy Wingo <wingo@pobox.com> * configure.ac: * testsuite/Makefile.am (SUBDIRS): * testsuite/ghostpads/Makefile.am: * testsuite/ghostpads/ghostpads.c: A new test for ghost pads. * gst/gstpad.c (gst_pad_add_ghost_pad, gst_pad_remove_ghost_pad): These two routines are the only ones that set GST_GPAD_REALPAD(gpad), the ghost pad list, and the ghost pad's pad template. They should be made static, depending on ABI needs. (gst_real_pad_dispose): Handle the case of ghost pads without a parent. Assert after dealing with ghost pads that the ghost pad list is empty. (gst_ghost_pad_class_init): New property added, ::real-pad. Can be set after creation. (gst_ghost_pad_dispose): Set ::real-pad to NULL. (gst_ghost_pad_set_property, gst_ghost_pad_get_property): New functions. set_property will call add_ghost_pad/remove_ghost_pad as appropriate. (gst_ghost_pad_new): All the work is offloaded to g_object_new. * gst/gstelement.c (gst_element_add_pad): Handle ghost pads as well. (gst_element_add_ghost_pad): Remove code duplicated from _add_pad. (gst_element_remove_pad): Handle ghost pads as well. (gst_element_remove_ghost_pad): Deprecated (could be removed, depending on API-stability needs). 2004-02-05 Andy Wingo <wingo@pobox.com> * gst/gstbin.[ch]: (gst_bin_get_by_interface): GTypes are scalars, of course they're const
2004-02-11 13:26:04 +00:00
*
* For most filters, the caps returned by @getcaps is directly affected by the
* allowed caps on other pads. For demuxers and decoders, the caps returned by
* the srcpad's getcaps function is directly related to the stream data. Again,
* @getcaps should return the most specific caps it reasonably can, since this
* helps with autoplugging.
gst/gstpad.c (gst_pad_custom_new): Add a FIXME, this is a hacky way to do inheritance. Original commit message from CVS: 2004-02-10 Andy Wingo <wingo@pobox.com> * gst/gstpad.c (gst_pad_custom_new): Add a FIXME, this is a hacky way to do inheritance. (gst_pad_get_event_masks, gst_pad_get_event_masks_default) (gst_pad_get_query_types, gst_pad_get_query_types_default): Routine docs. (gst_pad_set_link_function, gst_pad_set_fixate_function) (gst_pad_set_getcaps_function): Doc from Dave's negotation random doc. (gst_pad_unlink, gst_pad_is_linked): Docs. (gst_pad_renegotiate): A brief description of capsnego. (gst_pad_try_set_caps): Document. (gst_pad_try_set_caps_nonfixed): Document. (gst_pad_can_link_filtered, gst_pad_link_filtered): Doc fixes. (gst_pad_set_parent): Deprecated (although not out of the API). (gst_pad_get_parent): Deprecated, although many plugins use this. (gst_pad_add_ghost_pad, gst_pad_remove_ghost_pad): Doc that these are private and will go away in 0.9. (gst_pad_perform_negotiate): Doc. (gst_pad_link_unnegotiate): I think this is meant to be static. (gst_pad_get_negotiated_caps, gst_pad_get_pad_template_caps) (gst_pad_template_get_caps_by_name, gst_pad_check_compatibility) (gst_pad_get_peer): Doc updates. (gst_pad_caps_change_notify): Doc. (gst_pad_alloc_buffer, gst_pad_push, gst_static_pad_template_get) (gst_ghost_pad_new): Doc fixes. * gst/gstobject.c (gst_object_get_parent, gst_object_unparent) (gst_object_check_uniqueness): * gst/gstelement.c (gst_element_add_pad) (gst_element_add_ghost_pad, gst_element_remove_pad) (gst_element_remove_ghost_pad, gst_element_get_pad) (gst_element_get_static_pad, gst_element_get_pad_list) (gst_element_class_get_pad_template_list) (gst_element_class_get_pad_template): Work on the docs. (gst_element_get_pad_template_list): Uses the class method. (gst_element_get_compatible_pad_template): Docs, and consolidate some test conditions. (gst_element_get_pad_from_template): New static function. (gst_element_request_compatible_pad): Docs, and work with non-request compatible templates. (gst_element_get_compatible_pad_filtered): Docs and remove redundant checks. (gst_element_get_compatible_pad, gst_element_link_pads_filtered) (gst_element_link_filtered, gst_element_link_many) (gst_element_link, gst_element_link_pads) (gst_element_unlink_many): Docs. 2004-02-05 Andy Wingo <wingo@pobox.com> * gst/gstpad.c (_gst_real_pad_fixate_accumulator): s/pointer/boxed/. * gst/gstmarshal.list (VOID:BOXED, BOXED:BOXED): New marshallers. * gst/gstpad.c (gst_real_pad_class_init): Use a BOXED:BOXED marshaller for ::fixate, and VOID:BOXED for ::caps-nego-failed, with the type=GST_TYPE_CAPS. This allows language bindings to know what kind of data they're dealing with. * gst/gstcaps.c (_gst_caps_value_init): GBoxed values initialize to NULL when g_value_init is called. GstCaps, which rolls its own type implementation, now does the same instead of allocating empty caps. (_gst_caps_initialize, _gst_caps_collect_value, _gst_caps_lcopy_value): Provide collect_value and lcopy_value type table methods. This allows G_VALUE_COLLECT to work. 2004-02-05 Andy Wingo <wingo@pobox.com> * configure.ac: * testsuite/Makefile.am (SUBDIRS): * testsuite/ghostpads/Makefile.am: * testsuite/ghostpads/ghostpads.c: A new test for ghost pads. * gst/gstpad.c (gst_pad_add_ghost_pad, gst_pad_remove_ghost_pad): These two routines are the only ones that set GST_GPAD_REALPAD(gpad), the ghost pad list, and the ghost pad's pad template. They should be made static, depending on ABI needs. (gst_real_pad_dispose): Handle the case of ghost pads without a parent. Assert after dealing with ghost pads that the ghost pad list is empty. (gst_ghost_pad_class_init): New property added, ::real-pad. Can be set after creation. (gst_ghost_pad_dispose): Set ::real-pad to NULL. (gst_ghost_pad_set_property, gst_ghost_pad_get_property): New functions. set_property will call add_ghost_pad/remove_ghost_pad as appropriate. (gst_ghost_pad_new): All the work is offloaded to g_object_new. * gst/gstelement.c (gst_element_add_pad): Handle ghost pads as well. (gst_element_add_ghost_pad): Remove code duplicated from _add_pad. (gst_element_remove_pad): Handle ghost pads as well. (gst_element_remove_ghost_pad): Deprecated (could be removed, depending on API-stability needs). 2004-02-05 Andy Wingo <wingo@pobox.com> * gst/gstbin.[ch]: (gst_bin_get_by_interface): GTypes are scalars, of course they're const
2004-02-11 13:26:04 +00:00
*
* Note that the return value from @getcaps is owned by the caller, so the caller
* should unref the caps after usage.
*/
void
gst_pad_set_getcaps_function (GstPad * pad, GstPadGetCapsFunction getcaps)
{
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
g_return_if_fail (GST_IS_PAD (pad));
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
GST_PAD_GETCAPSFUNC (pad) = getcaps;
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_DEBUG_OBJECT (GST_CAT_PADS, pad, "getcapsfunc set to %s",
GST_DEBUG_FUNCPTR_NAME (getcaps));
}
/**
* gst_pad_set_acceptcaps_function:
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
* @pad: a #GstPad.
* @acceptcaps: the #GstPadAcceptCapsFunction to set.
*
* Sets the given acceptcaps function for the pad. The acceptcaps function
* will be called to check if the pad can accept the given caps. Setting the
* acceptcaps function to NULL restores the default behaviour of allowing
* any caps that matches the caps from gst_pad_get_caps.
*/
void
gst_pad_set_acceptcaps_function (GstPad * pad,
GstPadAcceptCapsFunction acceptcaps)
{
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
g_return_if_fail (GST_IS_PAD (pad));
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
GST_PAD_ACCEPTCAPSFUNC (pad) = acceptcaps;
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_DEBUG_OBJECT (GST_CAT_PADS, pad, "acceptcapsfunc set to %s",
GST_DEBUG_FUNCPTR_NAME (acceptcaps));
}
/**
* gst_pad_set_fixatecaps_function:
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
* @pad: a #GstPad.
* @fixatecaps: the #GstPadFixateCapsFunction to set.
*
* Sets the given fixatecaps function for the pad. The fixatecaps function
* will be called whenever the default values for a GstCaps needs to be
* filled in.
*/
void
gst_pad_set_fixatecaps_function (GstPad * pad,
GstPadFixateCapsFunction fixatecaps)
{
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
g_return_if_fail (GST_IS_PAD (pad));
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
GST_PAD_FIXATECAPSFUNC (pad) = fixatecaps;
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_DEBUG_OBJECT (GST_CAT_PADS, pad, "fixatecapsfunc set to %s",
GST_DEBUG_FUNCPTR_NAME (fixatecaps));
}
/**
* gst_pad_set_setcaps_function:
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
* @pad: a #GstPad.
* @setcaps: the #GstPadSetCapsFunction to set.
*
* Sets the given setcaps function for the pad. The setcaps function
* will be called whenever a buffer with a new media type is pushed or
* pulled from the pad. The pad/element needs to update its internal
* structures to process the new media type. If this new type is not
* acceptable, the setcaps function should return FALSE.
*/
void
gst_pad_set_setcaps_function (GstPad * pad, GstPadSetCapsFunction setcaps)
{
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
g_return_if_fail (GST_IS_PAD (pad));
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
GST_PAD_SETCAPSFUNC (pad) = setcaps;
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_DEBUG_OBJECT (GST_CAT_PADS, pad, "setcapsfunc set to %s",
GST_DEBUG_FUNCPTR_NAME (setcaps));
}
/**
* gst_pad_set_bufferalloc_function:
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
* @pad: a sink #GstPad.
* @bufalloc: the #GstPadBufferAllocFunction to set.
*
* Sets the given bufferalloc function for the pad. Note that the
* bufferalloc function can only be set on sinkpads.
*/
void
gst_pad_set_bufferalloc_function (GstPad * pad,
GstPadBufferAllocFunction bufalloc)
{
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
g_return_if_fail (GST_IS_PAD (pad));
g_return_if_fail (GST_PAD_IS_SINK (pad));
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
GST_PAD_BUFFERALLOCFUNC (pad) = bufalloc;
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_DEBUG_OBJECT (GST_CAT_PADS, pad, "bufferallocfunc set to %s",
GST_DEBUG_FUNCPTR_NAME (bufalloc));
}
/**
* gst_pad_unlink:
* @srcpad: the source #GstPad to unlink.
* @sinkpad: the sink #GstPad to unlink.
*
gst/gstpad.c (gst_pad_custom_new): Add a FIXME, this is a hacky way to do inheritance. Original commit message from CVS: 2004-02-10 Andy Wingo <wingo@pobox.com> * gst/gstpad.c (gst_pad_custom_new): Add a FIXME, this is a hacky way to do inheritance. (gst_pad_get_event_masks, gst_pad_get_event_masks_default) (gst_pad_get_query_types, gst_pad_get_query_types_default): Routine docs. (gst_pad_set_link_function, gst_pad_set_fixate_function) (gst_pad_set_getcaps_function): Doc from Dave's negotation random doc. (gst_pad_unlink, gst_pad_is_linked): Docs. (gst_pad_renegotiate): A brief description of capsnego. (gst_pad_try_set_caps): Document. (gst_pad_try_set_caps_nonfixed): Document. (gst_pad_can_link_filtered, gst_pad_link_filtered): Doc fixes. (gst_pad_set_parent): Deprecated (although not out of the API). (gst_pad_get_parent): Deprecated, although many plugins use this. (gst_pad_add_ghost_pad, gst_pad_remove_ghost_pad): Doc that these are private and will go away in 0.9. (gst_pad_perform_negotiate): Doc. (gst_pad_link_unnegotiate): I think this is meant to be static. (gst_pad_get_negotiated_caps, gst_pad_get_pad_template_caps) (gst_pad_template_get_caps_by_name, gst_pad_check_compatibility) (gst_pad_get_peer): Doc updates. (gst_pad_caps_change_notify): Doc. (gst_pad_alloc_buffer, gst_pad_push, gst_static_pad_template_get) (gst_ghost_pad_new): Doc fixes. * gst/gstobject.c (gst_object_get_parent, gst_object_unparent) (gst_object_check_uniqueness): * gst/gstelement.c (gst_element_add_pad) (gst_element_add_ghost_pad, gst_element_remove_pad) (gst_element_remove_ghost_pad, gst_element_get_pad) (gst_element_get_static_pad, gst_element_get_pad_list) (gst_element_class_get_pad_template_list) (gst_element_class_get_pad_template): Work on the docs. (gst_element_get_pad_template_list): Uses the class method. (gst_element_get_compatible_pad_template): Docs, and consolidate some test conditions. (gst_element_get_pad_from_template): New static function. (gst_element_request_compatible_pad): Docs, and work with non-request compatible templates. (gst_element_get_compatible_pad_filtered): Docs and remove redundant checks. (gst_element_get_compatible_pad, gst_element_link_pads_filtered) (gst_element_link_filtered, gst_element_link_many) (gst_element_link, gst_element_link_pads) (gst_element_unlink_many): Docs. 2004-02-05 Andy Wingo <wingo@pobox.com> * gst/gstpad.c (_gst_real_pad_fixate_accumulator): s/pointer/boxed/. * gst/gstmarshal.list (VOID:BOXED, BOXED:BOXED): New marshallers. * gst/gstpad.c (gst_real_pad_class_init): Use a BOXED:BOXED marshaller for ::fixate, and VOID:BOXED for ::caps-nego-failed, with the type=GST_TYPE_CAPS. This allows language bindings to know what kind of data they're dealing with. * gst/gstcaps.c (_gst_caps_value_init): GBoxed values initialize to NULL when g_value_init is called. GstCaps, which rolls its own type implementation, now does the same instead of allocating empty caps. (_gst_caps_initialize, _gst_caps_collect_value, _gst_caps_lcopy_value): Provide collect_value and lcopy_value type table methods. This allows G_VALUE_COLLECT to work. 2004-02-05 Andy Wingo <wingo@pobox.com> * configure.ac: * testsuite/Makefile.am (SUBDIRS): * testsuite/ghostpads/Makefile.am: * testsuite/ghostpads/ghostpads.c: A new test for ghost pads. * gst/gstpad.c (gst_pad_add_ghost_pad, gst_pad_remove_ghost_pad): These two routines are the only ones that set GST_GPAD_REALPAD(gpad), the ghost pad list, and the ghost pad's pad template. They should be made static, depending on ABI needs. (gst_real_pad_dispose): Handle the case of ghost pads without a parent. Assert after dealing with ghost pads that the ghost pad list is empty. (gst_ghost_pad_class_init): New property added, ::real-pad. Can be set after creation. (gst_ghost_pad_dispose): Set ::real-pad to NULL. (gst_ghost_pad_set_property, gst_ghost_pad_get_property): New functions. set_property will call add_ghost_pad/remove_ghost_pad as appropriate. (gst_ghost_pad_new): All the work is offloaded to g_object_new. * gst/gstelement.c (gst_element_add_pad): Handle ghost pads as well. (gst_element_add_ghost_pad): Remove code duplicated from _add_pad. (gst_element_remove_pad): Handle ghost pads as well. (gst_element_remove_ghost_pad): Deprecated (could be removed, depending on API-stability needs). 2004-02-05 Andy Wingo <wingo@pobox.com> * gst/gstbin.[ch]: (gst_bin_get_by_interface): GTypes are scalars, of course they're const
2004-02-11 13:26:04 +00:00
* Unlinks the source pad from the sink pad. Will emit the "unlinked" signal on
* both pads.
*
* Returns: TRUE if the pads were unlinked. This function returns FALSE if
* the pads were not linked together.
*
* MT safe.
*/
gboolean
gst_pad_unlink (GstPad * srcpad, GstPad * sinkpad)
{
g_return_val_if_fail (GST_IS_PAD (srcpad), FALSE);
g_return_val_if_fail (GST_IS_PAD (sinkpad), FALSE);
GST_CAT_INFO (GST_CAT_ELEMENT_PADS, "unlinking %s:%s(%p) and %s:%s(%p)",
GST_DEBUG_PAD_NAME (srcpad), srcpad,
GST_DEBUG_PAD_NAME (sinkpad), sinkpad);
GST_OBJECT_LOCK (srcpad);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
if (G_UNLIKELY (GST_PAD_DIRECTION (srcpad) != GST_PAD_SRC))
goto not_srcpad;
GST_OBJECT_LOCK (sinkpad);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
if (G_UNLIKELY (GST_PAD_DIRECTION (sinkpad) != GST_PAD_SINK))
goto not_sinkpad;
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
if (G_UNLIKELY (GST_PAD_PEER (srcpad) != sinkpad))
goto not_linked_together;
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
if (GST_PAD_UNLINKFUNC (srcpad)) {
GST_PAD_UNLINKFUNC (srcpad) (srcpad);
}
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
if (GST_PAD_UNLINKFUNC (sinkpad)) {
GST_PAD_UNLINKFUNC (sinkpad) (sinkpad);
}
/* first clear peers */
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
GST_PAD_PEER (srcpad) = NULL;
GST_PAD_PEER (sinkpad) = NULL;
GST_OBJECT_UNLOCK (sinkpad);
GST_OBJECT_UNLOCK (srcpad);
/* fire off a signal to each of the pads telling them
* that they've been unlinked */
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
g_signal_emit (G_OBJECT (srcpad), gst_pad_signals[PAD_UNLINKED], 0, sinkpad);
g_signal_emit (G_OBJECT (sinkpad), gst_pad_signals[PAD_UNLINKED], 0, srcpad);
GST_CAT_INFO (GST_CAT_ELEMENT_PADS, "unlinked %s:%s and %s:%s",
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
GST_DEBUG_PAD_NAME (srcpad), GST_DEBUG_PAD_NAME (sinkpad));
return TRUE;
not_srcpad:
{
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
g_critical ("pad %s is not a source pad", GST_PAD_NAME (srcpad));
GST_OBJECT_UNLOCK (srcpad);
return FALSE;
}
not_sinkpad:
{
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
g_critical ("pad %s is not a sink pad", GST_PAD_NAME (sinkpad));
GST_OBJECT_UNLOCK (sinkpad);
GST_OBJECT_UNLOCK (srcpad);
return FALSE;
}
not_linked_together:
{
/* we do not emit a warning in this case because unlinking cannot
* be made MT safe.*/
GST_OBJECT_UNLOCK (sinkpad);
GST_OBJECT_UNLOCK (srcpad);
return FALSE;
}
}
/**
* gst_pad_is_linked:
* @pad: pad to check
*
gst/gstpad.c (gst_pad_custom_new): Add a FIXME, this is a hacky way to do inheritance. Original commit message from CVS: 2004-02-10 Andy Wingo <wingo@pobox.com> * gst/gstpad.c (gst_pad_custom_new): Add a FIXME, this is a hacky way to do inheritance. (gst_pad_get_event_masks, gst_pad_get_event_masks_default) (gst_pad_get_query_types, gst_pad_get_query_types_default): Routine docs. (gst_pad_set_link_function, gst_pad_set_fixate_function) (gst_pad_set_getcaps_function): Doc from Dave's negotation random doc. (gst_pad_unlink, gst_pad_is_linked): Docs. (gst_pad_renegotiate): A brief description of capsnego. (gst_pad_try_set_caps): Document. (gst_pad_try_set_caps_nonfixed): Document. (gst_pad_can_link_filtered, gst_pad_link_filtered): Doc fixes. (gst_pad_set_parent): Deprecated (although not out of the API). (gst_pad_get_parent): Deprecated, although many plugins use this. (gst_pad_add_ghost_pad, gst_pad_remove_ghost_pad): Doc that these are private and will go away in 0.9. (gst_pad_perform_negotiate): Doc. (gst_pad_link_unnegotiate): I think this is meant to be static. (gst_pad_get_negotiated_caps, gst_pad_get_pad_template_caps) (gst_pad_template_get_caps_by_name, gst_pad_check_compatibility) (gst_pad_get_peer): Doc updates. (gst_pad_caps_change_notify): Doc. (gst_pad_alloc_buffer, gst_pad_push, gst_static_pad_template_get) (gst_ghost_pad_new): Doc fixes. * gst/gstobject.c (gst_object_get_parent, gst_object_unparent) (gst_object_check_uniqueness): * gst/gstelement.c (gst_element_add_pad) (gst_element_add_ghost_pad, gst_element_remove_pad) (gst_element_remove_ghost_pad, gst_element_get_pad) (gst_element_get_static_pad, gst_element_get_pad_list) (gst_element_class_get_pad_template_list) (gst_element_class_get_pad_template): Work on the docs. (gst_element_get_pad_template_list): Uses the class method. (gst_element_get_compatible_pad_template): Docs, and consolidate some test conditions. (gst_element_get_pad_from_template): New static function. (gst_element_request_compatible_pad): Docs, and work with non-request compatible templates. (gst_element_get_compatible_pad_filtered): Docs and remove redundant checks. (gst_element_get_compatible_pad, gst_element_link_pads_filtered) (gst_element_link_filtered, gst_element_link_many) (gst_element_link, gst_element_link_pads) (gst_element_unlink_many): Docs. 2004-02-05 Andy Wingo <wingo@pobox.com> * gst/gstpad.c (_gst_real_pad_fixate_accumulator): s/pointer/boxed/. * gst/gstmarshal.list (VOID:BOXED, BOXED:BOXED): New marshallers. * gst/gstpad.c (gst_real_pad_class_init): Use a BOXED:BOXED marshaller for ::fixate, and VOID:BOXED for ::caps-nego-failed, with the type=GST_TYPE_CAPS. This allows language bindings to know what kind of data they're dealing with. * gst/gstcaps.c (_gst_caps_value_init): GBoxed values initialize to NULL when g_value_init is called. GstCaps, which rolls its own type implementation, now does the same instead of allocating empty caps. (_gst_caps_initialize, _gst_caps_collect_value, _gst_caps_lcopy_value): Provide collect_value and lcopy_value type table methods. This allows G_VALUE_COLLECT to work. 2004-02-05 Andy Wingo <wingo@pobox.com> * configure.ac: * testsuite/Makefile.am (SUBDIRS): * testsuite/ghostpads/Makefile.am: * testsuite/ghostpads/ghostpads.c: A new test for ghost pads. * gst/gstpad.c (gst_pad_add_ghost_pad, gst_pad_remove_ghost_pad): These two routines are the only ones that set GST_GPAD_REALPAD(gpad), the ghost pad list, and the ghost pad's pad template. They should be made static, depending on ABI needs. (gst_real_pad_dispose): Handle the case of ghost pads without a parent. Assert after dealing with ghost pads that the ghost pad list is empty. (gst_ghost_pad_class_init): New property added, ::real-pad. Can be set after creation. (gst_ghost_pad_dispose): Set ::real-pad to NULL. (gst_ghost_pad_set_property, gst_ghost_pad_get_property): New functions. set_property will call add_ghost_pad/remove_ghost_pad as appropriate. (gst_ghost_pad_new): All the work is offloaded to g_object_new. * gst/gstelement.c (gst_element_add_pad): Handle ghost pads as well. (gst_element_add_ghost_pad): Remove code duplicated from _add_pad. (gst_element_remove_pad): Handle ghost pads as well. (gst_element_remove_ghost_pad): Deprecated (could be removed, depending on API-stability needs). 2004-02-05 Andy Wingo <wingo@pobox.com> * gst/gstbin.[ch]: (gst_bin_get_by_interface): GTypes are scalars, of course they're const
2004-02-11 13:26:04 +00:00
* Checks if a @pad is linked to another pad or not.
*
gst/gstpad.c (gst_pad_custom_new): Add a FIXME, this is a hacky way to do inheritance. Original commit message from CVS: 2004-02-10 Andy Wingo <wingo@pobox.com> * gst/gstpad.c (gst_pad_custom_new): Add a FIXME, this is a hacky way to do inheritance. (gst_pad_get_event_masks, gst_pad_get_event_masks_default) (gst_pad_get_query_types, gst_pad_get_query_types_default): Routine docs. (gst_pad_set_link_function, gst_pad_set_fixate_function) (gst_pad_set_getcaps_function): Doc from Dave's negotation random doc. (gst_pad_unlink, gst_pad_is_linked): Docs. (gst_pad_renegotiate): A brief description of capsnego. (gst_pad_try_set_caps): Document. (gst_pad_try_set_caps_nonfixed): Document. (gst_pad_can_link_filtered, gst_pad_link_filtered): Doc fixes. (gst_pad_set_parent): Deprecated (although not out of the API). (gst_pad_get_parent): Deprecated, although many plugins use this. (gst_pad_add_ghost_pad, gst_pad_remove_ghost_pad): Doc that these are private and will go away in 0.9. (gst_pad_perform_negotiate): Doc. (gst_pad_link_unnegotiate): I think this is meant to be static. (gst_pad_get_negotiated_caps, gst_pad_get_pad_template_caps) (gst_pad_template_get_caps_by_name, gst_pad_check_compatibility) (gst_pad_get_peer): Doc updates. (gst_pad_caps_change_notify): Doc. (gst_pad_alloc_buffer, gst_pad_push, gst_static_pad_template_get) (gst_ghost_pad_new): Doc fixes. * gst/gstobject.c (gst_object_get_parent, gst_object_unparent) (gst_object_check_uniqueness): * gst/gstelement.c (gst_element_add_pad) (gst_element_add_ghost_pad, gst_element_remove_pad) (gst_element_remove_ghost_pad, gst_element_get_pad) (gst_element_get_static_pad, gst_element_get_pad_list) (gst_element_class_get_pad_template_list) (gst_element_class_get_pad_template): Work on the docs. (gst_element_get_pad_template_list): Uses the class method. (gst_element_get_compatible_pad_template): Docs, and consolidate some test conditions. (gst_element_get_pad_from_template): New static function. (gst_element_request_compatible_pad): Docs, and work with non-request compatible templates. (gst_element_get_compatible_pad_filtered): Docs and remove redundant checks. (gst_element_get_compatible_pad, gst_element_link_pads_filtered) (gst_element_link_filtered, gst_element_link_many) (gst_element_link, gst_element_link_pads) (gst_element_unlink_many): Docs. 2004-02-05 Andy Wingo <wingo@pobox.com> * gst/gstpad.c (_gst_real_pad_fixate_accumulator): s/pointer/boxed/. * gst/gstmarshal.list (VOID:BOXED, BOXED:BOXED): New marshallers. * gst/gstpad.c (gst_real_pad_class_init): Use a BOXED:BOXED marshaller for ::fixate, and VOID:BOXED for ::caps-nego-failed, with the type=GST_TYPE_CAPS. This allows language bindings to know what kind of data they're dealing with. * gst/gstcaps.c (_gst_caps_value_init): GBoxed values initialize to NULL when g_value_init is called. GstCaps, which rolls its own type implementation, now does the same instead of allocating empty caps. (_gst_caps_initialize, _gst_caps_collect_value, _gst_caps_lcopy_value): Provide collect_value and lcopy_value type table methods. This allows G_VALUE_COLLECT to work. 2004-02-05 Andy Wingo <wingo@pobox.com> * configure.ac: * testsuite/Makefile.am (SUBDIRS): * testsuite/ghostpads/Makefile.am: * testsuite/ghostpads/ghostpads.c: A new test for ghost pads. * gst/gstpad.c (gst_pad_add_ghost_pad, gst_pad_remove_ghost_pad): These two routines are the only ones that set GST_GPAD_REALPAD(gpad), the ghost pad list, and the ghost pad's pad template. They should be made static, depending on ABI needs. (gst_real_pad_dispose): Handle the case of ghost pads without a parent. Assert after dealing with ghost pads that the ghost pad list is empty. (gst_ghost_pad_class_init): New property added, ::real-pad. Can be set after creation. (gst_ghost_pad_dispose): Set ::real-pad to NULL. (gst_ghost_pad_set_property, gst_ghost_pad_get_property): New functions. set_property will call add_ghost_pad/remove_ghost_pad as appropriate. (gst_ghost_pad_new): All the work is offloaded to g_object_new. * gst/gstelement.c (gst_element_add_pad): Handle ghost pads as well. (gst_element_add_ghost_pad): Remove code duplicated from _add_pad. (gst_element_remove_pad): Handle ghost pads as well. (gst_element_remove_ghost_pad): Deprecated (could be removed, depending on API-stability needs). 2004-02-05 Andy Wingo <wingo@pobox.com> * gst/gstbin.[ch]: (gst_bin_get_by_interface): GTypes are scalars, of course they're const
2004-02-11 13:26:04 +00:00
* Returns: TRUE if the pad is linked, FALSE otherwise.
*
* MT safe.
*/
gboolean
gst_pad_is_linked (GstPad * pad)
{
gboolean result;
g_return_val_if_fail (GST_IS_PAD (pad), FALSE);
GST_OBJECT_LOCK (pad);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
result = (GST_PAD_PEER (pad) != NULL);
GST_OBJECT_UNLOCK (pad);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
return result;
}
/* get the caps from both pads and see if the intersection
* is not empty.
*
* This function should be called with the pad LOCK on both
* pads
*/
static gboolean
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
gst_pad_link_check_compatible_unlocked (GstPad * src, GstPad * sink)
{
GstCaps *srccaps;
GstCaps *sinkcaps;
GstCaps *icaps;
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
srccaps = gst_pad_get_caps_unlocked (src);
sinkcaps = gst_pad_get_caps_unlocked (sink);
GST_CAT_DEBUG (GST_CAT_CAPS, "src caps %" GST_PTR_FORMAT, srccaps);
GST_CAT_DEBUG (GST_CAT_CAPS, "sink caps %" GST_PTR_FORMAT, sinkcaps);
/* if we have caps on both pads we can check the intersection. If one
* of the caps is NULL, we return TRUE. */
if (srccaps == NULL || sinkcaps == NULL)
goto done;
icaps = gst_caps_intersect (srccaps, sinkcaps);
gst_caps_unref (srccaps);
gst_caps_unref (sinkcaps);
if (icaps == NULL)
goto was_null;
GST_CAT_DEBUG (GST_CAT_CAPS,
"intersection caps %p %" GST_PTR_FORMAT, icaps, icaps);
if (gst_caps_is_empty (icaps))
goto was_empty;
gst_caps_unref (icaps);
done:
return TRUE;
/* incompatible cases */
was_null:
{
GST_CAT_DEBUG (GST_CAT_CAPS, "intersection gave NULL");
return FALSE;
}
was_empty:
{
GST_CAT_DEBUG (GST_CAT_CAPS, "intersection is EMPTY");
gst_caps_unref (icaps);
return FALSE;
}
}
/* check if the grandparents of both pads are the same.
* This check is required so that we don't try to link
* pads from elements in different bins without ghostpads.
*
* The LOCK should be held on both pads
*/
static gboolean
gst_pad_link_check_hierarchy (GstPad * src, GstPad * sink)
{
GstObject *psrc, *psink;
psrc = GST_OBJECT_PARENT (src);
psink = GST_OBJECT_PARENT (sink);
/* if one of the pads has no parent, we allow the link */
if (G_UNLIKELY (psrc == NULL || psink == NULL))
goto no_parent;
/* if the parents are the same, we have a loop */
if (G_UNLIKELY (psrc == psink))
goto same_parents;
/* if they both have a parent, we check the grandparents. We can not lock
* the parent because we hold on the child (pad) and the locking order is
* parent >> child. */
psrc = GST_OBJECT_PARENT (psrc);
psink = GST_OBJECT_PARENT (psink);
/* if they have grandparents but they are not the same */
if (G_UNLIKELY (psrc != psink))
goto wrong_grandparents;
return TRUE;
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
/* ERRORS */
no_parent:
{
GST_CAT_DEBUG (GST_CAT_CAPS,
"one of the pads has no parent %" GST_PTR_FORMAT " and %"
GST_PTR_FORMAT, psrc, psink);
return TRUE;
}
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
same_parents:
{
GST_CAT_DEBUG (GST_CAT_CAPS, "pads have same parent %" GST_PTR_FORMAT,
psrc);
return FALSE;
}
wrong_grandparents:
{
GST_CAT_DEBUG (GST_CAT_CAPS,
"pads have different grandparents %" GST_PTR_FORMAT " and %"
GST_PTR_FORMAT, psrc, psink);
return FALSE;
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
}
}
/* FIXME leftover from an attempt at refactoring... */
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
/* call with the two pads unlocked, when this function returns GST_PAD_LINK_OK,
* the two pads will be locked in the srcpad, sinkpad order. */
static GstPadLinkReturn
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
gst_pad_link_prepare (GstPad * srcpad, GstPad * sinkpad)
{
/* generic checks */
g_return_val_if_fail (GST_IS_PAD (srcpad), GST_PAD_LINK_REFUSED);
g_return_val_if_fail (GST_IS_PAD (sinkpad), GST_PAD_LINK_REFUSED);
GST_CAT_INFO (GST_CAT_PADS, "trying to link %s:%s and %s:%s",
GST_DEBUG_PAD_NAME (srcpad), GST_DEBUG_PAD_NAME (sinkpad));
GST_OBJECT_LOCK (srcpad);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
if (G_UNLIKELY (GST_PAD_DIRECTION (srcpad) != GST_PAD_SRC))
goto not_srcpad;
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
if (G_UNLIKELY (GST_PAD_PEER (srcpad) != NULL))
goto src_was_linked;
GST_OBJECT_LOCK (sinkpad);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
if (G_UNLIKELY (GST_PAD_DIRECTION (sinkpad) != GST_PAD_SINK))
goto not_sinkpad;
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
if (G_UNLIKELY (GST_PAD_PEER (sinkpad) != NULL))
goto sink_was_linked;
/* check hierarchy, pads can only be linked if the grandparents
* are the same. */
if (!gst_pad_link_check_hierarchy (srcpad, sinkpad))
goto wrong_hierarchy;
/* check pad caps for non-empty intersection */
if (!gst_pad_link_check_compatible_unlocked (srcpad, sinkpad))
goto no_format;
/* FIXME check pad scheduling for non-empty intersection */
return GST_PAD_LINK_OK;
not_srcpad:
{
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
g_critical ("pad %s is not a source pad", GST_PAD_NAME (srcpad));
GST_OBJECT_UNLOCK (srcpad);
return GST_PAD_LINK_WRONG_DIRECTION;
}
src_was_linked:
{
gst/gstbin.c: (de)activate src pads before calling state_change on the childs. Original commit message from CVS: * gst/gstbin.c: (activate_pads), (iterator_activate_fold_with_resync), (gst_bin_src_pads_activate), (gst_bin_change_state_func): (de)activate src pads before calling state_change on the childs. This is to avoid the case where a src ghostpad is blocked (holding the stream lock), which would block the deactivation of the ghostpad's target pad. * gst/gstghostpad.c: (gst_proxy_pad_do_query_type), (gst_proxy_pad_do_event), (gst_proxy_pad_do_query), (gst_proxy_pad_do_internal_link), (gst_proxy_pad_do_bufferalloc), (gst_proxy_pad_do_chain), (gst_proxy_pad_do_getrange), (gst_proxy_pad_do_checkgetrange), (gst_proxy_pad_do_getcaps), (gst_proxy_pad_do_acceptcaps), (gst_proxy_pad_do_fixatecaps), (gst_proxy_pad_do_setcaps), (gst_proxy_pad_set_target_unlocked), (gst_proxy_pad_set_target), (gst_proxy_pad_get_internal), (gst_proxy_pad_dispose), (gst_proxy_pad_init), (gst_ghost_pad_parent_set), (gst_ghost_pad_parent_unset), (gst_ghost_pad_class_init), (gst_ghost_pad_internal_do_activate_push), (gst_ghost_pad_internal_do_activate_pull), (gst_ghost_pad_do_activate_push), (gst_ghost_pad_do_activate_pull), (gst_ghost_pad_do_link), (gst_ghost_pad_do_unlink), (gst_ghost_pad_dispose), (gst_ghost_pad_new_no_target), (gst_ghost_pad_new), (gst_ghost_pad_set_target): GhostPads now create their internal GstProxyPad at creation (and not when they're linked, as it was being done previously). The internal and target pads are linked straight away. The data will also travel through the other pad in order to make pad blocking and probes non-hackish (the probe/block now really happens on the GhostPad and not on the target). * gst/gstpad.c: (gst_pad_set_blocked_async), (gst_pad_link_prepare), (gst_pad_push_event): Remove previous ghostpad cruft. * gst/gstutils.c: (gst_pad_add_data_probe), (gst_pad_add_event_probe), (gst_pad_add_buffer_probe), (gst_pad_remove_data_probe), (gst_pad_remove_event_probe), (gst_pad_remove_buffer_probe): Remove previous ghost pad cruft. Added more detailed debug statements. * tests/check/gst/gstghostpad.c: (GST_START_TEST): Fix the testsuite for refcounting changes. The comments about who has references were correct, but the refcount being checked wasn't the same (!?!).
2006-07-11 16:20:09 +00:00
GST_CAT_INFO (GST_CAT_PADS, "src %s:%s was already linked to %s:%s",
GST_DEBUG_PAD_NAME (srcpad),
GST_DEBUG_PAD_NAME (GST_PAD_PEER (srcpad)));
/* we do not emit a warning in this case because unlinking cannot
* be made MT safe.*/
GST_OBJECT_UNLOCK (srcpad);
return GST_PAD_LINK_WAS_LINKED;
}
not_sinkpad:
{
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
g_critical ("pad %s is not a sink pad", GST_PAD_NAME (sinkpad));
GST_OBJECT_UNLOCK (sinkpad);
GST_OBJECT_UNLOCK (srcpad);
return GST_PAD_LINK_WRONG_DIRECTION;
}
sink_was_linked:
{
gst/gstbin.c: (de)activate src pads before calling state_change on the childs. Original commit message from CVS: * gst/gstbin.c: (activate_pads), (iterator_activate_fold_with_resync), (gst_bin_src_pads_activate), (gst_bin_change_state_func): (de)activate src pads before calling state_change on the childs. This is to avoid the case where a src ghostpad is blocked (holding the stream lock), which would block the deactivation of the ghostpad's target pad. * gst/gstghostpad.c: (gst_proxy_pad_do_query_type), (gst_proxy_pad_do_event), (gst_proxy_pad_do_query), (gst_proxy_pad_do_internal_link), (gst_proxy_pad_do_bufferalloc), (gst_proxy_pad_do_chain), (gst_proxy_pad_do_getrange), (gst_proxy_pad_do_checkgetrange), (gst_proxy_pad_do_getcaps), (gst_proxy_pad_do_acceptcaps), (gst_proxy_pad_do_fixatecaps), (gst_proxy_pad_do_setcaps), (gst_proxy_pad_set_target_unlocked), (gst_proxy_pad_set_target), (gst_proxy_pad_get_internal), (gst_proxy_pad_dispose), (gst_proxy_pad_init), (gst_ghost_pad_parent_set), (gst_ghost_pad_parent_unset), (gst_ghost_pad_class_init), (gst_ghost_pad_internal_do_activate_push), (gst_ghost_pad_internal_do_activate_pull), (gst_ghost_pad_do_activate_push), (gst_ghost_pad_do_activate_pull), (gst_ghost_pad_do_link), (gst_ghost_pad_do_unlink), (gst_ghost_pad_dispose), (gst_ghost_pad_new_no_target), (gst_ghost_pad_new), (gst_ghost_pad_set_target): GhostPads now create their internal GstProxyPad at creation (and not when they're linked, as it was being done previously). The internal and target pads are linked straight away. The data will also travel through the other pad in order to make pad blocking and probes non-hackish (the probe/block now really happens on the GhostPad and not on the target). * gst/gstpad.c: (gst_pad_set_blocked_async), (gst_pad_link_prepare), (gst_pad_push_event): Remove previous ghostpad cruft. * gst/gstutils.c: (gst_pad_add_data_probe), (gst_pad_add_event_probe), (gst_pad_add_buffer_probe), (gst_pad_remove_data_probe), (gst_pad_remove_event_probe), (gst_pad_remove_buffer_probe): Remove previous ghost pad cruft. Added more detailed debug statements. * tests/check/gst/gstghostpad.c: (GST_START_TEST): Fix the testsuite for refcounting changes. The comments about who has references were correct, but the refcount being checked wasn't the same (!?!).
2006-07-11 16:20:09 +00:00
GST_CAT_INFO (GST_CAT_PADS, "sink %s:%s was already linked to %s:%s",
GST_DEBUG_PAD_NAME (sinkpad),
GST_DEBUG_PAD_NAME (GST_PAD_PEER (sinkpad)));
/* we do not emit a warning in this case because unlinking cannot
* be made MT safe.*/
GST_OBJECT_UNLOCK (sinkpad);
GST_OBJECT_UNLOCK (srcpad);
return GST_PAD_LINK_WAS_LINKED;
}
wrong_hierarchy:
{
GST_CAT_INFO (GST_CAT_PADS, "pads have wrong hierarchy");
GST_OBJECT_UNLOCK (sinkpad);
GST_OBJECT_UNLOCK (srcpad);
return GST_PAD_LINK_WRONG_HIERARCHY;
}
no_format:
{
GST_CAT_INFO (GST_CAT_PADS, "caps are incompatible");
GST_OBJECT_UNLOCK (sinkpad);
GST_OBJECT_UNLOCK (srcpad);
return GST_PAD_LINK_NOFORMAT;
}
}
/**
* gst_pad_link:
* @srcpad: the source #GstPad to link.
* @sinkpad: the sink #GstPad to link.
*
* Links the source pad and the sink pad.
*
* Returns: A result code indicating if the connection worked or
* what went wrong.
*
* MT Safe.
*/
GstPadLinkReturn
gst_pad_link (GstPad * srcpad, GstPad * sinkpad)
{
GstPadLinkReturn result;
/* prepare will also lock the two pads */
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
result = gst_pad_link_prepare (srcpad, sinkpad);
if (result != GST_PAD_LINK_OK)
goto prepare_failed;
/* must set peers before calling the link function */
GST_PAD_PEER (srcpad) = sinkpad;
GST_PAD_PEER (sinkpad) = srcpad;
GST_OBJECT_UNLOCK (sinkpad);
GST_OBJECT_UNLOCK (srcpad);
/* FIXME released the locks here, concurrent thread might link
* something else. */
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
if (GST_PAD_LINKFUNC (srcpad)) {
/* this one will call the peer link function */
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
result = GST_PAD_LINKFUNC (srcpad) (srcpad, sinkpad);
} else if (GST_PAD_LINKFUNC (sinkpad)) {
/* if no source link function, we need to call the sink link
* function ourselves. */
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
result = GST_PAD_LINKFUNC (sinkpad) (sinkpad, srcpad);
} else {
result = GST_PAD_LINK_OK;
}
GST_OBJECT_LOCK (srcpad);
GST_OBJECT_LOCK (sinkpad);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
if (result == GST_PAD_LINK_OK) {
GST_OBJECT_UNLOCK (sinkpad);
GST_OBJECT_UNLOCK (srcpad);
/* fire off a signal to each of the pads telling them
* that they've been linked */
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
g_signal_emit (G_OBJECT (srcpad), gst_pad_signals[PAD_LINKED], 0, sinkpad);
g_signal_emit (G_OBJECT (sinkpad), gst_pad_signals[PAD_LINKED], 0, srcpad);
GST_CAT_INFO (GST_CAT_PADS, "linked %s:%s and %s:%s, successful",
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
GST_DEBUG_PAD_NAME (srcpad), GST_DEBUG_PAD_NAME (sinkpad));
} else {
GST_CAT_INFO (GST_CAT_PADS, "link between %s:%s and %s:%s failed",
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
GST_DEBUG_PAD_NAME (srcpad), GST_DEBUG_PAD_NAME (sinkpad));
GST_PAD_PEER (srcpad) = NULL;
GST_PAD_PEER (sinkpad) = NULL;
GST_OBJECT_UNLOCK (sinkpad);
GST_OBJECT_UNLOCK (srcpad);
}
return result;
prepare_failed:
{
return result;
}
}
static void
gst_pad_set_pad_template (GstPad * pad, GstPadTemplate * templ)
{
GstPadTemplate **template_p;
/* this function would need checks if it weren't static */
GST_OBJECT_LOCK (pad);
template_p = &pad->padtemplate;
gst_object_replace ((GstObject **) template_p, (GstObject *) templ);
GST_OBJECT_UNLOCK (pad);
if (templ)
gst_pad_template_pad_created (templ, pad);
}
gst/gstelement.c (gst_element_dispose): Protect against multiple invocations. Original commit message from CVS: 2004-02-24 Andy Wingo <wingo@pobox.com> * gst/gstelement.c (gst_element_dispose): Protect against multiple invocations. * gst/schedulers/gstoptimalscheduler.c I added a mess of prototypes at the top of the file by way of documentation. Some of the operations on chains and groups were re-organized. (create_group): Added a type argument so if the group is enabled, the setup_group_scheduler knows what to do. (group_elements): Added a type argument here, too, to be passed on to create_group. (group_element_set_enabled): If an unlinked PLAYING element is added to a bin, we have to create a new group to hold the element, and this function will be called before the group is added to the chain. Thus we have a valid case for group->chain==NULL. Instead of calling chain_group_set_enabled, just set the flag on the group (the chain's status will be set when the group is added to it). (gst_opt_scheduler_state_transition, chain_group_set_enabled): Setup the group scheduler when the group is enabled, not specifically when an element goes PAUSED->PLAYING. This means PLAYING elements can be added, linked, and scheduled into a PLAYING pipeline, as was intended. (add_to_group): Don't ref the group twice. I don't know when this double-ref got in here. Removing it has the potential to cause segfaults if other parts of the scheduler are buggy. If you find that the scheduler is segfaulting for you, put in an extra ref here and see if that hacks over the underlying issue. Of course, then find out what code is unreffing a group it doesn't own... (create_group): Make the extra refcount floating, and remove it after adding the element. This means that... (unref_group): Destroy when the refcount reaches 0, not 1, like every other refcounted object in the known universe. (remove_from_group): When a group becomes empty, set it to be not active, and remove it from its chain. Don't unref it again, there's no floating reference any more. (destroy_group): We have to remove the group from the chain in remove_from_group (rather than here) to break refcounting cycles (the chain always has a ref on the group). So assert that group->chain==NULL. (ref_group_by_count): Removed, it was commented out anyway. (merge_chains): Use the remove_from_chain and add_to_chain primitives to do the reparenting, instead of rolling our own implementation. (add_to_chain): The first non-disabled group in the chain's group list will be the entry point for the chain. Because buffers can accumulate in loop elements' peer bufpens, we preferentially schedule loop groups before get groups to avoid unnecessary execution of get-based groups when the bufpens are already full. (gst_opt_scheduler_schedule_run_queue): Debug fixes. (get_group_schedule_function): Ditto. (loop_group_schedule_function): Ditto. (gst_opt_scheduler_loop_wrapper): Ditto. (gst_opt_scheduler_iterate): Ditto. I understand the opt scheduler now, yippee! * gst/gstpad.c: All throughout, added FIXMEs to look at for 0.9. (gst_pad_get_name, gst_pad_set_chain_function) (gst_pad_set_get_function, gst_pad_set_event_function) (gst_pad_set_event_mask_function, gst_pad_get_event_masks) (gst_pad_get_event_masks_default, gst_pad_set_convert_function) (gst_pad_set_query_function, gst_pad_get_query_types) (gst_pad_get_query_types_default) (gst_pad_set_internal_link_function) (gst_pad_set_formats_function, gst_pad_set_link_function) (gst_pad_set_fixate_function, gst_pad_set_getcaps_function) (gst_pad_set_bufferalloc_function, gst_pad_unlink) (gst_pad_renegotiate, gst_pad_set_parent, gst_pad_get_parent) (gst_pad_add_ghost_pad, gst_pad_proxy_getcaps) (gst_pad_proxy_pad_link, gst_pad_proxy_fixate) (gst_pad_get_pad_template_caps, gst_pad_check_compatibility) (gst_pad_get_peer, gst_pad_get_allowed_caps) (gst_pad_alloc_buffer, gst_pad_push, gst_pad_pull) (gst_pad_selectv, gst_pad_select, gst_pad_template_get_caps) (gst_pad_event_default_dispatch, gst_pad_event_default) (gst_pad_dispatcher, gst_pad_send_event, gst_pad_convert_default) (gst_pad_convert, gst_pad_query_default, gst_pad_query) (gst_pad_get_formats_default, gst_pad_get_formats): Better argument checks, and some doc fixes. (gst_pad_custom_new_from_template): Um, does anyone use these functions? Actually make a custom pad instead of a normal one. (gst_pad_try_set_caps): Transpose some checks. (gst_pad_try_set_caps_nonfixed): Same, and use a macro to check if the pad is in negotiation. (gst_pad_try_relink_filtered): Use pad_link_prepare. * gst/gstelement.c: Remove prototypes also defined in gstclock.h. * gst/gstelement.h: * gst/gstclock.h: Un-deprecate the old clocking API, as discussed on the list.
2004-02-25 13:16:12 +00:00
/**
* gst_pad_get_pad_template:
gst/gstelement.c (gst_element_dispose): Protect against multiple invocations. Original commit message from CVS: 2004-02-24 Andy Wingo <wingo@pobox.com> * gst/gstelement.c (gst_element_dispose): Protect against multiple invocations. * gst/schedulers/gstoptimalscheduler.c I added a mess of prototypes at the top of the file by way of documentation. Some of the operations on chains and groups were re-organized. (create_group): Added a type argument so if the group is enabled, the setup_group_scheduler knows what to do. (group_elements): Added a type argument here, too, to be passed on to create_group. (group_element_set_enabled): If an unlinked PLAYING element is added to a bin, we have to create a new group to hold the element, and this function will be called before the group is added to the chain. Thus we have a valid case for group->chain==NULL. Instead of calling chain_group_set_enabled, just set the flag on the group (the chain's status will be set when the group is added to it). (gst_opt_scheduler_state_transition, chain_group_set_enabled): Setup the group scheduler when the group is enabled, not specifically when an element goes PAUSED->PLAYING. This means PLAYING elements can be added, linked, and scheduled into a PLAYING pipeline, as was intended. (add_to_group): Don't ref the group twice. I don't know when this double-ref got in here. Removing it has the potential to cause segfaults if other parts of the scheduler are buggy. If you find that the scheduler is segfaulting for you, put in an extra ref here and see if that hacks over the underlying issue. Of course, then find out what code is unreffing a group it doesn't own... (create_group): Make the extra refcount floating, and remove it after adding the element. This means that... (unref_group): Destroy when the refcount reaches 0, not 1, like every other refcounted object in the known universe. (remove_from_group): When a group becomes empty, set it to be not active, and remove it from its chain. Don't unref it again, there's no floating reference any more. (destroy_group): We have to remove the group from the chain in remove_from_group (rather than here) to break refcounting cycles (the chain always has a ref on the group). So assert that group->chain==NULL. (ref_group_by_count): Removed, it was commented out anyway. (merge_chains): Use the remove_from_chain and add_to_chain primitives to do the reparenting, instead of rolling our own implementation. (add_to_chain): The first non-disabled group in the chain's group list will be the entry point for the chain. Because buffers can accumulate in loop elements' peer bufpens, we preferentially schedule loop groups before get groups to avoid unnecessary execution of get-based groups when the bufpens are already full. (gst_opt_scheduler_schedule_run_queue): Debug fixes. (get_group_schedule_function): Ditto. (loop_group_schedule_function): Ditto. (gst_opt_scheduler_loop_wrapper): Ditto. (gst_opt_scheduler_iterate): Ditto. I understand the opt scheduler now, yippee! * gst/gstpad.c: All throughout, added FIXMEs to look at for 0.9. (gst_pad_get_name, gst_pad_set_chain_function) (gst_pad_set_get_function, gst_pad_set_event_function) (gst_pad_set_event_mask_function, gst_pad_get_event_masks) (gst_pad_get_event_masks_default, gst_pad_set_convert_function) (gst_pad_set_query_function, gst_pad_get_query_types) (gst_pad_get_query_types_default) (gst_pad_set_internal_link_function) (gst_pad_set_formats_function, gst_pad_set_link_function) (gst_pad_set_fixate_function, gst_pad_set_getcaps_function) (gst_pad_set_bufferalloc_function, gst_pad_unlink) (gst_pad_renegotiate, gst_pad_set_parent, gst_pad_get_parent) (gst_pad_add_ghost_pad, gst_pad_proxy_getcaps) (gst_pad_proxy_pad_link, gst_pad_proxy_fixate) (gst_pad_get_pad_template_caps, gst_pad_check_compatibility) (gst_pad_get_peer, gst_pad_get_allowed_caps) (gst_pad_alloc_buffer, gst_pad_push, gst_pad_pull) (gst_pad_selectv, gst_pad_select, gst_pad_template_get_caps) (gst_pad_event_default_dispatch, gst_pad_event_default) (gst_pad_dispatcher, gst_pad_send_event, gst_pad_convert_default) (gst_pad_convert, gst_pad_query_default, gst_pad_query) (gst_pad_get_formats_default, gst_pad_get_formats): Better argument checks, and some doc fixes. (gst_pad_custom_new_from_template): Um, does anyone use these functions? Actually make a custom pad instead of a normal one. (gst_pad_try_set_caps): Transpose some checks. (gst_pad_try_set_caps_nonfixed): Same, and use a macro to check if the pad is in negotiation. (gst_pad_try_relink_filtered): Use pad_link_prepare. * gst/gstelement.c: Remove prototypes also defined in gstclock.h. * gst/gstelement.h: * gst/gstclock.h: Un-deprecate the old clocking API, as discussed on the list.
2004-02-25 13:16:12 +00:00
* @pad: a #GstPad.
*
gst/gstelement.c (gst_element_dispose): Protect against multiple invocations. Original commit message from CVS: 2004-02-24 Andy Wingo <wingo@pobox.com> * gst/gstelement.c (gst_element_dispose): Protect against multiple invocations. * gst/schedulers/gstoptimalscheduler.c I added a mess of prototypes at the top of the file by way of documentation. Some of the operations on chains and groups were re-organized. (create_group): Added a type argument so if the group is enabled, the setup_group_scheduler knows what to do. (group_elements): Added a type argument here, too, to be passed on to create_group. (group_element_set_enabled): If an unlinked PLAYING element is added to a bin, we have to create a new group to hold the element, and this function will be called before the group is added to the chain. Thus we have a valid case for group->chain==NULL. Instead of calling chain_group_set_enabled, just set the flag on the group (the chain's status will be set when the group is added to it). (gst_opt_scheduler_state_transition, chain_group_set_enabled): Setup the group scheduler when the group is enabled, not specifically when an element goes PAUSED->PLAYING. This means PLAYING elements can be added, linked, and scheduled into a PLAYING pipeline, as was intended. (add_to_group): Don't ref the group twice. I don't know when this double-ref got in here. Removing it has the potential to cause segfaults if other parts of the scheduler are buggy. If you find that the scheduler is segfaulting for you, put in an extra ref here and see if that hacks over the underlying issue. Of course, then find out what code is unreffing a group it doesn't own... (create_group): Make the extra refcount floating, and remove it after adding the element. This means that... (unref_group): Destroy when the refcount reaches 0, not 1, like every other refcounted object in the known universe. (remove_from_group): When a group becomes empty, set it to be not active, and remove it from its chain. Don't unref it again, there's no floating reference any more. (destroy_group): We have to remove the group from the chain in remove_from_group (rather than here) to break refcounting cycles (the chain always has a ref on the group). So assert that group->chain==NULL. (ref_group_by_count): Removed, it was commented out anyway. (merge_chains): Use the remove_from_chain and add_to_chain primitives to do the reparenting, instead of rolling our own implementation. (add_to_chain): The first non-disabled group in the chain's group list will be the entry point for the chain. Because buffers can accumulate in loop elements' peer bufpens, we preferentially schedule loop groups before get groups to avoid unnecessary execution of get-based groups when the bufpens are already full. (gst_opt_scheduler_schedule_run_queue): Debug fixes. (get_group_schedule_function): Ditto. (loop_group_schedule_function): Ditto. (gst_opt_scheduler_loop_wrapper): Ditto. (gst_opt_scheduler_iterate): Ditto. I understand the opt scheduler now, yippee! * gst/gstpad.c: All throughout, added FIXMEs to look at for 0.9. (gst_pad_get_name, gst_pad_set_chain_function) (gst_pad_set_get_function, gst_pad_set_event_function) (gst_pad_set_event_mask_function, gst_pad_get_event_masks) (gst_pad_get_event_masks_default, gst_pad_set_convert_function) (gst_pad_set_query_function, gst_pad_get_query_types) (gst_pad_get_query_types_default) (gst_pad_set_internal_link_function) (gst_pad_set_formats_function, gst_pad_set_link_function) (gst_pad_set_fixate_function, gst_pad_set_getcaps_function) (gst_pad_set_bufferalloc_function, gst_pad_unlink) (gst_pad_renegotiate, gst_pad_set_parent, gst_pad_get_parent) (gst_pad_add_ghost_pad, gst_pad_proxy_getcaps) (gst_pad_proxy_pad_link, gst_pad_proxy_fixate) (gst_pad_get_pad_template_caps, gst_pad_check_compatibility) (gst_pad_get_peer, gst_pad_get_allowed_caps) (gst_pad_alloc_buffer, gst_pad_push, gst_pad_pull) (gst_pad_selectv, gst_pad_select, gst_pad_template_get_caps) (gst_pad_event_default_dispatch, gst_pad_event_default) (gst_pad_dispatcher, gst_pad_send_event, gst_pad_convert_default) (gst_pad_convert, gst_pad_query_default, gst_pad_query) (gst_pad_get_formats_default, gst_pad_get_formats): Better argument checks, and some doc fixes. (gst_pad_custom_new_from_template): Um, does anyone use these functions? Actually make a custom pad instead of a normal one. (gst_pad_try_set_caps): Transpose some checks. (gst_pad_try_set_caps_nonfixed): Same, and use a macro to check if the pad is in negotiation. (gst_pad_try_relink_filtered): Use pad_link_prepare. * gst/gstelement.c: Remove prototypes also defined in gstclock.h. * gst/gstelement.h: * gst/gstclock.h: Un-deprecate the old clocking API, as discussed on the list.
2004-02-25 13:16:12 +00:00
* Gets the template for @pad.
*
gst/gstelement.c (gst_element_dispose): Protect against multiple invocations. Original commit message from CVS: 2004-02-24 Andy Wingo <wingo@pobox.com> * gst/gstelement.c (gst_element_dispose): Protect against multiple invocations. * gst/schedulers/gstoptimalscheduler.c I added a mess of prototypes at the top of the file by way of documentation. Some of the operations on chains and groups were re-organized. (create_group): Added a type argument so if the group is enabled, the setup_group_scheduler knows what to do. (group_elements): Added a type argument here, too, to be passed on to create_group. (group_element_set_enabled): If an unlinked PLAYING element is added to a bin, we have to create a new group to hold the element, and this function will be called before the group is added to the chain. Thus we have a valid case for group->chain==NULL. Instead of calling chain_group_set_enabled, just set the flag on the group (the chain's status will be set when the group is added to it). (gst_opt_scheduler_state_transition, chain_group_set_enabled): Setup the group scheduler when the group is enabled, not specifically when an element goes PAUSED->PLAYING. This means PLAYING elements can be added, linked, and scheduled into a PLAYING pipeline, as was intended. (add_to_group): Don't ref the group twice. I don't know when this double-ref got in here. Removing it has the potential to cause segfaults if other parts of the scheduler are buggy. If you find that the scheduler is segfaulting for you, put in an extra ref here and see if that hacks over the underlying issue. Of course, then find out what code is unreffing a group it doesn't own... (create_group): Make the extra refcount floating, and remove it after adding the element. This means that... (unref_group): Destroy when the refcount reaches 0, not 1, like every other refcounted object in the known universe. (remove_from_group): When a group becomes empty, set it to be not active, and remove it from its chain. Don't unref it again, there's no floating reference any more. (destroy_group): We have to remove the group from the chain in remove_from_group (rather than here) to break refcounting cycles (the chain always has a ref on the group). So assert that group->chain==NULL. (ref_group_by_count): Removed, it was commented out anyway. (merge_chains): Use the remove_from_chain and add_to_chain primitives to do the reparenting, instead of rolling our own implementation. (add_to_chain): The first non-disabled group in the chain's group list will be the entry point for the chain. Because buffers can accumulate in loop elements' peer bufpens, we preferentially schedule loop groups before get groups to avoid unnecessary execution of get-based groups when the bufpens are already full. (gst_opt_scheduler_schedule_run_queue): Debug fixes. (get_group_schedule_function): Ditto. (loop_group_schedule_function): Ditto. (gst_opt_scheduler_loop_wrapper): Ditto. (gst_opt_scheduler_iterate): Ditto. I understand the opt scheduler now, yippee! * gst/gstpad.c: All throughout, added FIXMEs to look at for 0.9. (gst_pad_get_name, gst_pad_set_chain_function) (gst_pad_set_get_function, gst_pad_set_event_function) (gst_pad_set_event_mask_function, gst_pad_get_event_masks) (gst_pad_get_event_masks_default, gst_pad_set_convert_function) (gst_pad_set_query_function, gst_pad_get_query_types) (gst_pad_get_query_types_default) (gst_pad_set_internal_link_function) (gst_pad_set_formats_function, gst_pad_set_link_function) (gst_pad_set_fixate_function, gst_pad_set_getcaps_function) (gst_pad_set_bufferalloc_function, gst_pad_unlink) (gst_pad_renegotiate, gst_pad_set_parent, gst_pad_get_parent) (gst_pad_add_ghost_pad, gst_pad_proxy_getcaps) (gst_pad_proxy_pad_link, gst_pad_proxy_fixate) (gst_pad_get_pad_template_caps, gst_pad_check_compatibility) (gst_pad_get_peer, gst_pad_get_allowed_caps) (gst_pad_alloc_buffer, gst_pad_push, gst_pad_pull) (gst_pad_selectv, gst_pad_select, gst_pad_template_get_caps) (gst_pad_event_default_dispatch, gst_pad_event_default) (gst_pad_dispatcher, gst_pad_send_event, gst_pad_convert_default) (gst_pad_convert, gst_pad_query_default, gst_pad_query) (gst_pad_get_formats_default, gst_pad_get_formats): Better argument checks, and some doc fixes. (gst_pad_custom_new_from_template): Um, does anyone use these functions? Actually make a custom pad instead of a normal one. (gst_pad_try_set_caps): Transpose some checks. (gst_pad_try_set_caps_nonfixed): Same, and use a macro to check if the pad is in negotiation. (gst_pad_try_relink_filtered): Use pad_link_prepare. * gst/gstelement.c: Remove prototypes also defined in gstclock.h. * gst/gstelement.h: * gst/gstclock.h: Un-deprecate the old clocking API, as discussed on the list.
2004-02-25 13:16:12 +00:00
* Returns: the #GstPadTemplate from which this pad was instantiated, or %NULL
* if this pad has no template.
*
* FIXME: currently returns an unrefcounted padtemplate.
*/
GstPadTemplate *
gst_pad_get_pad_template (GstPad * pad)
{
g_return_val_if_fail (GST_IS_PAD (pad), NULL);
return GST_PAD_PAD_TEMPLATE (pad);
}
/* should be called with the pad LOCK held */
/* refs the caps, so caller is responsible for getting it unreffed */
static GstCaps *
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
gst_pad_get_caps_unlocked (GstPad * pad)
{
GstCaps *result = NULL;
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_DEBUG_OBJECT (GST_CAT_CAPS, pad, "get pad caps");
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
if (GST_PAD_GETCAPSFUNC (pad)) {
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_DEBUG_OBJECT (GST_CAT_CAPS, pad,
"dispatching to pad getcaps function");
renamed GST_FLAGS macros to GST_OBJECT_FLAGS moved bitshift from macro to enum definition Original commit message from CVS: * check/gst/gstbin.c: (GST_START_TEST): * docs/gst/gstreamer-sections.txt: * gst/base/gstbasesink.c: (gst_base_sink_init): * gst/base/gstbasesrc.c: (gst_base_src_init), (gst_base_src_get_range), (gst_base_src_check_get_range), (gst_base_src_start), (gst_base_src_stop): * gst/base/gstbasesrc.h: * gst/elements/gstfakesrc.c: (gst_fake_src_set_property): * gst/gstbin.c: (gst_bin_add_func), (gst_bin_remove_func), (bin_element_is_sink), (reset_degree), (gst_bin_element_set_state), (bin_bus_handler): * gst/gstbin.h: * gst/gstbuffer.h: * gst/gstbus.c: (gst_bus_post), (gst_bus_set_flushing): * gst/gstbus.h: * gst/gstelement.c: (gst_element_is_locked_state), (gst_element_set_locked_state), (gst_element_commit_state), (gst_element_set_state): * gst/gstelement.h: * gst/gstindex.c: (gst_index_init): * gst/gstindex.h: * gst/gstminiobject.h: * gst/gstobject.c: (gst_object_init), (gst_object_sink), (gst_object_set_parent): * gst/gstobject.h: * gst/gstpad.c: (gst_pad_set_blocked_async), (gst_pad_is_blocked), (gst_pad_get_caps_unlocked), (gst_pad_set_caps): * gst/gstpad.h: * gst/gstpadtemplate.h: * gst/gstpipeline.c: (gst_pipeline_provide_clock_func), (gst_pipeline_use_clock), (gst_pipeline_auto_clock): * gst/gstpipeline.h: * gst/indexers/gstfileindex.c: (gst_file_index_load), (gst_file_index_commit): * testsuite/bytestream/filepadsink.c: (gst_fp_sink_init): * testsuite/pad/link.c: (gst_test_src_init), (gst_test_filter_init), (gst_test_sink_init): * testsuite/states/locked.c: (main): renamed GST_FLAGS macros to GST_OBJECT_FLAGS moved bitshift from macro to enum definition
2005-10-12 14:28:39 +00:00
GST_OBJECT_FLAG_SET (pad, GST_PAD_IN_GETCAPS);
GST_OBJECT_UNLOCK (pad);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
result = GST_PAD_GETCAPSFUNC (pad) (pad);
GST_OBJECT_LOCK (pad);
renamed GST_FLAGS macros to GST_OBJECT_FLAGS moved bitshift from macro to enum definition Original commit message from CVS: * check/gst/gstbin.c: (GST_START_TEST): * docs/gst/gstreamer-sections.txt: * gst/base/gstbasesink.c: (gst_base_sink_init): * gst/base/gstbasesrc.c: (gst_base_src_init), (gst_base_src_get_range), (gst_base_src_check_get_range), (gst_base_src_start), (gst_base_src_stop): * gst/base/gstbasesrc.h: * gst/elements/gstfakesrc.c: (gst_fake_src_set_property): * gst/gstbin.c: (gst_bin_add_func), (gst_bin_remove_func), (bin_element_is_sink), (reset_degree), (gst_bin_element_set_state), (bin_bus_handler): * gst/gstbin.h: * gst/gstbuffer.h: * gst/gstbus.c: (gst_bus_post), (gst_bus_set_flushing): * gst/gstbus.h: * gst/gstelement.c: (gst_element_is_locked_state), (gst_element_set_locked_state), (gst_element_commit_state), (gst_element_set_state): * gst/gstelement.h: * gst/gstindex.c: (gst_index_init): * gst/gstindex.h: * gst/gstminiobject.h: * gst/gstobject.c: (gst_object_init), (gst_object_sink), (gst_object_set_parent): * gst/gstobject.h: * gst/gstpad.c: (gst_pad_set_blocked_async), (gst_pad_is_blocked), (gst_pad_get_caps_unlocked), (gst_pad_set_caps): * gst/gstpad.h: * gst/gstpadtemplate.h: * gst/gstpipeline.c: (gst_pipeline_provide_clock_func), (gst_pipeline_use_clock), (gst_pipeline_auto_clock): * gst/gstpipeline.h: * gst/indexers/gstfileindex.c: (gst_file_index_load), (gst_file_index_commit): * testsuite/bytestream/filepadsink.c: (gst_fp_sink_init): * testsuite/pad/link.c: (gst_test_src_init), (gst_test_filter_init), (gst_test_sink_init): * testsuite/states/locked.c: (main): renamed GST_FLAGS macros to GST_OBJECT_FLAGS moved bitshift from macro to enum definition
2005-10-12 14:28:39 +00:00
GST_OBJECT_FLAG_UNSET (pad, GST_PAD_IN_GETCAPS);
if (result == NULL) {
g_critical ("pad %s:%s returned NULL caps from getcaps function",
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
GST_DEBUG_PAD_NAME (pad));
} else {
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_DEBUG_OBJECT (GST_CAT_CAPS, pad,
"pad getcaps returned %" GST_PTR_FORMAT, result);
#ifndef G_DISABLE_ASSERT
/* check that the returned caps are a real subset of the template caps */
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
if (GST_PAD_PAD_TEMPLATE (pad)) {
const GstCaps *templ_caps =
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
GST_PAD_TEMPLATE_CAPS (GST_PAD_PAD_TEMPLATE (pad));
if (!gst_caps_is_subset (result, templ_caps)) {
GstCaps *temp;
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
GST_CAT_ERROR_OBJECT (GST_CAT_CAPS, pad,
"pad returned caps %" GST_PTR_FORMAT
" which are not a real subset of its template caps %"
GST_PTR_FORMAT, result, templ_caps);
g_warning
("pad %s:%s returned caps which are not a real "
"subset of its template caps", GST_DEBUG_PAD_NAME (pad));
temp = gst_caps_intersect (templ_caps, result);
gst_caps_unref (result);
result = temp;
}
}
#endif
goto done;
}
}
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
if (GST_PAD_PAD_TEMPLATE (pad)) {
GstPadTemplate *templ = GST_PAD_PAD_TEMPLATE (pad);
result = GST_PAD_TEMPLATE_CAPS (templ);
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_DEBUG_OBJECT (GST_CAT_CAPS, pad,
"using pad template %p with caps %p %" GST_PTR_FORMAT, templ, result,
result);
result = gst_caps_ref (result);
goto done;
}
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
if (GST_PAD_CAPS (pad)) {
result = GST_PAD_CAPS (pad);
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_DEBUG_OBJECT (GST_CAT_CAPS, pad,
"using pad caps %p %" GST_PTR_FORMAT, result, result);
result = gst_caps_ref (result);
goto done;
}
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_DEBUG_OBJECT (GST_CAT_CAPS, pad, "pad has no caps");
result = gst_caps_new_empty ();
done:
return result;
}
/**
* gst_pad_get_caps:
* @pad: a #GstPad to get the capabilities of.
*
* Gets the capabilities this pad can produce or consume.
* Note that this method doesn't necessarily return the caps set by
* gst_pad_set_caps() - use #GST_PAD_CAPS for that instead.
* gst_pad_get_caps returns all possible caps a pad can operate with, using
* the pad's get_caps function;
* this returns the pad template caps if not explicitly set.
*
* Returns: a newly allocated copy of the #GstCaps of this pad.
*
* MT safe.
*/
GstCaps *
gst_pad_get_caps (GstPad * pad)
{
GstCaps *result = NULL;
gst/gstelement.c (gst_element_dispose): Protect against multiple invocations. Original commit message from CVS: 2004-02-24 Andy Wingo <wingo@pobox.com> * gst/gstelement.c (gst_element_dispose): Protect against multiple invocations. * gst/schedulers/gstoptimalscheduler.c I added a mess of prototypes at the top of the file by way of documentation. Some of the operations on chains and groups were re-organized. (create_group): Added a type argument so if the group is enabled, the setup_group_scheduler knows what to do. (group_elements): Added a type argument here, too, to be passed on to create_group. (group_element_set_enabled): If an unlinked PLAYING element is added to a bin, we have to create a new group to hold the element, and this function will be called before the group is added to the chain. Thus we have a valid case for group->chain==NULL. Instead of calling chain_group_set_enabled, just set the flag on the group (the chain's status will be set when the group is added to it). (gst_opt_scheduler_state_transition, chain_group_set_enabled): Setup the group scheduler when the group is enabled, not specifically when an element goes PAUSED->PLAYING. This means PLAYING elements can be added, linked, and scheduled into a PLAYING pipeline, as was intended. (add_to_group): Don't ref the group twice. I don't know when this double-ref got in here. Removing it has the potential to cause segfaults if other parts of the scheduler are buggy. If you find that the scheduler is segfaulting for you, put in an extra ref here and see if that hacks over the underlying issue. Of course, then find out what code is unreffing a group it doesn't own... (create_group): Make the extra refcount floating, and remove it after adding the element. This means that... (unref_group): Destroy when the refcount reaches 0, not 1, like every other refcounted object in the known universe. (remove_from_group): When a group becomes empty, set it to be not active, and remove it from its chain. Don't unref it again, there's no floating reference any more. (destroy_group): We have to remove the group from the chain in remove_from_group (rather than here) to break refcounting cycles (the chain always has a ref on the group). So assert that group->chain==NULL. (ref_group_by_count): Removed, it was commented out anyway. (merge_chains): Use the remove_from_chain and add_to_chain primitives to do the reparenting, instead of rolling our own implementation. (add_to_chain): The first non-disabled group in the chain's group list will be the entry point for the chain. Because buffers can accumulate in loop elements' peer bufpens, we preferentially schedule loop groups before get groups to avoid unnecessary execution of get-based groups when the bufpens are already full. (gst_opt_scheduler_schedule_run_queue): Debug fixes. (get_group_schedule_function): Ditto. (loop_group_schedule_function): Ditto. (gst_opt_scheduler_loop_wrapper): Ditto. (gst_opt_scheduler_iterate): Ditto. I understand the opt scheduler now, yippee! * gst/gstpad.c: All throughout, added FIXMEs to look at for 0.9. (gst_pad_get_name, gst_pad_set_chain_function) (gst_pad_set_get_function, gst_pad_set_event_function) (gst_pad_set_event_mask_function, gst_pad_get_event_masks) (gst_pad_get_event_masks_default, gst_pad_set_convert_function) (gst_pad_set_query_function, gst_pad_get_query_types) (gst_pad_get_query_types_default) (gst_pad_set_internal_link_function) (gst_pad_set_formats_function, gst_pad_set_link_function) (gst_pad_set_fixate_function, gst_pad_set_getcaps_function) (gst_pad_set_bufferalloc_function, gst_pad_unlink) (gst_pad_renegotiate, gst_pad_set_parent, gst_pad_get_parent) (gst_pad_add_ghost_pad, gst_pad_proxy_getcaps) (gst_pad_proxy_pad_link, gst_pad_proxy_fixate) (gst_pad_get_pad_template_caps, gst_pad_check_compatibility) (gst_pad_get_peer, gst_pad_get_allowed_caps) (gst_pad_alloc_buffer, gst_pad_push, gst_pad_pull) (gst_pad_selectv, gst_pad_select, gst_pad_template_get_caps) (gst_pad_event_default_dispatch, gst_pad_event_default) (gst_pad_dispatcher, gst_pad_send_event, gst_pad_convert_default) (gst_pad_convert, gst_pad_query_default, gst_pad_query) (gst_pad_get_formats_default, gst_pad_get_formats): Better argument checks, and some doc fixes. (gst_pad_custom_new_from_template): Um, does anyone use these functions? Actually make a custom pad instead of a normal one. (gst_pad_try_set_caps): Transpose some checks. (gst_pad_try_set_caps_nonfixed): Same, and use a macro to check if the pad is in negotiation. (gst_pad_try_relink_filtered): Use pad_link_prepare. * gst/gstelement.c: Remove prototypes also defined in gstclock.h. * gst/gstelement.h: * gst/gstclock.h: Un-deprecate the old clocking API, as discussed on the list.
2004-02-25 13:16:12 +00:00
g_return_val_if_fail (GST_IS_PAD (pad), NULL);
GST_OBJECT_LOCK (pad);
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_DEBUG_OBJECT (GST_CAT_CAPS, pad, "get pad caps");
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
result = gst_pad_get_caps_unlocked (pad);
GST_OBJECT_UNLOCK (pad);
return result;
}
/**
* gst_pad_peer_get_caps:
* @pad: a #GstPad to get the peer capabilities of.
*
* Gets the capabilities of the peer connected to this pad.
*
* Returns: the #GstCaps of the peer pad. This function returns a new caps, so use
* gst_caps_unref to get rid of it. this function returns NULL if there is no
* peer pad.
*/
GstCaps *
gst_pad_peer_get_caps (GstPad * pad)
{
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
GstPad *peerpad;
GstCaps *result = NULL;
g_return_val_if_fail (GST_IS_PAD (pad), NULL);
GST_OBJECT_LOCK (pad);
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_DEBUG_OBJECT (GST_CAT_CAPS, pad, "get peer caps");
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
peerpad = GST_PAD_PEER (pad);
if (G_UNLIKELY (peerpad == NULL))
goto no_peer;
gst_object_ref (peerpad);
GST_OBJECT_UNLOCK (pad);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
result = gst_pad_get_caps (peerpad);
gst_object_unref (peerpad);
return result;
no_peer:
{
GST_OBJECT_UNLOCK (pad);
return NULL;
}
}
static gboolean
fixate_value (GValue * dest, const GValue * src)
{
if (G_VALUE_TYPE (src) == GST_TYPE_INT_RANGE) {
g_value_init (dest, G_TYPE_INT);
g_value_set_int (dest, gst_value_get_int_range_min (src));
} else if (G_VALUE_TYPE (src) == GST_TYPE_DOUBLE_RANGE) {
g_value_init (dest, G_TYPE_DOUBLE);
g_value_set_double (dest, gst_value_get_double_range_min (src));
} else if (G_VALUE_TYPE (src) == GST_TYPE_FRACTION_RANGE) {
gst_value_init_and_copy (dest, gst_value_get_fraction_range_min (src));
} else if (G_VALUE_TYPE (src) == GST_TYPE_LIST) {
GValue temp = { 0 };
gst_value_init_and_copy (&temp, gst_value_list_get_value (src, 0));
if (!fixate_value (dest, &temp))
gst_value_init_and_copy (dest, &temp);
g_value_unset (&temp);
} else if (G_VALUE_TYPE (src) == GST_TYPE_ARRAY) {
gboolean res = FALSE;
guint n;
g_value_init (dest, GST_TYPE_ARRAY);
for (n = 0; n < gst_value_array_get_size (src); n++) {
GValue kid = { 0 };
const GValue *orig_kid = gst_value_array_get_value (src, n);
if (!fixate_value (&kid, orig_kid))
gst_value_init_and_copy (&kid, orig_kid);
else
res = TRUE;
gst_value_array_append_value (dest, &kid);
g_value_unset (&kid);
}
if (!res)
g_value_unset (dest);
return res;
} else {
return FALSE;
}
return TRUE;
}
static gboolean
gst_pad_default_fixate (GQuark field_id, const GValue * value, gpointer data)
{
GstStructure *s = data;
GValue v = { 0 };
if (fixate_value (&v, value)) {
gst_structure_id_set_value (s, field_id, &v);
g_value_unset (&v);
}
return TRUE;
}
/**
* gst_pad_fixate_caps:
* @pad: a #GstPad to fixate
* @caps: the #GstCaps to fixate
*
* Fixate a caps on the given pad. Modifies the caps in place, so you should
* make sure that the caps are actually writable (see gst_caps_make_writable()).
*/
tests/network-clock.scm: Commentary update. Original commit message from CVS: 2005-07-01 Andy Wingo <wingo@pobox.com> * tests/network-clock.scm: Commentary update. * gst/elements/gstidentity.c (PROP_DUPLICATE): Gone daddy gone. Didn't really make sense, not implementable with basetransform, etc. (gst_identity_transform): Unref inbuf via make_writable. Feeble attempt at implementing the sync property, needs an unlock method. * gst/base/gstbasetransform.c (gst_base_transform_transform_caps): New func, by default returns the same caps (the identity transformation). (gst_base_transform_getcaps): Uses transform_caps to return something sensible. (gst_base_transform_setcaps): Complicated logic to get caps on both pads, even if they are different, and to call set_caps once for every time both pads get their caps set. (gst_base_transform_handle_buffer): Give the ref to the transform function. Allows in-place modification of the buffer. * gst/base/gstbasetransform.h (transform_caps): New class method. Given caps on one side, what can I do on the other. (set_caps): Take two caps, one for each side of the element. * gst/gstpad.h: * gst/gstpad.c (gst_pad_fixate_caps): Change prototype to modify caps in place. This is safe because we can check the mutability of the caps, and a good idea because fixate functions are just called as a matter of last resort. (Not actually implemented.) (gst_pad_set_caps): If the caps we're setting is actually the same as the existing pad caps, just update the pointer without calling setcaps. Assert that caps is either NULL or fixed, as per the docs. * gst/gstghostpad.c: Update for fixate changes.
2005-07-01 16:46:59 +00:00
void
gst_pad_fixate_caps (GstPad * pad, GstCaps * caps)
{
GstPadFixateCapsFunction fixatefunc;
guint n;
g_return_if_fail (GST_IS_PAD (pad));
g_return_if_fail (caps != NULL);
if (gst_caps_is_fixed (caps))
return;
fixatefunc = GST_PAD_FIXATECAPSFUNC (pad);
if (fixatefunc) {
fixatefunc (pad, caps);
}
/* default fixation */
for (n = 0; n < gst_caps_get_size (caps); n++) {
GstStructure *s = gst_caps_get_structure (caps, n);
gst_structure_foreach (s, gst_pad_default_fixate, s);
}
}
/* Default accept caps implementation just checks against
* against the allowed caps for the pad */
static gboolean
gst_pad_acceptcaps_default (GstPad * pad, GstCaps * caps)
{
/* get the caps and see if it intersects to something
* not empty */
GstCaps *intersect;
GstCaps *allowed;
gboolean result = FALSE;
allowed = gst_pad_get_caps (pad);
if (!allowed)
goto nothing_allowed;
intersect = gst_caps_intersect (allowed, caps);
result = !gst_caps_is_empty (intersect);
if (!result)
GST_DEBUG_OBJECT (pad, "intersection gave empty caps");
gst_caps_unref (allowed);
gst_caps_unref (intersect);
return result;
/* ERRORS */
nothing_allowed:
{
GST_DEBUG_OBJECT (pad, "no caps allowed on the pad");
return FALSE;
}
}
/**
* gst_pad_accept_caps:
* @pad: a #GstPad to check
* @caps: a #GstCaps to check on the pad
*
* Check if the given pad accepts the caps.
*
* Returns: TRUE if the pad can accept the caps.
*/
gboolean
gst_pad_accept_caps (GstPad * pad, GstCaps * caps)
{
gboolean result;
GstPadAcceptCapsFunction acceptfunc;
GstCaps *existing = NULL;
g_return_val_if_fail (GST_IS_PAD (pad), FALSE);
/* any pad can be unnegotiated */
if (caps == NULL)
return TRUE;
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
/* lock for checking the existing caps */
GST_OBJECT_LOCK (pad);
acceptfunc = GST_PAD_ACCEPTCAPSFUNC (pad);
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_DEBUG_OBJECT (GST_CAT_CAPS, pad, "accept caps of %p", caps);
/* The current caps on a pad are trivially acceptable */
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
if (G_LIKELY ((existing = GST_PAD_CAPS (pad)))) {
if (caps == existing || gst_caps_is_equal (caps, existing))
goto is_same_caps;
}
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_OBJECT_UNLOCK (pad);
if (G_LIKELY (acceptfunc)) {
/* we can call the function */
result = acceptfunc (pad, caps);
GST_DEBUG_OBJECT (pad, "acceptfunc returned %d", result);
} else {
/* Only null if the element explicitly unset it */
result = gst_pad_acceptcaps_default (pad, caps);
GST_DEBUG_OBJECT (pad, "default acceptcaps returned %d", result);
}
return result;
is_same_caps:
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
{
GST_DEBUG_OBJECT (pad, "pad had same caps");
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_OBJECT_UNLOCK (pad);
return TRUE;
}
}
/**
* gst_pad_peer_accept_caps:
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
* @pad: a #GstPad to check the peer of
* @caps: a #GstCaps to check on the pad
*
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
* Check if the peer of @pad accepts @caps. If @pad has no peer, this function
* returns TRUE.
*
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
* Returns: TRUE if the peer of @pad can accept the caps or @pad has no peer.
*/
gboolean
gst_pad_peer_accept_caps (GstPad * pad, GstCaps * caps)
{
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
GstPad *peerpad;
gboolean result;
g_return_val_if_fail (GST_IS_PAD (pad), FALSE);
GST_OBJECT_LOCK (pad);
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_DEBUG_OBJECT (GST_CAT_CAPS, pad, "peer accept caps of (%p)", pad);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
peerpad = GST_PAD_PEER (pad);
if (G_UNLIKELY (peerpad == NULL))
goto no_peer;
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
result = gst_pad_accept_caps (peerpad, caps);
GST_OBJECT_UNLOCK (pad);
return result;
no_peer:
{
GST_OBJECT_UNLOCK (pad);
return TRUE;
}
}
/**
* gst_pad_set_caps:
* @pad: a #GstPad to set the capabilities of.
* @caps: a #GstCaps to set.
*
* Sets the capabilities of this pad. The caps must be fixed. Any previous
* caps on the pad will be unreffed. This function refs the caps so you should
* unref if as soon as you don't need it anymore.
* It is possible to set NULL caps, which will make the pad unnegotiated
* again.
*
* Returns: TRUE if the caps could be set. FALSE if the caps were not fixed
* or bad parameters were provided to this function.
*
* MT safe.
*/
gboolean
gst_pad_set_caps (GstPad * pad, GstCaps * caps)
{
GstPadSetCapsFunction setcaps;
tests/network-clock.scm: Commentary update. Original commit message from CVS: 2005-07-01 Andy Wingo <wingo@pobox.com> * tests/network-clock.scm: Commentary update. * gst/elements/gstidentity.c (PROP_DUPLICATE): Gone daddy gone. Didn't really make sense, not implementable with basetransform, etc. (gst_identity_transform): Unref inbuf via make_writable. Feeble attempt at implementing the sync property, needs an unlock method. * gst/base/gstbasetransform.c (gst_base_transform_transform_caps): New func, by default returns the same caps (the identity transformation). (gst_base_transform_getcaps): Uses transform_caps to return something sensible. (gst_base_transform_setcaps): Complicated logic to get caps on both pads, even if they are different, and to call set_caps once for every time both pads get their caps set. (gst_base_transform_handle_buffer): Give the ref to the transform function. Allows in-place modification of the buffer. * gst/base/gstbasetransform.h (transform_caps): New class method. Given caps on one side, what can I do on the other. (set_caps): Take two caps, one for each side of the element. * gst/gstpad.h: * gst/gstpad.c (gst_pad_fixate_caps): Change prototype to modify caps in place. This is safe because we can check the mutability of the caps, and a good idea because fixate functions are just called as a matter of last resort. (Not actually implemented.) (gst_pad_set_caps): If the caps we're setting is actually the same as the existing pad caps, just update the pointer without calling setcaps. Assert that caps is either NULL or fixed, as per the docs. * gst/gstghostpad.c: Update for fixate changes.
2005-07-01 16:46:59 +00:00
GstCaps *existing;
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
g_return_val_if_fail (GST_IS_PAD (pad), FALSE);
tests/network-clock.scm: Commentary update. Original commit message from CVS: 2005-07-01 Andy Wingo <wingo@pobox.com> * tests/network-clock.scm: Commentary update. * gst/elements/gstidentity.c (PROP_DUPLICATE): Gone daddy gone. Didn't really make sense, not implementable with basetransform, etc. (gst_identity_transform): Unref inbuf via make_writable. Feeble attempt at implementing the sync property, needs an unlock method. * gst/base/gstbasetransform.c (gst_base_transform_transform_caps): New func, by default returns the same caps (the identity transformation). (gst_base_transform_getcaps): Uses transform_caps to return something sensible. (gst_base_transform_setcaps): Complicated logic to get caps on both pads, even if they are different, and to call set_caps once for every time both pads get their caps set. (gst_base_transform_handle_buffer): Give the ref to the transform function. Allows in-place modification of the buffer. * gst/base/gstbasetransform.h (transform_caps): New class method. Given caps on one side, what can I do on the other. (set_caps): Take two caps, one for each side of the element. * gst/gstpad.h: * gst/gstpad.c (gst_pad_fixate_caps): Change prototype to modify caps in place. This is safe because we can check the mutability of the caps, and a good idea because fixate functions are just called as a matter of last resort. (Not actually implemented.) (gst_pad_set_caps): If the caps we're setting is actually the same as the existing pad caps, just update the pointer without calling setcaps. Assert that caps is either NULL or fixed, as per the docs. * gst/gstghostpad.c: Update for fixate changes.
2005-07-01 16:46:59 +00:00
g_return_val_if_fail (caps == NULL || gst_caps_is_fixed (caps), FALSE);
GST_OBJECT_LOCK (pad);
tests/network-clock.scm: Commentary update. Original commit message from CVS: 2005-07-01 Andy Wingo <wingo@pobox.com> * tests/network-clock.scm: Commentary update. * gst/elements/gstidentity.c (PROP_DUPLICATE): Gone daddy gone. Didn't really make sense, not implementable with basetransform, etc. (gst_identity_transform): Unref inbuf via make_writable. Feeble attempt at implementing the sync property, needs an unlock method. * gst/base/gstbasetransform.c (gst_base_transform_transform_caps): New func, by default returns the same caps (the identity transformation). (gst_base_transform_getcaps): Uses transform_caps to return something sensible. (gst_base_transform_setcaps): Complicated logic to get caps on both pads, even if they are different, and to call set_caps once for every time both pads get their caps set. (gst_base_transform_handle_buffer): Give the ref to the transform function. Allows in-place modification of the buffer. * gst/base/gstbasetransform.h (transform_caps): New class method. Given caps on one side, what can I do on the other. (set_caps): Take two caps, one for each side of the element. * gst/gstpad.h: * gst/gstpad.c (gst_pad_fixate_caps): Change prototype to modify caps in place. This is safe because we can check the mutability of the caps, and a good idea because fixate functions are just called as a matter of last resort. (Not actually implemented.) (gst_pad_set_caps): If the caps we're setting is actually the same as the existing pad caps, just update the pointer without calling setcaps. Assert that caps is either NULL or fixed, as per the docs. * gst/gstghostpad.c: Update for fixate changes.
2005-07-01 16:46:59 +00:00
existing = GST_PAD_CAPS (pad);
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
if (existing == caps)
goto was_ok;
if (gst_caps_is_equal (caps, existing))
tests/network-clock.scm: Commentary update. Original commit message from CVS: 2005-07-01 Andy Wingo <wingo@pobox.com> * tests/network-clock.scm: Commentary update. * gst/elements/gstidentity.c (PROP_DUPLICATE): Gone daddy gone. Didn't really make sense, not implementable with basetransform, etc. (gst_identity_transform): Unref inbuf via make_writable. Feeble attempt at implementing the sync property, needs an unlock method. * gst/base/gstbasetransform.c (gst_base_transform_transform_caps): New func, by default returns the same caps (the identity transformation). (gst_base_transform_getcaps): Uses transform_caps to return something sensible. (gst_base_transform_setcaps): Complicated logic to get caps on both pads, even if they are different, and to call set_caps once for every time both pads get their caps set. (gst_base_transform_handle_buffer): Give the ref to the transform function. Allows in-place modification of the buffer. * gst/base/gstbasetransform.h (transform_caps): New class method. Given caps on one side, what can I do on the other. (set_caps): Take two caps, one for each side of the element. * gst/gstpad.h: * gst/gstpad.c (gst_pad_fixate_caps): Change prototype to modify caps in place. This is safe because we can check the mutability of the caps, and a good idea because fixate functions are just called as a matter of last resort. (Not actually implemented.) (gst_pad_set_caps): If the caps we're setting is actually the same as the existing pad caps, just update the pointer without calling setcaps. Assert that caps is either NULL or fixed, as per the docs. * gst/gstghostpad.c: Update for fixate changes.
2005-07-01 16:46:59 +00:00
goto setting_same_caps;
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
setcaps = GST_PAD_SETCAPSFUNC (pad);
/* call setcaps function to configure the pad only if the
* caps is not NULL */
if (setcaps != NULL && caps) {
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
if (!GST_PAD_IS_IN_SETCAPS (pad)) {
renamed GST_FLAGS macros to GST_OBJECT_FLAGS moved bitshift from macro to enum definition Original commit message from CVS: * check/gst/gstbin.c: (GST_START_TEST): * docs/gst/gstreamer-sections.txt: * gst/base/gstbasesink.c: (gst_base_sink_init): * gst/base/gstbasesrc.c: (gst_base_src_init), (gst_base_src_get_range), (gst_base_src_check_get_range), (gst_base_src_start), (gst_base_src_stop): * gst/base/gstbasesrc.h: * gst/elements/gstfakesrc.c: (gst_fake_src_set_property): * gst/gstbin.c: (gst_bin_add_func), (gst_bin_remove_func), (bin_element_is_sink), (reset_degree), (gst_bin_element_set_state), (bin_bus_handler): * gst/gstbin.h: * gst/gstbuffer.h: * gst/gstbus.c: (gst_bus_post), (gst_bus_set_flushing): * gst/gstbus.h: * gst/gstelement.c: (gst_element_is_locked_state), (gst_element_set_locked_state), (gst_element_commit_state), (gst_element_set_state): * gst/gstelement.h: * gst/gstindex.c: (gst_index_init): * gst/gstindex.h: * gst/gstminiobject.h: * gst/gstobject.c: (gst_object_init), (gst_object_sink), (gst_object_set_parent): * gst/gstobject.h: * gst/gstpad.c: (gst_pad_set_blocked_async), (gst_pad_is_blocked), (gst_pad_get_caps_unlocked), (gst_pad_set_caps): * gst/gstpad.h: * gst/gstpadtemplate.h: * gst/gstpipeline.c: (gst_pipeline_provide_clock_func), (gst_pipeline_use_clock), (gst_pipeline_auto_clock): * gst/gstpipeline.h: * gst/indexers/gstfileindex.c: (gst_file_index_load), (gst_file_index_commit): * testsuite/bytestream/filepadsink.c: (gst_fp_sink_init): * testsuite/pad/link.c: (gst_test_src_init), (gst_test_filter_init), (gst_test_sink_init): * testsuite/states/locked.c: (main): renamed GST_FLAGS macros to GST_OBJECT_FLAGS moved bitshift from macro to enum definition
2005-10-12 14:28:39 +00:00
GST_OBJECT_FLAG_SET (pad, GST_PAD_IN_SETCAPS);
GST_OBJECT_UNLOCK (pad);
if (!setcaps (pad, caps))
goto could_not_set;
GST_OBJECT_LOCK (pad);
renamed GST_FLAGS macros to GST_OBJECT_FLAGS moved bitshift from macro to enum definition Original commit message from CVS: * check/gst/gstbin.c: (GST_START_TEST): * docs/gst/gstreamer-sections.txt: * gst/base/gstbasesink.c: (gst_base_sink_init): * gst/base/gstbasesrc.c: (gst_base_src_init), (gst_base_src_get_range), (gst_base_src_check_get_range), (gst_base_src_start), (gst_base_src_stop): * gst/base/gstbasesrc.h: * gst/elements/gstfakesrc.c: (gst_fake_src_set_property): * gst/gstbin.c: (gst_bin_add_func), (gst_bin_remove_func), (bin_element_is_sink), (reset_degree), (gst_bin_element_set_state), (bin_bus_handler): * gst/gstbin.h: * gst/gstbuffer.h: * gst/gstbus.c: (gst_bus_post), (gst_bus_set_flushing): * gst/gstbus.h: * gst/gstelement.c: (gst_element_is_locked_state), (gst_element_set_locked_state), (gst_element_commit_state), (gst_element_set_state): * gst/gstelement.h: * gst/gstindex.c: (gst_index_init): * gst/gstindex.h: * gst/gstminiobject.h: * gst/gstobject.c: (gst_object_init), (gst_object_sink), (gst_object_set_parent): * gst/gstobject.h: * gst/gstpad.c: (gst_pad_set_blocked_async), (gst_pad_is_blocked), (gst_pad_get_caps_unlocked), (gst_pad_set_caps): * gst/gstpad.h: * gst/gstpadtemplate.h: * gst/gstpipeline.c: (gst_pipeline_provide_clock_func), (gst_pipeline_use_clock), (gst_pipeline_auto_clock): * gst/gstpipeline.h: * gst/indexers/gstfileindex.c: (gst_file_index_load), (gst_file_index_commit): * testsuite/bytestream/filepadsink.c: (gst_fp_sink_init): * testsuite/pad/link.c: (gst_test_src_init), (gst_test_filter_init), (gst_test_sink_init): * testsuite/states/locked.c: (main): renamed GST_FLAGS macros to GST_OBJECT_FLAGS moved bitshift from macro to enum definition
2005-10-12 14:28:39 +00:00
GST_OBJECT_FLAG_UNSET (pad, GST_PAD_IN_SETCAPS);
} else {
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_DEBUG_OBJECT (GST_CAT_CAPS, pad, "pad was dispatching");
}
}
gst_caps_replace (&GST_PAD_CAPS (pad), caps);
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_DEBUG_OBJECT (GST_CAT_CAPS, pad, "caps %" GST_PTR_FORMAT, caps);
GST_OBJECT_UNLOCK (pad);
g_object_notify (G_OBJECT (pad), "caps");
return TRUE;
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
was_ok:
{
GST_OBJECT_UNLOCK (pad);
return TRUE;
}
tests/network-clock.scm: Commentary update. Original commit message from CVS: 2005-07-01 Andy Wingo <wingo@pobox.com> * tests/network-clock.scm: Commentary update. * gst/elements/gstidentity.c (PROP_DUPLICATE): Gone daddy gone. Didn't really make sense, not implementable with basetransform, etc. (gst_identity_transform): Unref inbuf via make_writable. Feeble attempt at implementing the sync property, needs an unlock method. * gst/base/gstbasetransform.c (gst_base_transform_transform_caps): New func, by default returns the same caps (the identity transformation). (gst_base_transform_getcaps): Uses transform_caps to return something sensible. (gst_base_transform_setcaps): Complicated logic to get caps on both pads, even if they are different, and to call set_caps once for every time both pads get their caps set. (gst_base_transform_handle_buffer): Give the ref to the transform function. Allows in-place modification of the buffer. * gst/base/gstbasetransform.h (transform_caps): New class method. Given caps on one side, what can I do on the other. (set_caps): Take two caps, one for each side of the element. * gst/gstpad.h: * gst/gstpad.c (gst_pad_fixate_caps): Change prototype to modify caps in place. This is safe because we can check the mutability of the caps, and a good idea because fixate functions are just called as a matter of last resort. (Not actually implemented.) (gst_pad_set_caps): If the caps we're setting is actually the same as the existing pad caps, just update the pointer without calling setcaps. Assert that caps is either NULL or fixed, as per the docs. * gst/gstghostpad.c: Update for fixate changes.
2005-07-01 16:46:59 +00:00
setting_same_caps:
{
gst_caps_replace (&GST_PAD_CAPS (pad), caps);
GST_CAT_DEBUG_OBJECT (GST_CAT_CAPS, pad,
"caps %" GST_PTR_FORMAT " same as existing, updating ptr only", caps);
GST_OBJECT_UNLOCK (pad);
tests/network-clock.scm: Commentary update. Original commit message from CVS: 2005-07-01 Andy Wingo <wingo@pobox.com> * tests/network-clock.scm: Commentary update. * gst/elements/gstidentity.c (PROP_DUPLICATE): Gone daddy gone. Didn't really make sense, not implementable with basetransform, etc. (gst_identity_transform): Unref inbuf via make_writable. Feeble attempt at implementing the sync property, needs an unlock method. * gst/base/gstbasetransform.c (gst_base_transform_transform_caps): New func, by default returns the same caps (the identity transformation). (gst_base_transform_getcaps): Uses transform_caps to return something sensible. (gst_base_transform_setcaps): Complicated logic to get caps on both pads, even if they are different, and to call set_caps once for every time both pads get their caps set. (gst_base_transform_handle_buffer): Give the ref to the transform function. Allows in-place modification of the buffer. * gst/base/gstbasetransform.h (transform_caps): New class method. Given caps on one side, what can I do on the other. (set_caps): Take two caps, one for each side of the element. * gst/gstpad.h: * gst/gstpad.c (gst_pad_fixate_caps): Change prototype to modify caps in place. This is safe because we can check the mutability of the caps, and a good idea because fixate functions are just called as a matter of last resort. (Not actually implemented.) (gst_pad_set_caps): If the caps we're setting is actually the same as the existing pad caps, just update the pointer without calling setcaps. Assert that caps is either NULL or fixed, as per the docs. * gst/gstghostpad.c: Update for fixate changes.
2005-07-01 16:46:59 +00:00
return TRUE;
}
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
/* ERRORS */
could_not_set:
{
GST_OBJECT_LOCK (pad);
renamed GST_FLAGS macros to GST_OBJECT_FLAGS moved bitshift from macro to enum definition Original commit message from CVS: * check/gst/gstbin.c: (GST_START_TEST): * docs/gst/gstreamer-sections.txt: * gst/base/gstbasesink.c: (gst_base_sink_init): * gst/base/gstbasesrc.c: (gst_base_src_init), (gst_base_src_get_range), (gst_base_src_check_get_range), (gst_base_src_start), (gst_base_src_stop): * gst/base/gstbasesrc.h: * gst/elements/gstfakesrc.c: (gst_fake_src_set_property): * gst/gstbin.c: (gst_bin_add_func), (gst_bin_remove_func), (bin_element_is_sink), (reset_degree), (gst_bin_element_set_state), (bin_bus_handler): * gst/gstbin.h: * gst/gstbuffer.h: * gst/gstbus.c: (gst_bus_post), (gst_bus_set_flushing): * gst/gstbus.h: * gst/gstelement.c: (gst_element_is_locked_state), (gst_element_set_locked_state), (gst_element_commit_state), (gst_element_set_state): * gst/gstelement.h: * gst/gstindex.c: (gst_index_init): * gst/gstindex.h: * gst/gstminiobject.h: * gst/gstobject.c: (gst_object_init), (gst_object_sink), (gst_object_set_parent): * gst/gstobject.h: * gst/gstpad.c: (gst_pad_set_blocked_async), (gst_pad_is_blocked), (gst_pad_get_caps_unlocked), (gst_pad_set_caps): * gst/gstpad.h: * gst/gstpadtemplate.h: * gst/gstpipeline.c: (gst_pipeline_provide_clock_func), (gst_pipeline_use_clock), (gst_pipeline_auto_clock): * gst/gstpipeline.h: * gst/indexers/gstfileindex.c: (gst_file_index_load), (gst_file_index_commit): * testsuite/bytestream/filepadsink.c: (gst_fp_sink_init): * testsuite/pad/link.c: (gst_test_src_init), (gst_test_filter_init), (gst_test_sink_init): * testsuite/states/locked.c: (main): renamed GST_FLAGS macros to GST_OBJECT_FLAGS moved bitshift from macro to enum definition
2005-10-12 14:28:39 +00:00
GST_OBJECT_FLAG_UNSET (pad, GST_PAD_IN_SETCAPS);
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_DEBUG_OBJECT (GST_CAT_CAPS, pad,
"caps %" GST_PTR_FORMAT " could not be set", caps);
GST_OBJECT_UNLOCK (pad);
return FALSE;
}
}
static gboolean
gst_pad_configure_sink (GstPad * pad, GstCaps * caps)
{
GstPadSetCapsFunction setcaps;
gboolean res;
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
setcaps = GST_PAD_SETCAPSFUNC (pad);
/* See if pad accepts the caps - only needed if
* no setcaps function */
if (setcaps == NULL)
if (!gst_pad_accept_caps (pad, caps))
goto not_accepted;
/* set caps on pad if call succeeds */
res = gst_pad_set_caps (pad, caps);
/* no need to unref the caps here, set_caps takes a ref and
* our ref goes away when we leave this function. */
return res;
not_accepted:
{
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_DEBUG_OBJECT (GST_CAT_CAPS, pad,
"caps %" GST_PTR_FORMAT " not accepted", caps);
return FALSE;
}
}
/* returns TRUE if the src pad could be configured to accept the given caps */
static gboolean
gst_pad_configure_src (GstPad * pad, GstCaps * caps, gboolean dosetcaps)
{
GstPadSetCapsFunction setcaps;
gboolean res;
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
setcaps = GST_PAD_SETCAPSFUNC (pad);
/* See if pad accepts the caps - only needed if
* no setcaps function */
if (setcaps == NULL)
if (!gst_pad_accept_caps (pad, caps))
goto not_accepted;
if (dosetcaps)
res = gst_pad_set_caps (pad, caps);
else
res = TRUE;
return res;
not_accepted:
{
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_DEBUG_OBJECT (GST_CAT_CAPS, pad,
"caps %" GST_PTR_FORMAT " not accepted", caps);
return FALSE;
}
}
/**
* gst_pad_get_pad_template_caps:
* @pad: a #GstPad to get the template capabilities from.
*
gst/gstpad.c (gst_pad_custom_new): Add a FIXME, this is a hacky way to do inheritance. Original commit message from CVS: 2004-02-10 Andy Wingo <wingo@pobox.com> * gst/gstpad.c (gst_pad_custom_new): Add a FIXME, this is a hacky way to do inheritance. (gst_pad_get_event_masks, gst_pad_get_event_masks_default) (gst_pad_get_query_types, gst_pad_get_query_types_default): Routine docs. (gst_pad_set_link_function, gst_pad_set_fixate_function) (gst_pad_set_getcaps_function): Doc from Dave's negotation random doc. (gst_pad_unlink, gst_pad_is_linked): Docs. (gst_pad_renegotiate): A brief description of capsnego. (gst_pad_try_set_caps): Document. (gst_pad_try_set_caps_nonfixed): Document. (gst_pad_can_link_filtered, gst_pad_link_filtered): Doc fixes. (gst_pad_set_parent): Deprecated (although not out of the API). (gst_pad_get_parent): Deprecated, although many plugins use this. (gst_pad_add_ghost_pad, gst_pad_remove_ghost_pad): Doc that these are private and will go away in 0.9. (gst_pad_perform_negotiate): Doc. (gst_pad_link_unnegotiate): I think this is meant to be static. (gst_pad_get_negotiated_caps, gst_pad_get_pad_template_caps) (gst_pad_template_get_caps_by_name, gst_pad_check_compatibility) (gst_pad_get_peer): Doc updates. (gst_pad_caps_change_notify): Doc. (gst_pad_alloc_buffer, gst_pad_push, gst_static_pad_template_get) (gst_ghost_pad_new): Doc fixes. * gst/gstobject.c (gst_object_get_parent, gst_object_unparent) (gst_object_check_uniqueness): * gst/gstelement.c (gst_element_add_pad) (gst_element_add_ghost_pad, gst_element_remove_pad) (gst_element_remove_ghost_pad, gst_element_get_pad) (gst_element_get_static_pad, gst_element_get_pad_list) (gst_element_class_get_pad_template_list) (gst_element_class_get_pad_template): Work on the docs. (gst_element_get_pad_template_list): Uses the class method. (gst_element_get_compatible_pad_template): Docs, and consolidate some test conditions. (gst_element_get_pad_from_template): New static function. (gst_element_request_compatible_pad): Docs, and work with non-request compatible templates. (gst_element_get_compatible_pad_filtered): Docs and remove redundant checks. (gst_element_get_compatible_pad, gst_element_link_pads_filtered) (gst_element_link_filtered, gst_element_link_many) (gst_element_link, gst_element_link_pads) (gst_element_unlink_many): Docs. 2004-02-05 Andy Wingo <wingo@pobox.com> * gst/gstpad.c (_gst_real_pad_fixate_accumulator): s/pointer/boxed/. * gst/gstmarshal.list (VOID:BOXED, BOXED:BOXED): New marshallers. * gst/gstpad.c (gst_real_pad_class_init): Use a BOXED:BOXED marshaller for ::fixate, and VOID:BOXED for ::caps-nego-failed, with the type=GST_TYPE_CAPS. This allows language bindings to know what kind of data they're dealing with. * gst/gstcaps.c (_gst_caps_value_init): GBoxed values initialize to NULL when g_value_init is called. GstCaps, which rolls its own type implementation, now does the same instead of allocating empty caps. (_gst_caps_initialize, _gst_caps_collect_value, _gst_caps_lcopy_value): Provide collect_value and lcopy_value type table methods. This allows G_VALUE_COLLECT to work. 2004-02-05 Andy Wingo <wingo@pobox.com> * configure.ac: * testsuite/Makefile.am (SUBDIRS): * testsuite/ghostpads/Makefile.am: * testsuite/ghostpads/ghostpads.c: A new test for ghost pads. * gst/gstpad.c (gst_pad_add_ghost_pad, gst_pad_remove_ghost_pad): These two routines are the only ones that set GST_GPAD_REALPAD(gpad), the ghost pad list, and the ghost pad's pad template. They should be made static, depending on ABI needs. (gst_real_pad_dispose): Handle the case of ghost pads without a parent. Assert after dealing with ghost pads that the ghost pad list is empty. (gst_ghost_pad_class_init): New property added, ::real-pad. Can be set after creation. (gst_ghost_pad_dispose): Set ::real-pad to NULL. (gst_ghost_pad_set_property, gst_ghost_pad_get_property): New functions. set_property will call add_ghost_pad/remove_ghost_pad as appropriate. (gst_ghost_pad_new): All the work is offloaded to g_object_new. * gst/gstelement.c (gst_element_add_pad): Handle ghost pads as well. (gst_element_add_ghost_pad): Remove code duplicated from _add_pad. (gst_element_remove_pad): Handle ghost pads as well. (gst_element_remove_ghost_pad): Deprecated (could be removed, depending on API-stability needs). 2004-02-05 Andy Wingo <wingo@pobox.com> * gst/gstbin.[ch]: (gst_bin_get_by_interface): GTypes are scalars, of course they're const
2004-02-11 13:26:04 +00:00
* Gets the capabilities for @pad's template.
*
gst/gstpad.c (_gst_real_pad_fixate_accumulator): s/pointer/boxed/. Original commit message from CVS: 2004-02-05 Andy Wingo <wingo@pobox.com> * gst/gstpad.c (_gst_real_pad_fixate_accumulator): s/pointer/boxed/. * gst/gstmarshal.list (VOID:BOXED, BOXED:BOXED): New marshallers. * gst/gstpad.c (gst_real_pad_class_init): Use a BOXED:BOXED marshaller for ::fixate, and VOID:BOXED for ::caps-nego-failed, with the type=GST_TYPE_CAPS. This allows language bindings to know what kind of data they're dealing with. * gst/gstcaps.c (_gst_caps_value_init): GBoxed values initialize to NULL when g_value_init is called. GstCaps, which rolls its own type implementation, now does the same instead of allocating empty caps. (_gst_caps_initialize, _gst_caps_collect_value, _gst_caps_lcopy_value): Provide collect_value and lcopy_value type table methods. This allows G_VALUE_COLLECT to work. 2004-02-05 Andy Wingo <wingo@pobox.com> * configure.ac: * testsuite/Makefile.am (SUBDIRS): * testsuite/ghostpads/Makefile.am: * testsuite/ghostpads/ghostpads.c: A new test for ghost pads. * gst/gstpad.c (gst_pad_add_ghost_pad, gst_pad_remove_ghost_pad): These two routines are the only ones that set GST_GPAD_REALPAD(gpad), the ghost pad list, and the ghost pad's pad template. They should be made static, depending on ABI needs. (gst_real_pad_dispose): Handle the case of ghost pads without a parent. Assert after dealing with ghost pads that the ghost pad list is empty. (gst_ghost_pad_class_init): New property added, ::real-pad. Can be set after creation. (gst_ghost_pad_dispose): Set ::real-pad to NULL. (gst_ghost_pad_set_property, gst_ghost_pad_get_property): New functions. set_property will call add_ghost_pad/remove_ghost_pad as appropriate. (gst_ghost_pad_new): All the work is offloaded to g_object_new. * gst/gstelement.c (gst_element_add_pad): Handle ghost pads as well. (gst_element_add_ghost_pad): Remove code duplicated from _add_pad. (gst_element_remove_pad): Handle ghost pads as well. (gst_element_remove_ghost_pad): Deprecated (could be removed, depending on API-stability needs). 2004-02-05 Andy Wingo <wingo@pobox.com> * gst/gstbin.[ch]: (gst_bin_get_by_interface): GTypes are scalars, of course they're const
2004-02-05 13:51:26 +00:00
* Returns: the #GstCaps of this pad template. If you intend to keep a reference
* on the caps, make a copy (see gst_caps_copy ()).
*/
const GstCaps *
gst_pad_get_pad_template_caps (GstPad * pad)
{
static GstStaticCaps anycaps = GST_STATIC_CAPS ("ANY");
g_return_val_if_fail (GST_IS_PAD (pad), NULL);
if (GST_PAD_PAD_TEMPLATE (pad))
return GST_PAD_TEMPLATE_CAPS (GST_PAD_PAD_TEMPLATE (pad));
return gst_static_caps_get (&anycaps);
}
/**
* gst_pad_get_peer:
* @pad: a #GstPad to get the peer of.
*
* Gets the peer of @pad. This function refs the peer pad so
* you need to unref it after use.
*
* Returns: the peer #GstPad. Unref after usage.
*
* MT safe.
*/
GstPad *
gst_pad_get_peer (GstPad * pad)
{
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
GstPad *result;
g_return_val_if_fail (GST_IS_PAD (pad), NULL);
GST_OBJECT_LOCK (pad);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
result = GST_PAD_PEER (pad);
if (result)
gst_object_ref (result);
GST_OBJECT_UNLOCK (pad);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
return result;
}
/**
* gst_pad_get_allowed_caps:
* @srcpad: a #GstPad, it must a a source pad.
*
* Gets the capabilities of the allowed media types that can flow through
* @srcpad and its peer. The pad must be a source pad.
gst/gstelement.c (gst_element_dispose): Protect against multiple invocations. Original commit message from CVS: 2004-02-24 Andy Wingo <wingo@pobox.com> * gst/gstelement.c (gst_element_dispose): Protect against multiple invocations. * gst/schedulers/gstoptimalscheduler.c I added a mess of prototypes at the top of the file by way of documentation. Some of the operations on chains and groups were re-organized. (create_group): Added a type argument so if the group is enabled, the setup_group_scheduler knows what to do. (group_elements): Added a type argument here, too, to be passed on to create_group. (group_element_set_enabled): If an unlinked PLAYING element is added to a bin, we have to create a new group to hold the element, and this function will be called before the group is added to the chain. Thus we have a valid case for group->chain==NULL. Instead of calling chain_group_set_enabled, just set the flag on the group (the chain's status will be set when the group is added to it). (gst_opt_scheduler_state_transition, chain_group_set_enabled): Setup the group scheduler when the group is enabled, not specifically when an element goes PAUSED->PLAYING. This means PLAYING elements can be added, linked, and scheduled into a PLAYING pipeline, as was intended. (add_to_group): Don't ref the group twice. I don't know when this double-ref got in here. Removing it has the potential to cause segfaults if other parts of the scheduler are buggy. If you find that the scheduler is segfaulting for you, put in an extra ref here and see if that hacks over the underlying issue. Of course, then find out what code is unreffing a group it doesn't own... (create_group): Make the extra refcount floating, and remove it after adding the element. This means that... (unref_group): Destroy when the refcount reaches 0, not 1, like every other refcounted object in the known universe. (remove_from_group): When a group becomes empty, set it to be not active, and remove it from its chain. Don't unref it again, there's no floating reference any more. (destroy_group): We have to remove the group from the chain in remove_from_group (rather than here) to break refcounting cycles (the chain always has a ref on the group). So assert that group->chain==NULL. (ref_group_by_count): Removed, it was commented out anyway. (merge_chains): Use the remove_from_chain and add_to_chain primitives to do the reparenting, instead of rolling our own implementation. (add_to_chain): The first non-disabled group in the chain's group list will be the entry point for the chain. Because buffers can accumulate in loop elements' peer bufpens, we preferentially schedule loop groups before get groups to avoid unnecessary execution of get-based groups when the bufpens are already full. (gst_opt_scheduler_schedule_run_queue): Debug fixes. (get_group_schedule_function): Ditto. (loop_group_schedule_function): Ditto. (gst_opt_scheduler_loop_wrapper): Ditto. (gst_opt_scheduler_iterate): Ditto. I understand the opt scheduler now, yippee! * gst/gstpad.c: All throughout, added FIXMEs to look at for 0.9. (gst_pad_get_name, gst_pad_set_chain_function) (gst_pad_set_get_function, gst_pad_set_event_function) (gst_pad_set_event_mask_function, gst_pad_get_event_masks) (gst_pad_get_event_masks_default, gst_pad_set_convert_function) (gst_pad_set_query_function, gst_pad_get_query_types) (gst_pad_get_query_types_default) (gst_pad_set_internal_link_function) (gst_pad_set_formats_function, gst_pad_set_link_function) (gst_pad_set_fixate_function, gst_pad_set_getcaps_function) (gst_pad_set_bufferalloc_function, gst_pad_unlink) (gst_pad_renegotiate, gst_pad_set_parent, gst_pad_get_parent) (gst_pad_add_ghost_pad, gst_pad_proxy_getcaps) (gst_pad_proxy_pad_link, gst_pad_proxy_fixate) (gst_pad_get_pad_template_caps, gst_pad_check_compatibility) (gst_pad_get_peer, gst_pad_get_allowed_caps) (gst_pad_alloc_buffer, gst_pad_push, gst_pad_pull) (gst_pad_selectv, gst_pad_select, gst_pad_template_get_caps) (gst_pad_event_default_dispatch, gst_pad_event_default) (gst_pad_dispatcher, gst_pad_send_event, gst_pad_convert_default) (gst_pad_convert, gst_pad_query_default, gst_pad_query) (gst_pad_get_formats_default, gst_pad_get_formats): Better argument checks, and some doc fixes. (gst_pad_custom_new_from_template): Um, does anyone use these functions? Actually make a custom pad instead of a normal one. (gst_pad_try_set_caps): Transpose some checks. (gst_pad_try_set_caps_nonfixed): Same, and use a macro to check if the pad is in negotiation. (gst_pad_try_relink_filtered): Use pad_link_prepare. * gst/gstelement.c: Remove prototypes also defined in gstclock.h. * gst/gstelement.h: * gst/gstclock.h: Un-deprecate the old clocking API, as discussed on the list.
2004-02-25 13:16:12 +00:00
* The caller must free the resulting caps.
*
* Returns: the allowed #GstCaps of the pad link. Free the caps when
* you no longer need it. This function returns NULL when the @srcpad has no
* peer.
*
* MT safe.
*/
GstCaps *
gst_pad_get_allowed_caps (GstPad * srcpad)
{
GstCaps *mycaps;
GstCaps *caps;
GstCaps *peercaps;
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
GstPad *peer;
g_return_val_if_fail (GST_IS_PAD (srcpad), NULL);
g_return_val_if_fail (GST_PAD_IS_SRC (srcpad), NULL);
GST_OBJECT_LOCK (srcpad);
peer = GST_PAD_PEER (srcpad);
if (G_UNLIKELY (peer == NULL))
goto no_peer;
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_DEBUG_OBJECT (GST_CAT_PROPERTIES, srcpad, "getting allowed caps");
gst_object_ref (peer);
GST_OBJECT_UNLOCK (srcpad);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
mycaps = gst_pad_get_caps (srcpad);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
peercaps = gst_pad_get_caps (peer);
gst_object_unref (peer);
caps = gst_caps_intersect (mycaps, peercaps);
gst_caps_unref (peercaps);
gst_caps_unref (mycaps);
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_DEBUG_OBJECT (GST_CAT_CAPS, srcpad, "allowed caps %" GST_PTR_FORMAT,
caps);
return caps;
no_peer:
{
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_DEBUG_OBJECT (GST_CAT_PROPERTIES, srcpad, "no peer");
GST_OBJECT_UNLOCK (srcpad);
return NULL;
}
}
gst/gstpad.c (gst_pad_custom_new): Add a FIXME, this is a hacky way to do inheritance. Original commit message from CVS: 2004-02-10 Andy Wingo <wingo@pobox.com> * gst/gstpad.c (gst_pad_custom_new): Add a FIXME, this is a hacky way to do inheritance. (gst_pad_get_event_masks, gst_pad_get_event_masks_default) (gst_pad_get_query_types, gst_pad_get_query_types_default): Routine docs. (gst_pad_set_link_function, gst_pad_set_fixate_function) (gst_pad_set_getcaps_function): Doc from Dave's negotation random doc. (gst_pad_unlink, gst_pad_is_linked): Docs. (gst_pad_renegotiate): A brief description of capsnego. (gst_pad_try_set_caps): Document. (gst_pad_try_set_caps_nonfixed): Document. (gst_pad_can_link_filtered, gst_pad_link_filtered): Doc fixes. (gst_pad_set_parent): Deprecated (although not out of the API). (gst_pad_get_parent): Deprecated, although many plugins use this. (gst_pad_add_ghost_pad, gst_pad_remove_ghost_pad): Doc that these are private and will go away in 0.9. (gst_pad_perform_negotiate): Doc. (gst_pad_link_unnegotiate): I think this is meant to be static. (gst_pad_get_negotiated_caps, gst_pad_get_pad_template_caps) (gst_pad_template_get_caps_by_name, gst_pad_check_compatibility) (gst_pad_get_peer): Doc updates. (gst_pad_caps_change_notify): Doc. (gst_pad_alloc_buffer, gst_pad_push, gst_static_pad_template_get) (gst_ghost_pad_new): Doc fixes. * gst/gstobject.c (gst_object_get_parent, gst_object_unparent) (gst_object_check_uniqueness): * gst/gstelement.c (gst_element_add_pad) (gst_element_add_ghost_pad, gst_element_remove_pad) (gst_element_remove_ghost_pad, gst_element_get_pad) (gst_element_get_static_pad, gst_element_get_pad_list) (gst_element_class_get_pad_template_list) (gst_element_class_get_pad_template): Work on the docs. (gst_element_get_pad_template_list): Uses the class method. (gst_element_get_compatible_pad_template): Docs, and consolidate some test conditions. (gst_element_get_pad_from_template): New static function. (gst_element_request_compatible_pad): Docs, and work with non-request compatible templates. (gst_element_get_compatible_pad_filtered): Docs and remove redundant checks. (gst_element_get_compatible_pad, gst_element_link_pads_filtered) (gst_element_link_filtered, gst_element_link_many) (gst_element_link, gst_element_link_pads) (gst_element_unlink_many): Docs. 2004-02-05 Andy Wingo <wingo@pobox.com> * gst/gstpad.c (_gst_real_pad_fixate_accumulator): s/pointer/boxed/. * gst/gstmarshal.list (VOID:BOXED, BOXED:BOXED): New marshallers. * gst/gstpad.c (gst_real_pad_class_init): Use a BOXED:BOXED marshaller for ::fixate, and VOID:BOXED for ::caps-nego-failed, with the type=GST_TYPE_CAPS. This allows language bindings to know what kind of data they're dealing with. * gst/gstcaps.c (_gst_caps_value_init): GBoxed values initialize to NULL when g_value_init is called. GstCaps, which rolls its own type implementation, now does the same instead of allocating empty caps. (_gst_caps_initialize, _gst_caps_collect_value, _gst_caps_lcopy_value): Provide collect_value and lcopy_value type table methods. This allows G_VALUE_COLLECT to work. 2004-02-05 Andy Wingo <wingo@pobox.com> * configure.ac: * testsuite/Makefile.am (SUBDIRS): * testsuite/ghostpads/Makefile.am: * testsuite/ghostpads/ghostpads.c: A new test for ghost pads. * gst/gstpad.c (gst_pad_add_ghost_pad, gst_pad_remove_ghost_pad): These two routines are the only ones that set GST_GPAD_REALPAD(gpad), the ghost pad list, and the ghost pad's pad template. They should be made static, depending on ABI needs. (gst_real_pad_dispose): Handle the case of ghost pads without a parent. Assert after dealing with ghost pads that the ghost pad list is empty. (gst_ghost_pad_class_init): New property added, ::real-pad. Can be set after creation. (gst_ghost_pad_dispose): Set ::real-pad to NULL. (gst_ghost_pad_set_property, gst_ghost_pad_get_property): New functions. set_property will call add_ghost_pad/remove_ghost_pad as appropriate. (gst_ghost_pad_new): All the work is offloaded to g_object_new. * gst/gstelement.c (gst_element_add_pad): Handle ghost pads as well. (gst_element_add_ghost_pad): Remove code duplicated from _add_pad. (gst_element_remove_pad): Handle ghost pads as well. (gst_element_remove_ghost_pad): Deprecated (could be removed, depending on API-stability needs). 2004-02-05 Andy Wingo <wingo@pobox.com> * gst/gstbin.[ch]: (gst_bin_get_by_interface): GTypes are scalars, of course they're const
2004-02-11 13:26:04 +00:00
/**
* gst_pad_get_negotiated_caps:
* @pad: a #GstPad.
*
* Gets the capabilities of the media type that currently flows through @pad
* and its peer.
*
* This function can be used on both src and sinkpads. Note that srcpads are
* always negotiated before sinkpads so it is possible that the negotiated caps
* on the srcpad do not match the negotiated caps of the peer.
*
* Returns: the negotiated #GstCaps of the pad link. Free the caps when
* you no longer need it. This function returns NULL when the @pad has no
* peer or is not negotiated yet.
gst/gstpad.c (gst_pad_custom_new): Add a FIXME, this is a hacky way to do inheritance. Original commit message from CVS: 2004-02-10 Andy Wingo <wingo@pobox.com> * gst/gstpad.c (gst_pad_custom_new): Add a FIXME, this is a hacky way to do inheritance. (gst_pad_get_event_masks, gst_pad_get_event_masks_default) (gst_pad_get_query_types, gst_pad_get_query_types_default): Routine docs. (gst_pad_set_link_function, gst_pad_set_fixate_function) (gst_pad_set_getcaps_function): Doc from Dave's negotation random doc. (gst_pad_unlink, gst_pad_is_linked): Docs. (gst_pad_renegotiate): A brief description of capsnego. (gst_pad_try_set_caps): Document. (gst_pad_try_set_caps_nonfixed): Document. (gst_pad_can_link_filtered, gst_pad_link_filtered): Doc fixes. (gst_pad_set_parent): Deprecated (although not out of the API). (gst_pad_get_parent): Deprecated, although many plugins use this. (gst_pad_add_ghost_pad, gst_pad_remove_ghost_pad): Doc that these are private and will go away in 0.9. (gst_pad_perform_negotiate): Doc. (gst_pad_link_unnegotiate): I think this is meant to be static. (gst_pad_get_negotiated_caps, gst_pad_get_pad_template_caps) (gst_pad_template_get_caps_by_name, gst_pad_check_compatibility) (gst_pad_get_peer): Doc updates. (gst_pad_caps_change_notify): Doc. (gst_pad_alloc_buffer, gst_pad_push, gst_static_pad_template_get) (gst_ghost_pad_new): Doc fixes. * gst/gstobject.c (gst_object_get_parent, gst_object_unparent) (gst_object_check_uniqueness): * gst/gstelement.c (gst_element_add_pad) (gst_element_add_ghost_pad, gst_element_remove_pad) (gst_element_remove_ghost_pad, gst_element_get_pad) (gst_element_get_static_pad, gst_element_get_pad_list) (gst_element_class_get_pad_template_list) (gst_element_class_get_pad_template): Work on the docs. (gst_element_get_pad_template_list): Uses the class method. (gst_element_get_compatible_pad_template): Docs, and consolidate some test conditions. (gst_element_get_pad_from_template): New static function. (gst_element_request_compatible_pad): Docs, and work with non-request compatible templates. (gst_element_get_compatible_pad_filtered): Docs and remove redundant checks. (gst_element_get_compatible_pad, gst_element_link_pads_filtered) (gst_element_link_filtered, gst_element_link_many) (gst_element_link, gst_element_link_pads) (gst_element_unlink_many): Docs. 2004-02-05 Andy Wingo <wingo@pobox.com> * gst/gstpad.c (_gst_real_pad_fixate_accumulator): s/pointer/boxed/. * gst/gstmarshal.list (VOID:BOXED, BOXED:BOXED): New marshallers. * gst/gstpad.c (gst_real_pad_class_init): Use a BOXED:BOXED marshaller for ::fixate, and VOID:BOXED for ::caps-nego-failed, with the type=GST_TYPE_CAPS. This allows language bindings to know what kind of data they're dealing with. * gst/gstcaps.c (_gst_caps_value_init): GBoxed values initialize to NULL when g_value_init is called. GstCaps, which rolls its own type implementation, now does the same instead of allocating empty caps. (_gst_caps_initialize, _gst_caps_collect_value, _gst_caps_lcopy_value): Provide collect_value and lcopy_value type table methods. This allows G_VALUE_COLLECT to work. 2004-02-05 Andy Wingo <wingo@pobox.com> * configure.ac: * testsuite/Makefile.am (SUBDIRS): * testsuite/ghostpads/Makefile.am: * testsuite/ghostpads/ghostpads.c: A new test for ghost pads. * gst/gstpad.c (gst_pad_add_ghost_pad, gst_pad_remove_ghost_pad): These two routines are the only ones that set GST_GPAD_REALPAD(gpad), the ghost pad list, and the ghost pad's pad template. They should be made static, depending on ABI needs. (gst_real_pad_dispose): Handle the case of ghost pads without a parent. Assert after dealing with ghost pads that the ghost pad list is empty. (gst_ghost_pad_class_init): New property added, ::real-pad. Can be set after creation. (gst_ghost_pad_dispose): Set ::real-pad to NULL. (gst_ghost_pad_set_property, gst_ghost_pad_get_property): New functions. set_property will call add_ghost_pad/remove_ghost_pad as appropriate. (gst_ghost_pad_new): All the work is offloaded to g_object_new. * gst/gstelement.c (gst_element_add_pad): Handle ghost pads as well. (gst_element_add_ghost_pad): Remove code duplicated from _add_pad. (gst_element_remove_pad): Handle ghost pads as well. (gst_element_remove_ghost_pad): Deprecated (could be removed, depending on API-stability needs). 2004-02-05 Andy Wingo <wingo@pobox.com> * gst/gstbin.[ch]: (gst_bin_get_by_interface): GTypes are scalars, of course they're const
2004-02-11 13:26:04 +00:00
*
* MT safe.
gst/gstpad.c (gst_pad_custom_new): Add a FIXME, this is a hacky way to do inheritance. Original commit message from CVS: 2004-02-10 Andy Wingo <wingo@pobox.com> * gst/gstpad.c (gst_pad_custom_new): Add a FIXME, this is a hacky way to do inheritance. (gst_pad_get_event_masks, gst_pad_get_event_masks_default) (gst_pad_get_query_types, gst_pad_get_query_types_default): Routine docs. (gst_pad_set_link_function, gst_pad_set_fixate_function) (gst_pad_set_getcaps_function): Doc from Dave's negotation random doc. (gst_pad_unlink, gst_pad_is_linked): Docs. (gst_pad_renegotiate): A brief description of capsnego. (gst_pad_try_set_caps): Document. (gst_pad_try_set_caps_nonfixed): Document. (gst_pad_can_link_filtered, gst_pad_link_filtered): Doc fixes. (gst_pad_set_parent): Deprecated (although not out of the API). (gst_pad_get_parent): Deprecated, although many plugins use this. (gst_pad_add_ghost_pad, gst_pad_remove_ghost_pad): Doc that these are private and will go away in 0.9. (gst_pad_perform_negotiate): Doc. (gst_pad_link_unnegotiate): I think this is meant to be static. (gst_pad_get_negotiated_caps, gst_pad_get_pad_template_caps) (gst_pad_template_get_caps_by_name, gst_pad_check_compatibility) (gst_pad_get_peer): Doc updates. (gst_pad_caps_change_notify): Doc. (gst_pad_alloc_buffer, gst_pad_push, gst_static_pad_template_get) (gst_ghost_pad_new): Doc fixes. * gst/gstobject.c (gst_object_get_parent, gst_object_unparent) (gst_object_check_uniqueness): * gst/gstelement.c (gst_element_add_pad) (gst_element_add_ghost_pad, gst_element_remove_pad) (gst_element_remove_ghost_pad, gst_element_get_pad) (gst_element_get_static_pad, gst_element_get_pad_list) (gst_element_class_get_pad_template_list) (gst_element_class_get_pad_template): Work on the docs. (gst_element_get_pad_template_list): Uses the class method. (gst_element_get_compatible_pad_template): Docs, and consolidate some test conditions. (gst_element_get_pad_from_template): New static function. (gst_element_request_compatible_pad): Docs, and work with non-request compatible templates. (gst_element_get_compatible_pad_filtered): Docs and remove redundant checks. (gst_element_get_compatible_pad, gst_element_link_pads_filtered) (gst_element_link_filtered, gst_element_link_many) (gst_element_link, gst_element_link_pads) (gst_element_unlink_many): Docs. 2004-02-05 Andy Wingo <wingo@pobox.com> * gst/gstpad.c (_gst_real_pad_fixate_accumulator): s/pointer/boxed/. * gst/gstmarshal.list (VOID:BOXED, BOXED:BOXED): New marshallers. * gst/gstpad.c (gst_real_pad_class_init): Use a BOXED:BOXED marshaller for ::fixate, and VOID:BOXED for ::caps-nego-failed, with the type=GST_TYPE_CAPS. This allows language bindings to know what kind of data they're dealing with. * gst/gstcaps.c (_gst_caps_value_init): GBoxed values initialize to NULL when g_value_init is called. GstCaps, which rolls its own type implementation, now does the same instead of allocating empty caps. (_gst_caps_initialize, _gst_caps_collect_value, _gst_caps_lcopy_value): Provide collect_value and lcopy_value type table methods. This allows G_VALUE_COLLECT to work. 2004-02-05 Andy Wingo <wingo@pobox.com> * configure.ac: * testsuite/Makefile.am (SUBDIRS): * testsuite/ghostpads/Makefile.am: * testsuite/ghostpads/ghostpads.c: A new test for ghost pads. * gst/gstpad.c (gst_pad_add_ghost_pad, gst_pad_remove_ghost_pad): These two routines are the only ones that set GST_GPAD_REALPAD(gpad), the ghost pad list, and the ghost pad's pad template. They should be made static, depending on ABI needs. (gst_real_pad_dispose): Handle the case of ghost pads without a parent. Assert after dealing with ghost pads that the ghost pad list is empty. (gst_ghost_pad_class_init): New property added, ::real-pad. Can be set after creation. (gst_ghost_pad_dispose): Set ::real-pad to NULL. (gst_ghost_pad_set_property, gst_ghost_pad_get_property): New functions. set_property will call add_ghost_pad/remove_ghost_pad as appropriate. (gst_ghost_pad_new): All the work is offloaded to g_object_new. * gst/gstelement.c (gst_element_add_pad): Handle ghost pads as well. (gst_element_add_ghost_pad): Remove code duplicated from _add_pad. (gst_element_remove_pad): Handle ghost pads as well. (gst_element_remove_ghost_pad): Deprecated (could be removed, depending on API-stability needs). 2004-02-05 Andy Wingo <wingo@pobox.com> * gst/gstbin.[ch]: (gst_bin_get_by_interface): GTypes are scalars, of course they're const
2004-02-11 13:26:04 +00:00
*/
GstCaps *
gst_pad_get_negotiated_caps (GstPad * pad)
{
GstCaps *caps;
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
GstPad *peer;
g_return_val_if_fail (GST_IS_PAD (pad), NULL);
GST_OBJECT_LOCK (pad);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
if (G_UNLIKELY ((peer = GST_PAD_PEER (pad)) == NULL))
goto no_peer;
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_DEBUG_OBJECT (GST_CAT_PROPERTIES, pad, "getting negotiated caps");
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
caps = GST_PAD_CAPS (pad);
if (caps)
gst_caps_ref (caps);
GST_OBJECT_UNLOCK (pad);
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_DEBUG_OBJECT (GST_CAT_CAPS, pad, "negotiated caps %" GST_PTR_FORMAT,
caps);
return caps;
no_peer:
{
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_DEBUG_OBJECT (GST_CAT_PROPERTIES, pad, "no peer");
GST_OBJECT_UNLOCK (pad);
return NULL;
}
}
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
/* calls the buffer_alloc function on the given pad */
static GstFlowReturn
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
gst_pad_buffer_alloc_unchecked (GstPad * pad, guint64 offset, gint size,
GstCaps * caps, GstBuffer ** buf)
{
GstFlowReturn ret;
GstPadBufferAllocFunction bufferallocfunc;
GST_OBJECT_LOCK (pad);
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
/* when the pad is flushing we cannot give a buffer */
if (G_UNLIKELY (GST_PAD_IS_FLUSHING (pad)))
goto flushing;
bufferallocfunc = pad->bufferallocfunc;
if (offset == GST_BUFFER_OFFSET_NONE) {
GST_CAT_DEBUG_OBJECT (GST_CAT_PADS, pad,
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
"calling bufferallocfunc &%s (@%p) for size %d offset NONE",
GST_DEBUG_FUNCPTR_NAME (bufferallocfunc), bufferallocfunc, size);
} else {
GST_CAT_DEBUG_OBJECT (GST_CAT_PADS, pad,
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
"calling bufferallocfunc &%s (@%p) of for size %d offset %"
G_GUINT64_FORMAT, GST_DEBUG_FUNCPTR_NAME (bufferallocfunc),
bufferallocfunc, size, offset);
}
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_OBJECT_UNLOCK (pad);
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
/* G_LIKELY for now since most elements don't implement a buffer alloc
* function and there is no default alloc proxy function as this is usually
* not possible. */
if (G_LIKELY (bufferallocfunc == NULL))
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
goto fallback;
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
ret = bufferallocfunc (pad, offset, size, caps, buf);
if (G_UNLIKELY (ret != GST_FLOW_OK))
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
goto error;
/* no error, but NULL buffer means fallback to the default */
if (G_UNLIKELY (*buf == NULL))
goto fallback;
/* If the buffer alloc function didn't set up the caps like it should,
* do it for it */
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
if (G_UNLIKELY (caps && (GST_BUFFER_CAPS (*buf) == NULL))) {
GST_WARNING_OBJECT (pad,
"Buffer allocation function did not set caps. Setting");
gst_buffer_set_caps (*buf, caps);
}
return ret;
flushing:
{
/* pad was flushing */
GST_OBJECT_UNLOCK (pad);
GST_CAT_DEBUG_OBJECT (GST_CAT_PADS, pad, "pad was flushing");
return GST_FLOW_WRONG_STATE;
}
error:
{
GST_CAT_DEBUG_OBJECT (GST_CAT_PADS, pad,
"alloc function returned error (%d) %s", ret, gst_flow_get_name (ret));
return ret;
}
fallback:
{
/* fallback case, allocate a buffer of our own, add pad caps. */
GST_CAT_DEBUG_OBJECT (GST_CAT_PADS, pad, "fallback buffer alloc");
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
*buf = gst_buffer_new_and_alloc (size);
GST_BUFFER_OFFSET (*buf) = offset;
gst_buffer_set_caps (*buf, caps);
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
return GST_FLOW_OK;
}
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
}
static GstFlowReturn
gst_pad_alloc_buffer_full (GstPad * pad, guint64 offset, gint size,
GstCaps * caps, GstBuffer ** buf, gboolean setcaps)
{
GstPad *peer;
GstFlowReturn ret;
gboolean caps_changed;
g_return_val_if_fail (GST_IS_PAD (pad), GST_FLOW_ERROR);
g_return_val_if_fail (GST_PAD_IS_SRC (pad), GST_FLOW_ERROR);
g_return_val_if_fail (buf != NULL, GST_FLOW_ERROR);
GST_OBJECT_LOCK (pad);
while (G_UNLIKELY (GST_PAD_IS_BLOCKED (pad)))
if ((ret = handle_pad_block (pad)) != GST_FLOW_OK)
goto flushed;
if (G_UNLIKELY ((peer = GST_PAD_PEER (pad)) == NULL))
goto no_peer;
gst_object_ref (peer);
GST_OBJECT_UNLOCK (pad);
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
ret = gst_pad_buffer_alloc_unchecked (peer, offset, size, caps, buf);
gst_object_unref (peer);
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
if (G_UNLIKELY (ret != GST_FLOW_OK))
goto peer_error;
/* FIXME, move capnego this into a base class? */
caps = GST_BUFFER_CAPS (*buf);
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
/* Lock for checking caps, pretty pointless as the _pad_push() function might
* change it concurrently, one of the problems with automatic caps setting in
* pad_alloc_and_set_caps. Worst case, if does a check too much, but only when
* there is heavy renegotiation going on in both directions. */
GST_OBJECT_LOCK (pad);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
caps_changed = caps && caps != GST_PAD_CAPS (pad);
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_OBJECT_UNLOCK (pad);
/* we got a new datatype on the pad, see if it can handle it */
if (G_UNLIKELY (caps_changed)) {
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_DEBUG_OBJECT (pad, "caps changed to %p %" GST_PTR_FORMAT, caps, caps);
if (G_UNLIKELY (!gst_pad_configure_src (pad, caps, setcaps)))
goto not_negotiated;
}
return ret;
flushed:
{
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_DEBUG_OBJECT (GST_CAT_PADS, pad, "pad block stopped by flush");
GST_OBJECT_UNLOCK (pad);
return ret;
}
no_peer:
{
/* pad has no peer */
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_DEBUG_OBJECT (GST_CAT_PADS, pad,
"called bufferallocfunc but had no peer");
GST_OBJECT_UNLOCK (pad);
return GST_FLOW_NOT_LINKED;
}
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
peer_error:
{
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_LOG_OBJECT (GST_CAT_SCHEDULING, pad,
"alloc function returned error %s", gst_flow_get_name (ret));
return ret;
}
not_negotiated:
{
gst_buffer_unref (*buf);
*buf = NULL;
GST_CAT_LOG_OBJECT (GST_CAT_SCHEDULING, pad,
"alloc function returned unacceptable buffer");
return GST_FLOW_NOT_NEGOTIATED;
}
}
/**
* gst_pad_alloc_buffer:
* @pad: a source #GstPad
* @offset: the offset of the new buffer in the stream
* @size: the size of the new buffer
* @caps: the caps of the new buffer
* @buf: a newly allocated buffer
*
* Allocates a new, empty buffer optimized to push to pad @pad. This
* function only works if @pad is a source pad and has a peer.
*
* A new, empty #GstBuffer will be put in the @buf argument.
* You need to check the caps of the buffer after performing this
* function and renegotiate to the format if needed.
*
* Returns: a result code indicating success of the operation. Any
* result code other than #GST_FLOW_OK is an error and @buf should
* not be used.
* An error can occur if the pad is not connected or when the downstream
* peer elements cannot provide an acceptable buffer.
*
* MT safe.
*/
GstFlowReturn
gst_pad_alloc_buffer (GstPad * pad, guint64 offset, gint size, GstCaps * caps,
GstBuffer ** buf)
{
return gst_pad_alloc_buffer_full (pad, offset, size, caps, buf, FALSE);
}
/**
* gst_pad_alloc_buffer_and_set_caps:
* @pad: a source #GstPad
* @offset: the offset of the new buffer in the stream
* @size: the size of the new buffer
* @caps: the caps of the new buffer
* @buf: a newly allocated buffer
*
* In addition to the function gst_pad_alloc_buffer(), this function
* automatically calls gst_pad_set_caps() when the caps of the
* newly allocated buffer are different from the @pad caps.
*
* Returns: a result code indicating success of the operation. Any
* result code other than #GST_FLOW_OK is an error and @buf should
* not be used.
* An error can occur if the pad is not connected or when the downstream
* peer elements cannot provide an acceptable buffer.
*
* MT safe.
*/
GstFlowReturn
gst_pad_alloc_buffer_and_set_caps (GstPad * pad, guint64 offset, gint size,
GstCaps * caps, GstBuffer ** buf)
{
return gst_pad_alloc_buffer_full (pad, offset, size, caps, buf, TRUE);
}
/**
* gst_pad_get_internal_links_default:
* @pad: the #GstPad to get the internal links of.
*
* Gets a list of pads to which the given pad is linked to
* inside of the parent element.
* This is the default handler, and thus returns a list of all of the
* pads inside the parent element with opposite direction.
* The caller must free this list after use.
*
* Returns: a newly allocated #GList of pads, or NULL if the pad has no parent.
*
* Not MT safe.
*/
GList *
gst_pad_get_internal_links_default (GstPad * pad)
{
GList *res = NULL;
GstElement *parent;
GList *parent_pads;
GstPadDirection direction;
g_return_val_if_fail (GST_IS_PAD (pad), NULL);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
direction = pad->direction;
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
parent = GST_PAD_PARENT (pad);
if (!parent)
return NULL;
parent_pads = parent->pads;
while (parent_pads) {
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
GstPad *parent_pad = GST_PAD_CAST (parent_pads->data);
if (parent_pad->direction != direction) {
res = g_list_prepend (res, parent_pad);
}
parent_pads = g_list_next (parent_pads);
}
return res;
}
/**
* gst_pad_get_internal_links:
* @pad: the #GstPad to get the internal links of.
*
* Gets a list of pads to which the given pad is linked to
* inside of the parent element.
* The caller must free this list after use.
*
* Returns: a newly allocated #GList of pads.
*
* Not MT safe.
*/
GList *
gst_pad_get_internal_links (GstPad * pad)
{
GList *res = NULL;
g_return_val_if_fail (GST_IS_PAD (pad), NULL);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
if (GST_PAD_INTLINKFUNC (pad))
res = GST_PAD_INTLINKFUNC (pad) (pad);
return res;
}
static gboolean
gst_pad_event_default_dispatch (GstPad * pad, GstEvent * event)
{
GList *orig, *pads;
gboolean result;
GST_INFO_OBJECT (pad, "Sending event %p (%s) to all internally linked pads",
event, GST_EVENT_TYPE_NAME (event));
result = (GST_PAD_DIRECTION (pad) == GST_PAD_SINK);
orig = pads = gst_pad_get_internal_links (pad);
while (pads) {
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
GstPad *eventpad = GST_PAD_CAST (pads->data);
pads = g_list_next (pads);
if (GST_PAD_DIRECTION (eventpad) == GST_PAD_SRC) {
/* for each pad we send to, we should ref the event; it's up
* to downstream to unref again when handled. */
GST_LOG_OBJECT (pad, "Reffing and sending event %p (%s) to %s:%s",
event, GST_EVENT_TYPE_NAME (event), GST_DEBUG_PAD_NAME (eventpad));
gst_event_ref (event);
gst_pad_push_event (eventpad, event);
} else {
/* we only send the event on one pad, multi-sinkpad elements
* should implement a handler */
GST_LOG_OBJECT (pad, "sending event %p (%s) to one sink pad %s:%s",
event, GST_EVENT_TYPE_NAME (event), GST_DEBUG_PAD_NAME (eventpad));
result = gst_pad_push_event (eventpad, event);
goto done;
}
}
/* we handled the incoming event so we unref once */
GST_LOG_OBJECT (pad, "handled event %p, unreffing", event);
gst_event_unref (event);
done:
g_list_free (orig);
return result;
}
/**
* gst_pad_event_default:
* @pad: a #GstPad to call the default event handler on.
* @event: the #GstEvent to handle.
*
* Invokes the default event handler for the given pad. End-of-stream and
* discontinuity events are handled specially, and then the event is sent to all
* pads internally linked to @pad. Note that if there are many possible sink
* pads that are internally linked to @pad, only one will be sent an event.
* Multi-sinkpad elements should implement custom event handlers.
*
* Returns: TRUE if the event was sent succesfully.
*/
gboolean
gst_pad_event_default (GstPad * pad, GstEvent * event)
{
g_return_val_if_fail (GST_IS_PAD (pad), FALSE);
g_return_val_if_fail (event != NULL, FALSE);
switch (GST_EVENT_TYPE (event)) {
case GST_EVENT_EOS:
{
GST_DEBUG_OBJECT (pad, "pausing task because of eos");
gst_pad_pause_task (pad);
}
default:
break;
}
return gst_pad_event_default_dispatch (pad, event);
}
/**
* gst_pad_dispatcher:
* @pad: a #GstPad to dispatch.
* @dispatch: the #GstDispatcherFunction to call.
* @data: gpointer user data passed to the dispatcher function.
*
* Invokes the given dispatcher function on each respective peer of
* all pads that are internally linked to the given pad.
* The GstPadDispatcherFunction should return TRUE when no further pads
* need to be processed.
*
* Returns: TRUE if one of the dispatcher functions returned TRUE.
*/
gboolean
gst_pad_dispatcher (GstPad * pad, GstPadDispatcherFunction dispatch,
gpointer data)
{
gboolean res = FALSE;
GList *int_pads, *orig;
g_return_val_if_fail (GST_IS_PAD (pad), FALSE);
g_return_val_if_fail (dispatch != NULL, FALSE);
This is an attempt at not segfaulting on errors but reporting some usefull info instead. Original commit message from CVS: This is an attempt at not segfaulting on errors but reporting some usefull info instead. - bin changes so errors can propagate. - changed the _FAST macros to _CAST because that is what they do. - removed all references to cothreads out of the core, they are really a scheduler issue, handler with a sched_private gpointer. - added a live buffer count, for debugging buffer leaks. - added error checking in gst_scheduler_state_transition this solves the "out of cothreads" problem. - GST_ELEMENT_NO_ENTRY == GST_ELEMENT_INFINITE_LOOP - added 2 private element flasg for use by the scheduler (_COTHREAD_STOPPING) is now - added scheduler entry points: - _yield : to create possible scheduling points. - _interrupt: to stop execution of an element. - _error: to signal en error condition to the scheduler. - improved error messages for pads. - signal gst_element_error where appropriate. - added the a new bin to the parent before entering it so one can reference its children. - queue memleak fixes on dispose. - added possible deadlock detection in queue (turned off be default) - GstBasicScheduler is a real class of its own now, hiding its internal variables. - GST_ELEMENT_IS_COTHREAD_STOPPING is gone. either call explicit _yield operations, or make a sane loop. - Better state change handling in filesrc. Better error reporting/recovery too. - updated core plugins. - detect non decoupled elements on scheduler boundries and error.
2001-12-22 21:18:17 +00:00
orig = int_pads = gst_pad_get_internal_links (pad);
while (int_pads) {
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
GstPad *int_pad = GST_PAD_CAST (int_pads->data);
GstPad *int_peer = GST_PAD_PEER (int_pad);
if (int_peer) {
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
res = dispatch (int_peer, data);
if (res)
break;
}
int_pads = g_list_next (int_pads);
}
g_list_free (orig);
return res;
}
/**
* gst_pad_query:
* @pad: a #GstPad to invoke the default query on.
* @query: the #GstQuery to perform.
*
* Dispatches a query to a pad. The query should have been allocated by the
* caller via one of the type-specific allocation functions in gstquery.h. The
* element is responsible for filling the query with an appropriate response,
* which should then be parsed with a type-specific query parsing function.
*
* Again, the caller is responsible for both the allocation and deallocation of
* the query structure.
*
* Returns: TRUE if the query could be performed.
*/
gboolean
gst_pad_query (GstPad * pad, GstQuery * query)
{
GstPadQueryFunction func;
g_return_val_if_fail (GST_IS_PAD (pad), FALSE);
g_return_val_if_fail (GST_IS_QUERY (query), FALSE);
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_DEBUG_OBJECT (pad, "sending query %p", query);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
if ((func = GST_PAD_QUERYFUNC (pad)) == NULL)
goto no_func;
2004-04-04 02:36:12 +00:00
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
return func (pad, query);
2004-04-04 02:36:12 +00:00
no_func:
{
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_DEBUG_OBJECT (pad, "had no query function");
return FALSE;
2004-04-04 02:36:12 +00:00
}
}
/**
* gst_pad_query_default:
* @pad: a #GstPad to call the default query handler on.
* @query: the #GstQuery to handle.
*
* Invokes the default query handler for the given pad.
* The query is sent to all pads internally linked to @pad. Note that
* if there are many possible sink pads that are internally linked to
* @pad, only one will be sent the query.
* Multi-sinkpad elements should implement custom query handlers.
*
* Returns: TRUE if the query was performed succesfully.
*/
gst/: Simplify pad activation. Original commit message from CVS: * gst/base/Makefile.am: * gst/base/README: * gst/base/gstbasesink.c: (gst_basesink_get_type), (gst_basesink_base_init), (gst_basesink_class_init), (gst_basesink_pad_getcaps), (gst_basesink_init), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesink.h: * gst/base/gstbasetransform.c: (gst_base_transform_get_type), (gst_base_transform_base_init), (gst_base_transform_finalize), (gst_base_transform_class_init), (gst_base_transform_init), (gst_base_transform_proxy_getcaps), (gst_base_transform_setcaps), (gst_base_transform_event), (gst_base_transform_getrange), (gst_base_transform_chain), (gst_base_transform_handle_buffer), (gst_base_transform_set_property), (gst_base_transform_get_property), (gst_base_transform_sink_activate), (gst_base_transform_src_activate), (gst_base_transform_change_state): * gst/base/gstbasetransform.h: * gst/elements/gstidentity.c: (gst_identity_finalize), (gst_identity_class_init), (gst_identity_init), (gst_identity_event), (gst_identity_check_perfect), (gst_identity_transform), (gst_identity_set_property), (gst_identity_get_property), (gst_identity_change_state): * gst/elements/gstidentity.h: * gst/gstelement.c: (gst_element_get_state_func), (gst_element_lost_state), (gst_element_pads_activate): * gst/gstpad.c: (gst_pad_set_active), (gst_pad_peer_set_active), (gst_pad_check_pull_range), (gst_pad_pull_range): * gst/gstpad.h: Simplify pad activation. Added function to check if pull_range can be performed. Error out when pulling inactive or flushing pads. Removed const from refcounted types as it does not make sense. Simplify pad templates in basesink Added base class for simple 1-to-1 transforms. Make identity subclass the base transform.
2005-03-29 16:18:12 +00:00
gboolean
gst_pad_query_default (GstPad * pad, GstQuery * query)
gst/: Simplify pad activation. Original commit message from CVS: * gst/base/Makefile.am: * gst/base/README: * gst/base/gstbasesink.c: (gst_basesink_get_type), (gst_basesink_base_init), (gst_basesink_class_init), (gst_basesink_pad_getcaps), (gst_basesink_init), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesink.h: * gst/base/gstbasetransform.c: (gst_base_transform_get_type), (gst_base_transform_base_init), (gst_base_transform_finalize), (gst_base_transform_class_init), (gst_base_transform_init), (gst_base_transform_proxy_getcaps), (gst_base_transform_setcaps), (gst_base_transform_event), (gst_base_transform_getrange), (gst_base_transform_chain), (gst_base_transform_handle_buffer), (gst_base_transform_set_property), (gst_base_transform_get_property), (gst_base_transform_sink_activate), (gst_base_transform_src_activate), (gst_base_transform_change_state): * gst/base/gstbasetransform.h: * gst/elements/gstidentity.c: (gst_identity_finalize), (gst_identity_class_init), (gst_identity_init), (gst_identity_event), (gst_identity_check_perfect), (gst_identity_transform), (gst_identity_set_property), (gst_identity_get_property), (gst_identity_change_state): * gst/elements/gstidentity.h: * gst/gstelement.c: (gst_element_get_state_func), (gst_element_lost_state), (gst_element_pads_activate): * gst/gstpad.c: (gst_pad_set_active), (gst_pad_peer_set_active), (gst_pad_check_pull_range), (gst_pad_pull_range): * gst/gstpad.h: Simplify pad activation. Added function to check if pull_range can be performed. Error out when pulling inactive or flushing pads. Removed const from refcounted types as it does not make sense. Simplify pad templates in basesink Added base class for simple 1-to-1 transforms. Make identity subclass the base transform.
2005-03-29 16:18:12 +00:00
{
switch (GST_QUERY_TYPE (query)) {
case GST_QUERY_POSITION:
case GST_QUERY_SEEKING:
case GST_QUERY_FORMATS:
case GST_QUERY_LATENCY:
case GST_QUERY_JITTER:
case GST_QUERY_RATE:
case GST_QUERY_CONVERT:
default:
return gst_pad_dispatcher
(pad, (GstPadDispatcherFunction) gst_pad_query, query);
gst/: Simplify pad activation. Original commit message from CVS: * gst/base/Makefile.am: * gst/base/README: * gst/base/gstbasesink.c: (gst_basesink_get_type), (gst_basesink_base_init), (gst_basesink_class_init), (gst_basesink_pad_getcaps), (gst_basesink_init), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesink.h: * gst/base/gstbasetransform.c: (gst_base_transform_get_type), (gst_base_transform_base_init), (gst_base_transform_finalize), (gst_base_transform_class_init), (gst_base_transform_init), (gst_base_transform_proxy_getcaps), (gst_base_transform_setcaps), (gst_base_transform_event), (gst_base_transform_getrange), (gst_base_transform_chain), (gst_base_transform_handle_buffer), (gst_base_transform_set_property), (gst_base_transform_get_property), (gst_base_transform_sink_activate), (gst_base_transform_src_activate), (gst_base_transform_change_state): * gst/base/gstbasetransform.h: * gst/elements/gstidentity.c: (gst_identity_finalize), (gst_identity_class_init), (gst_identity_init), (gst_identity_event), (gst_identity_check_perfect), (gst_identity_transform), (gst_identity_set_property), (gst_identity_get_property), (gst_identity_change_state): * gst/elements/gstidentity.h: * gst/gstelement.c: (gst_element_get_state_func), (gst_element_lost_state), (gst_element_pads_activate): * gst/gstpad.c: (gst_pad_set_active), (gst_pad_peer_set_active), (gst_pad_check_pull_range), (gst_pad_pull_range): * gst/gstpad.h: Simplify pad activation. Added function to check if pull_range can be performed. Error out when pulling inactive or flushing pads. Removed const from refcounted types as it does not make sense. Simplify pad templates in basesink Added base class for simple 1-to-1 transforms. Make identity subclass the base transform.
2005-03-29 16:18:12 +00:00
}
}
#ifndef GST_DISABLE_LOADSAVE
/* FIXME: why isn't this on a GstElement ? */
/**
* gst_pad_load_and_link:
* @self: an #xmlNodePtr to read the description from.
* @parent: the #GstObject element that owns the pad.
*
* Reads the pad definition from the XML node and links the given pad
* in the element to a pad of an element up in the hierarchy.
*/
void
gst_pad_load_and_link (xmlNodePtr self, GstObject * parent)
{
xmlNodePtr field = self->xmlChildrenNode;
GstPad *pad = NULL, *targetpad;
gchar *peer = NULL;
gchar **split;
GstElement *target;
GstObject *grandparent;
gchar *name = NULL;
while (field) {
if (!strcmp ((char *) field->name, "name")) {
name = (gchar *) xmlNodeGetContent (field);
pad = gst_element_get_pad (GST_ELEMENT (parent), name);
g_free (name);
} else if (!strcmp ((char *) field->name, "peer")) {
peer = (gchar *) xmlNodeGetContent (field);
}
field = field->next;
}
g_return_if_fail (pad != NULL);
if (peer == NULL)
return;
split = g_strsplit (peer, ".", 2);
if (split[0] == NULL || split[1] == NULL) {
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_DEBUG_OBJECT (GST_CAT_XML, pad,
"Could not parse peer '%s', leaving unlinked", peer);
g_free (peer);
return;
}
g_free (peer);
g_return_if_fail (split[0] != NULL);
g_return_if_fail (split[1] != NULL);
grandparent = gst_object_get_parent (parent);
if (grandparent && GST_IS_BIN (grandparent)) {
target = gst_bin_get_by_name_recurse_up (GST_BIN (grandparent), split[0]);
} else
goto cleanup;
if (target == NULL)
goto cleanup;
targetpad = gst_element_get_pad (target, split[1]);
if (targetpad == NULL)
goto cleanup;
gst_pad_link (pad, targetpad);
cleanup:
g_strfreev (split);
}
/**
* gst_pad_save_thyself:
* @pad: a #GstPad to save.
* @parent: the parent #xmlNodePtr to save the description in.
*
* Saves the pad into an xml representation.
*
* Returns: the #xmlNodePtr representation of the pad.
*/
static xmlNodePtr
gst_pad_save_thyself (GstObject * object, xmlNodePtr parent)
{
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
GstPad *pad;
GstPad *peer;
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
g_return_val_if_fail (GST_IS_PAD (object), NULL);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
pad = GST_PAD (object);
xmlNewChild (parent, NULL, (xmlChar *) "name",
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
(xmlChar *) GST_PAD_NAME (pad));
if (GST_PAD_IS_SRC (pad)) {
xmlNewChild (parent, NULL, (xmlChar *) "direction", (xmlChar *) "source");
} else if (GST_PAD_IS_SINK (pad)) {
xmlNewChild (parent, NULL, (xmlChar *) "direction", (xmlChar *) "sink");
} else {
xmlNewChild (parent, NULL, (xmlChar *) "direction", (xmlChar *) "unknown");
}
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
if (GST_PAD_PEER (pad) != NULL) {
gchar *content;
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
peer = GST_PAD_PEER (pad);
/* first check to see if the peer's parent's parent is the same */
/* we just save it off */
content = g_strdup_printf ("%s.%s",
GST_OBJECT_NAME (GST_PAD_PARENT (peer)), GST_PAD_NAME (peer));
xmlNewChild (parent, NULL, (xmlChar *) "peer", (xmlChar *) content);
g_free (content);
} else
xmlNewChild (parent, NULL, (xmlChar *) "peer", NULL);
return parent;
}
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
#if 0
/**
* gst_ghost_pad_save_thyself:
* @pad: a ghost #GstPad to save.
* @parent: the parent #xmlNodePtr to save the description in.
*
* Saves the ghost pad into an xml representation.
*
* Returns: the #xmlNodePtr representation of the pad.
*/
xmlNodePtr
gst_ghost_pad_save_thyself (GstPad * pad, xmlNodePtr parent)
{
xmlNodePtr self;
g_return_val_if_fail (GST_IS_GHOST_PAD (pad), NULL);
self = xmlNewChild (parent, NULL, (xmlChar *) "ghostpad", NULL);
xmlNewChild (self, NULL, (xmlChar *) "name", (xmlChar *) GST_PAD_NAME (pad));
xmlNewChild (self, NULL, (xmlChar *) "parent",
(xmlChar *) GST_OBJECT_NAME (GST_PAD_PARENT (pad)));
gst/gstpad.c (_gst_real_pad_fixate_accumulator): s/pointer/boxed/. Original commit message from CVS: 2004-02-05 Andy Wingo <wingo@pobox.com> * gst/gstpad.c (_gst_real_pad_fixate_accumulator): s/pointer/boxed/. * gst/gstmarshal.list (VOID:BOXED, BOXED:BOXED): New marshallers. * gst/gstpad.c (gst_real_pad_class_init): Use a BOXED:BOXED marshaller for ::fixate, and VOID:BOXED for ::caps-nego-failed, with the type=GST_TYPE_CAPS. This allows language bindings to know what kind of data they're dealing with. * gst/gstcaps.c (_gst_caps_value_init): GBoxed values initialize to NULL when g_value_init is called. GstCaps, which rolls its own type implementation, now does the same instead of allocating empty caps. (_gst_caps_initialize, _gst_caps_collect_value, _gst_caps_lcopy_value): Provide collect_value and lcopy_value type table methods. This allows G_VALUE_COLLECT to work. 2004-02-05 Andy Wingo <wingo@pobox.com> * configure.ac: * testsuite/Makefile.am (SUBDIRS): * testsuite/ghostpads/Makefile.am: * testsuite/ghostpads/ghostpads.c: A new test for ghost pads. * gst/gstpad.c (gst_pad_add_ghost_pad, gst_pad_remove_ghost_pad): These two routines are the only ones that set GST_GPAD_REALPAD(gpad), the ghost pad list, and the ghost pad's pad template. They should be made static, depending on ABI needs. (gst_real_pad_dispose): Handle the case of ghost pads without a parent. Assert after dealing with ghost pads that the ghost pad list is empty. (gst_ghost_pad_class_init): New property added, ::real-pad. Can be set after creation. (gst_ghost_pad_dispose): Set ::real-pad to NULL. (gst_ghost_pad_set_property, gst_ghost_pad_get_property): New functions. set_property will call add_ghost_pad/remove_ghost_pad as appropriate. (gst_ghost_pad_new): All the work is offloaded to g_object_new. * gst/gstelement.c (gst_element_add_pad): Handle ghost pads as well. (gst_element_add_ghost_pad): Remove code duplicated from _add_pad. (gst_element_remove_pad): Handle ghost pads as well. (gst_element_remove_ghost_pad): Deprecated (could be removed, depending on API-stability needs). 2004-02-05 Andy Wingo <wingo@pobox.com> * gst/gstbin.[ch]: (gst_bin_get_by_interface): GTypes are scalars, of course they're const
2004-02-05 13:51:26 +00:00
/* FIXME FIXME FIXME! */
return self;
}
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
#endif /* 0 */
#endif /* GST_DISABLE_LOADSAVE */
gst/gstpad.c (_gst_real_pad_fixate_accumulator): s/pointer/boxed/. Original commit message from CVS: 2004-02-05 Andy Wingo <wingo@pobox.com> * gst/gstpad.c (_gst_real_pad_fixate_accumulator): s/pointer/boxed/. * gst/gstmarshal.list (VOID:BOXED, BOXED:BOXED): New marshallers. * gst/gstpad.c (gst_real_pad_class_init): Use a BOXED:BOXED marshaller for ::fixate, and VOID:BOXED for ::caps-nego-failed, with the type=GST_TYPE_CAPS. This allows language bindings to know what kind of data they're dealing with. * gst/gstcaps.c (_gst_caps_value_init): GBoxed values initialize to NULL when g_value_init is called. GstCaps, which rolls its own type implementation, now does the same instead of allocating empty caps. (_gst_caps_initialize, _gst_caps_collect_value, _gst_caps_lcopy_value): Provide collect_value and lcopy_value type table methods. This allows G_VALUE_COLLECT to work. 2004-02-05 Andy Wingo <wingo@pobox.com> * configure.ac: * testsuite/Makefile.am (SUBDIRS): * testsuite/ghostpads/Makefile.am: * testsuite/ghostpads/ghostpads.c: A new test for ghost pads. * gst/gstpad.c (gst_pad_add_ghost_pad, gst_pad_remove_ghost_pad): These two routines are the only ones that set GST_GPAD_REALPAD(gpad), the ghost pad list, and the ghost pad's pad template. They should be made static, depending on ABI needs. (gst_real_pad_dispose): Handle the case of ghost pads without a parent. Assert after dealing with ghost pads that the ghost pad list is empty. (gst_ghost_pad_class_init): New property added, ::real-pad. Can be set after creation. (gst_ghost_pad_dispose): Set ::real-pad to NULL. (gst_ghost_pad_set_property, gst_ghost_pad_get_property): New functions. set_property will call add_ghost_pad/remove_ghost_pad as appropriate. (gst_ghost_pad_new): All the work is offloaded to g_object_new. * gst/gstelement.c (gst_element_add_pad): Handle ghost pads as well. (gst_element_add_ghost_pad): Remove code duplicated from _add_pad. (gst_element_remove_pad): Handle ghost pads as well. (gst_element_remove_ghost_pad): Deprecated (could be removed, depending on API-stability needs). 2004-02-05 Andy Wingo <wingo@pobox.com> * gst/gstbin.[ch]: (gst_bin_get_by_interface): GTypes are scalars, of course they're const
2004-02-05 13:51:26 +00:00
/*
* should be called with pad OBJECT_LOCK and STREAM_LOCK held.
* GST_PAD_IS_BLOCKED (pad) == TRUE when this function is
* called.
*
* This function perform the pad blocking when an event, buffer push
* or buffer_alloc is performed on a _SRC_ pad. It blocks the
* streaming thread after informing the pad has been blocked.
*
* An application can with this method wait and block any streaming
* thread and perform operations such as seeking or linking.
*
* Two methods are available for notifying the application of the
* block:
* - the callback method, which happens in the STREAMING thread with
* the STREAM_LOCK held. With this method, the most usefull way of
* dealing with the callback is to post a message to the main thread
* where the pad block can then be handled outside of the streaming
* thread. With the last method one can perform all operations such
* as doing a state change, linking, unblocking, seeking etc on the
* pad.
* - the GCond signal method, which makes any thread unblock when
* the pad block happens.
*
* During the actual blocking state, the GST_PAD_BLOCKING flag is set.
* The GST_PAD_BLOCKING flag is unset when the pad was unblocked.
*
* MT safe.
*/
static GstFlowReturn
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
handle_pad_block (GstPad * pad)
{
GstPadBlockCallback callback;
gpointer user_data;
GstFlowReturn ret = GST_FLOW_OK;
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_LOG_OBJECT (GST_CAT_SCHEDULING, pad, "signal block taken");
/* flushing, don't bother trying to block and return WRONG_STATE
* right away */
if (GST_PAD_IS_FLUSHING (pad))
goto flushingnonref;
/* we grab an extra ref for the callbacks, this is probably not
* needed (callback code does not have a ref and cannot unref). I
* think this was done to make it possible to unref the element in
* the callback, which is in the end totally impossible as it
* requires grabbing the STREAM_LOCK and OBJECT_LOCK which are
* all taken when calling this function. */
gst_object_ref (pad);
/* we either have a callback installed to notify the block or
* some other thread is doing a GCond wait. */
callback = pad->block_callback;
if (callback) {
/* there is a callback installed, call it. We release the
* lock so that the callback can do something usefull with the
* pad */
user_data = pad->block_data;
GST_OBJECT_UNLOCK (pad);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
callback (pad, TRUE, user_data);
GST_OBJECT_LOCK (pad);
/* we released the lock, recheck flushing */
if (GST_PAD_IS_FLUSHING (pad))
goto flushing;
} else {
/* no callback, signal the thread that is doing a GCond wait
* if any. */
GST_PAD_BLOCK_BROADCAST (pad);
}
/* OBJECT_LOCK could have been released when we did the callback, which
* then could have made the pad unblock so we need to check the blocking
* condition again. */
while (GST_PAD_IS_BLOCKED (pad)) {
/* now we block the streaming thread. It can be unlocked when we
* deactivate the pad (which will also set the FLUSHING flag) or
* when the pad is unblocked. A flushing event will also unblock
* the pad after setting the FLUSHING flag. */
GST_CAT_LOG_OBJECT (GST_CAT_SCHEDULING, pad,
"Waiting to be unblocked or set flushing");
GST_OBJECT_FLAG_SET (pad, GST_PAD_BLOCKING);
GST_PAD_BLOCK_WAIT (pad);
GST_OBJECT_FLAG_UNSET (pad, GST_PAD_BLOCKING);
/* see if we got unblocked by a flush or not */
if (GST_PAD_IS_FLUSHING (pad))
goto flushing;
}
GST_CAT_LOG_OBJECT (GST_CAT_SCHEDULING, pad, "got unblocked");
gst/gstpad.c (_gst_real_pad_fixate_accumulator): s/pointer/boxed/. Original commit message from CVS: 2004-02-05 Andy Wingo <wingo@pobox.com> * gst/gstpad.c (_gst_real_pad_fixate_accumulator): s/pointer/boxed/. * gst/gstmarshal.list (VOID:BOXED, BOXED:BOXED): New marshallers. * gst/gstpad.c (gst_real_pad_class_init): Use a BOXED:BOXED marshaller for ::fixate, and VOID:BOXED for ::caps-nego-failed, with the type=GST_TYPE_CAPS. This allows language bindings to know what kind of data they're dealing with. * gst/gstcaps.c (_gst_caps_value_init): GBoxed values initialize to NULL when g_value_init is called. GstCaps, which rolls its own type implementation, now does the same instead of allocating empty caps. (_gst_caps_initialize, _gst_caps_collect_value, _gst_caps_lcopy_value): Provide collect_value and lcopy_value type table methods. This allows G_VALUE_COLLECT to work. 2004-02-05 Andy Wingo <wingo@pobox.com> * configure.ac: * testsuite/Makefile.am (SUBDIRS): * testsuite/ghostpads/Makefile.am: * testsuite/ghostpads/ghostpads.c: A new test for ghost pads. * gst/gstpad.c (gst_pad_add_ghost_pad, gst_pad_remove_ghost_pad): These two routines are the only ones that set GST_GPAD_REALPAD(gpad), the ghost pad list, and the ghost pad's pad template. They should be made static, depending on ABI needs. (gst_real_pad_dispose): Handle the case of ghost pads without a parent. Assert after dealing with ghost pads that the ghost pad list is empty. (gst_ghost_pad_class_init): New property added, ::real-pad. Can be set after creation. (gst_ghost_pad_dispose): Set ::real-pad to NULL. (gst_ghost_pad_set_property, gst_ghost_pad_get_property): New functions. set_property will call add_ghost_pad/remove_ghost_pad as appropriate. (gst_ghost_pad_new): All the work is offloaded to g_object_new. * gst/gstelement.c (gst_element_add_pad): Handle ghost pads as well. (gst_element_add_ghost_pad): Remove code duplicated from _add_pad. (gst_element_remove_pad): Handle ghost pads as well. (gst_element_remove_ghost_pad): Deprecated (could be removed, depending on API-stability needs). 2004-02-05 Andy Wingo <wingo@pobox.com> * gst/gstbin.[ch]: (gst_bin_get_by_interface): GTypes are scalars, of course they're const
2004-02-05 13:51:26 +00:00
/* when we get here, the pad is unblocked again and we perform
* the needed unblock code. */
callback = pad->block_callback;
if (callback) {
/* we need to call the callback */
user_data = pad->block_data;
GST_OBJECT_UNLOCK (pad);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
callback (pad, FALSE, user_data);
GST_OBJECT_LOCK (pad);
} else {
/* we need to signal the thread waiting on the GCond */
GST_PAD_BLOCK_BROADCAST (pad);
gst/gstpad.c (_gst_real_pad_fixate_accumulator): s/pointer/boxed/. Original commit message from CVS: 2004-02-05 Andy Wingo <wingo@pobox.com> * gst/gstpad.c (_gst_real_pad_fixate_accumulator): s/pointer/boxed/. * gst/gstmarshal.list (VOID:BOXED, BOXED:BOXED): New marshallers. * gst/gstpad.c (gst_real_pad_class_init): Use a BOXED:BOXED marshaller for ::fixate, and VOID:BOXED for ::caps-nego-failed, with the type=GST_TYPE_CAPS. This allows language bindings to know what kind of data they're dealing with. * gst/gstcaps.c (_gst_caps_value_init): GBoxed values initialize to NULL when g_value_init is called. GstCaps, which rolls its own type implementation, now does the same instead of allocating empty caps. (_gst_caps_initialize, _gst_caps_collect_value, _gst_caps_lcopy_value): Provide collect_value and lcopy_value type table methods. This allows G_VALUE_COLLECT to work. 2004-02-05 Andy Wingo <wingo@pobox.com> * configure.ac: * testsuite/Makefile.am (SUBDIRS): * testsuite/ghostpads/Makefile.am: * testsuite/ghostpads/ghostpads.c: A new test for ghost pads. * gst/gstpad.c (gst_pad_add_ghost_pad, gst_pad_remove_ghost_pad): These two routines are the only ones that set GST_GPAD_REALPAD(gpad), the ghost pad list, and the ghost pad's pad template. They should be made static, depending on ABI needs. (gst_real_pad_dispose): Handle the case of ghost pads without a parent. Assert after dealing with ghost pads that the ghost pad list is empty. (gst_ghost_pad_class_init): New property added, ::real-pad. Can be set after creation. (gst_ghost_pad_dispose): Set ::real-pad to NULL. (gst_ghost_pad_set_property, gst_ghost_pad_get_property): New functions. set_property will call add_ghost_pad/remove_ghost_pad as appropriate. (gst_ghost_pad_new): All the work is offloaded to g_object_new. * gst/gstelement.c (gst_element_add_pad): Handle ghost pads as well. (gst_element_add_ghost_pad): Remove code duplicated from _add_pad. (gst_element_remove_pad): Handle ghost pads as well. (gst_element_remove_ghost_pad): Deprecated (could be removed, depending on API-stability needs). 2004-02-05 Andy Wingo <wingo@pobox.com> * gst/gstbin.[ch]: (gst_bin_get_by_interface): GTypes are scalars, of course they're const
2004-02-05 13:51:26 +00:00
}
gst_object_unref (pad);
return ret;
flushingnonref:
{
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_LOG_OBJECT (GST_CAT_SCHEDULING, pad, "pad was flushing");
return GST_FLOW_WRONG_STATE;
}
flushing:
{
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_LOG_OBJECT (GST_CAT_SCHEDULING, pad, "pad became flushing");
gst_object_unref (pad);
return GST_FLOW_WRONG_STATE;
}
gst/gstpad.c (_gst_real_pad_fixate_accumulator): s/pointer/boxed/. Original commit message from CVS: 2004-02-05 Andy Wingo <wingo@pobox.com> * gst/gstpad.c (_gst_real_pad_fixate_accumulator): s/pointer/boxed/. * gst/gstmarshal.list (VOID:BOXED, BOXED:BOXED): New marshallers. * gst/gstpad.c (gst_real_pad_class_init): Use a BOXED:BOXED marshaller for ::fixate, and VOID:BOXED for ::caps-nego-failed, with the type=GST_TYPE_CAPS. This allows language bindings to know what kind of data they're dealing with. * gst/gstcaps.c (_gst_caps_value_init): GBoxed values initialize to NULL when g_value_init is called. GstCaps, which rolls its own type implementation, now does the same instead of allocating empty caps. (_gst_caps_initialize, _gst_caps_collect_value, _gst_caps_lcopy_value): Provide collect_value and lcopy_value type table methods. This allows G_VALUE_COLLECT to work. 2004-02-05 Andy Wingo <wingo@pobox.com> * configure.ac: * testsuite/Makefile.am (SUBDIRS): * testsuite/ghostpads/Makefile.am: * testsuite/ghostpads/ghostpads.c: A new test for ghost pads. * gst/gstpad.c (gst_pad_add_ghost_pad, gst_pad_remove_ghost_pad): These two routines are the only ones that set GST_GPAD_REALPAD(gpad), the ghost pad list, and the ghost pad's pad template. They should be made static, depending on ABI needs. (gst_real_pad_dispose): Handle the case of ghost pads without a parent. Assert after dealing with ghost pads that the ghost pad list is empty. (gst_ghost_pad_class_init): New property added, ::real-pad. Can be set after creation. (gst_ghost_pad_dispose): Set ::real-pad to NULL. (gst_ghost_pad_set_property, gst_ghost_pad_get_property): New functions. set_property will call add_ghost_pad/remove_ghost_pad as appropriate. (gst_ghost_pad_new): All the work is offloaded to g_object_new. * gst/gstelement.c (gst_element_add_pad): Handle ghost pads as well. (gst_element_add_ghost_pad): Remove code duplicated from _add_pad. (gst_element_remove_pad): Handle ghost pads as well. (gst_element_remove_ghost_pad): Deprecated (could be removed, depending on API-stability needs). 2004-02-05 Andy Wingo <wingo@pobox.com> * gst/gstbin.[ch]: (gst_bin_get_by_interface): GTypes are scalars, of course they're const
2004-02-05 13:51:26 +00:00
}
/**********************************************************************
* Data passing functions
*/
gst/gstpad.c (_gst_real_pad_fixate_accumulator): s/pointer/boxed/. Original commit message from CVS: 2004-02-05 Andy Wingo <wingo@pobox.com> * gst/gstpad.c (_gst_real_pad_fixate_accumulator): s/pointer/boxed/. * gst/gstmarshal.list (VOID:BOXED, BOXED:BOXED): New marshallers. * gst/gstpad.c (gst_real_pad_class_init): Use a BOXED:BOXED marshaller for ::fixate, and VOID:BOXED for ::caps-nego-failed, with the type=GST_TYPE_CAPS. This allows language bindings to know what kind of data they're dealing with. * gst/gstcaps.c (_gst_caps_value_init): GBoxed values initialize to NULL when g_value_init is called. GstCaps, which rolls its own type implementation, now does the same instead of allocating empty caps. (_gst_caps_initialize, _gst_caps_collect_value, _gst_caps_lcopy_value): Provide collect_value and lcopy_value type table methods. This allows G_VALUE_COLLECT to work. 2004-02-05 Andy Wingo <wingo@pobox.com> * configure.ac: * testsuite/Makefile.am (SUBDIRS): * testsuite/ghostpads/Makefile.am: * testsuite/ghostpads/ghostpads.c: A new test for ghost pads. * gst/gstpad.c (gst_pad_add_ghost_pad, gst_pad_remove_ghost_pad): These two routines are the only ones that set GST_GPAD_REALPAD(gpad), the ghost pad list, and the ghost pad's pad template. They should be made static, depending on ABI needs. (gst_real_pad_dispose): Handle the case of ghost pads without a parent. Assert after dealing with ghost pads that the ghost pad list is empty. (gst_ghost_pad_class_init): New property added, ::real-pad. Can be set after creation. (gst_ghost_pad_dispose): Set ::real-pad to NULL. (gst_ghost_pad_set_property, gst_ghost_pad_get_property): New functions. set_property will call add_ghost_pad/remove_ghost_pad as appropriate. (gst_ghost_pad_new): All the work is offloaded to g_object_new. * gst/gstelement.c (gst_element_add_pad): Handle ghost pads as well. (gst_element_add_ghost_pad): Remove code duplicated from _add_pad. (gst_element_remove_pad): Handle ghost pads as well. (gst_element_remove_ghost_pad): Deprecated (could be removed, depending on API-stability needs). 2004-02-05 Andy Wingo <wingo@pobox.com> * gst/gstbin.[ch]: (gst_bin_get_by_interface): GTypes are scalars, of course they're const
2004-02-05 13:51:26 +00:00
static gboolean
gst_pad_emit_have_data_signal (GstPad * pad, GstMiniObject * obj)
{
GValue ret = { 0 };
GValue args[2] = { {0}, {0} };
gboolean res;
GQuark detail;
/* init */
g_value_init (&ret, G_TYPE_BOOLEAN);
g_value_set_boolean (&ret, TRUE);
g_value_init (&args[0], GST_TYPE_PAD);
g_value_set_object (&args[0], pad);
g_value_init (&args[1], GST_TYPE_MINI_OBJECT); // G_TYPE_POINTER);
gst_value_set_mini_object (&args[1], obj);
if (GST_IS_EVENT (obj))
detail = event_quark;
else
detail = buffer_quark;
/* actually emit */
g_signal_emitv (args, gst_pad_signals[PAD_HAVE_DATA], detail, &ret);
res = g_value_get_boolean (&ret);
/* clean up */
g_value_unset (&ret);
g_value_unset (&args[0]);
g_value_unset (&args[1]);
return res;
}
/* this is the chain function that does not perform the additional argument
* checking for that little extra speed.
*/
static inline GstFlowReturn
gst_pad_chain_unchecked (GstPad * pad, GstBuffer * buffer)
{
GstCaps *caps;
gboolean caps_changed;
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
GstPadChainFunction chainfunc;
GstFlowReturn ret;
gboolean emit_signal;
GST_PAD_STREAM_LOCK (pad);
GST_OBJECT_LOCK (pad);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
if (G_UNLIKELY (GST_PAD_IS_FLUSHING (pad)))
goto flushing;
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
caps = GST_BUFFER_CAPS (buffer);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
caps_changed = caps && caps != GST_PAD_CAPS (pad);
emit_signal = GST_PAD_DO_BUFFER_SIGNALS (pad) > 0;
GST_OBJECT_UNLOCK (pad);
/* see if the signal should be emited, we emit before caps nego as
* we might drop the buffer and do capsnego for nothing. */
if (G_UNLIKELY (emit_signal)) {
if (!gst_pad_emit_have_data_signal (pad, GST_MINI_OBJECT (buffer)))
goto dropping;
}
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
/* we got a new datatype on the pad, see if it can handle it */
if (G_UNLIKELY (caps_changed)) {
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_DEBUG_OBJECT (pad, "caps changed to %p %" GST_PTR_FORMAT, caps, caps);
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
if (G_UNLIKELY (!gst_pad_configure_sink (pad, caps)))
goto not_negotiated;
}
/* NOTE: we read the chainfunc unlocked.
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
* we cannot hold the lock for the pad so we might send
* the data to the wrong function. This is not really a
* problem since functions are assigned at creation time
* and don't change that often... */
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
if (G_UNLIKELY ((chainfunc = GST_PAD_CHAINFUNC (pad)) == NULL))
goto no_function;
GST_CAT_LOG_OBJECT (GST_CAT_SCHEDULING, pad,
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
"calling chainfunction &%s", GST_DEBUG_FUNCPTR_NAME (chainfunc));
ret = chainfunc (pad, buffer);
GST_CAT_LOG_OBJECT (GST_CAT_SCHEDULING, pad,
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
"called chainfunction &%s, returned %s",
GST_DEBUG_FUNCPTR_NAME (chainfunc), gst_flow_get_name (ret));
GST_PAD_STREAM_UNLOCK (pad);
return ret;
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
/* ERRORS */
flushing:
{
gst_buffer_unref (buffer);
GST_CAT_LOG_OBJECT (GST_CAT_SCHEDULING, pad,
"pushing, but pad was flushing");
GST_OBJECT_UNLOCK (pad);
GST_PAD_STREAM_UNLOCK (pad);
return GST_FLOW_WRONG_STATE;
}
dropping:
{
gst_buffer_unref (buffer);
GST_DEBUG_OBJECT (pad, "Dropping buffer due to FALSE probe return");
GST_PAD_STREAM_UNLOCK (pad);
return GST_FLOW_OK;
}
not_negotiated:
{
gst_buffer_unref (buffer);
GST_CAT_LOG_OBJECT (GST_CAT_SCHEDULING, pad,
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
"pushing buffer but pad did not accept");
GST_PAD_STREAM_UNLOCK (pad);
return GST_FLOW_NOT_NEGOTIATED;
}
no_function:
{
gst_buffer_unref (buffer);
GST_CAT_LOG_OBJECT (GST_CAT_SCHEDULING, pad,
"pushing, but not chainhandler");
GST_ELEMENT_ERROR (GST_PAD_PARENT (pad), CORE, PAD, (NULL),
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
("push on pad %s:%s but it has no chainfunction",
GST_DEBUG_PAD_NAME (pad)));
GST_PAD_STREAM_UNLOCK (pad);
return GST_FLOW_NOT_SUPPORTED;
}
}
/**
* gst_pad_chain:
* @pad: a sink #GstPad, returns GST_FLOW_ERROR if not.
* @buffer: the #GstBuffer to send, return GST_FLOW_ERROR if not.
*
* Chain a buffer to @pad.
*
* The function returns #GST_FLOW_WRONG_STATE if the pad was flushing.
*
* If the caps on @buffer are different from the current caps on @pad, this
* function will call any setcaps function (see gst_pad_set_setcaps_function())
* installed on @pad. If the new caps are not acceptable for @pad, this function
* returns #GST_FLOW_NOT_NEGOTIATED.
*
* The function proceeds calling the chain function installed on @pad (see
* gst_pad_set_chain_function()) and the return value of that function is
* returned to the caller. #GST_FLOW_NOT_SUPPORTED is returned if @pad has no
* chain function.
*
* In all cases, success or failure, the caller loses its reference to @buffer
* after calling this function.
*
* Returns: a #GstFlowReturn from the pad.
*
* MT safe.
*/
GstFlowReturn
gst_pad_chain (GstPad * pad, GstBuffer * buffer)
{
g_return_val_if_fail (GST_IS_PAD (pad), GST_FLOW_ERROR);
g_return_val_if_fail (GST_PAD_DIRECTION (pad) == GST_PAD_SINK,
GST_FLOW_ERROR);
g_return_val_if_fail (GST_IS_BUFFER (buffer), GST_FLOW_ERROR);
return gst_pad_chain_unchecked (pad, buffer);
}
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
/**
* gst_pad_push:
* @pad: a source #GstPad, returns #GST_FLOW_ERROR if not.
* @buffer: the #GstBuffer to push returns GST_FLOW_ERROR if not.
*
* Pushes a buffer to the peer of @pad.
*
* This function will call an installed pad block before triggering any
* installed pad probes.
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
*
* If the caps on @buffer are different from the currently configured caps on
* @pad, this function will call any installed setcaps function on @pad (see
* gst_pad_set_setcaps_function()). In case of failure to renegotiate the new
* format, this function returns #GST_FLOW_NOT_NEGOTIATED.
*
* The function proceeds calling gst_pad_chain() on the peer pad and returns the
* value from that function. If @pad has no peer, #GST_FLOW_NOT_LINKED will be
* returned.
*
* In all cases, success or failure, the caller loses its reference to @buffer
* after calling this function.
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
*
* Returns: a #GstFlowReturn from the peer pad.
*
* MT safe.
*/
GstFlowReturn
gst_pad_push (GstPad * pad, GstBuffer * buffer)
{
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
GstPad *peer;
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
GstFlowReturn ret;
GstCaps *caps;
gboolean caps_changed;
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
g_return_val_if_fail (GST_IS_PAD (pad), GST_FLOW_ERROR);
g_return_val_if_fail (GST_PAD_DIRECTION (pad) == GST_PAD_SRC, GST_FLOW_ERROR);
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
g_return_val_if_fail (GST_IS_BUFFER (buffer), GST_FLOW_ERROR);
GST_OBJECT_LOCK (pad);
/* FIXME: this check can go away; pad_set_blocked could be implemented with
* probes completely or probes with an extended pad block. */
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
while (G_UNLIKELY (GST_PAD_IS_BLOCKED (pad)))
if ((ret = handle_pad_block (pad)) != GST_FLOW_OK)
goto flushed;
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
/* we emit signals on the pad arg, the peer will have a chance to
* emit in the _chain() function */
if (G_UNLIKELY (GST_PAD_DO_BUFFER_SIGNALS (pad) > 0)) {
/* unlock before emitting */
GST_OBJECT_UNLOCK (pad);
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
/* if the signal handler returned FALSE, it means we should just drop the
* buffer */
if (!gst_pad_emit_have_data_signal (pad, GST_MINI_OBJECT (buffer)))
goto dropped;
GST_OBJECT_LOCK (pad);
}
if (G_UNLIKELY ((peer = GST_PAD_PEER (pad)) == NULL))
goto not_linked;
/* take ref to peer pad before releasing the lock */
gst_object_ref (peer);
/* Before pushing the buffer to the peer pad, ensure that caps
* are set on this pad */
caps = GST_BUFFER_CAPS (buffer);
caps_changed = caps && caps != GST_PAD_CAPS (pad);
GST_OBJECT_UNLOCK (pad);
/* we got a new datatype from the pad, it had better handle it */
if (G_UNLIKELY (caps_changed)) {
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_DEBUG_OBJECT (pad, "caps changed to %p %" GST_PTR_FORMAT, caps, caps);
if (G_UNLIKELY (!gst_pad_configure_src (pad, caps, TRUE)))
goto not_negotiated;
}
ret = gst_pad_chain_unchecked (peer, buffer);
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
gst_object_unref (peer);
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
return ret;
/* ERROR recovery here */
flushed:
{
gst_buffer_unref (buffer);
GST_DEBUG_OBJECT (pad, "pad block stopped by flush");
GST_OBJECT_UNLOCK (pad);
return ret;
}
dropped:
{
gst_buffer_unref (buffer);
GST_DEBUG_OBJECT (pad, "Dropping buffer due to FALSE probe return");
return GST_FLOW_OK;
}
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
not_linked:
{
gst_buffer_unref (buffer);
GST_CAT_LOG_OBJECT (GST_CAT_SCHEDULING, pad,
"pushing, but it was not linked");
GST_OBJECT_UNLOCK (pad);
return GST_FLOW_NOT_LINKED;
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
}
not_negotiated:
{
gst_buffer_unref (buffer);
gst_object_unref (peer);
GST_CAT_DEBUG_OBJECT (GST_CAT_SCHEDULING, pad,
"element pushed buffer then refused to accept the caps");
return GST_FLOW_NOT_NEGOTIATED;
}
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
}
/**
* gst_pad_check_pull_range:
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
* @pad: a sink #GstPad.
*
* Checks if a gst_pad_pull_range() can be performed on the peer
* source pad. This function is used by plugins that want to check
* if they can use random access on the peer source pad.
*
* The peer sourcepad can implement a custom #GstPadCheckGetRangeFunction
* if it needs to perform some logic to determine if pull_range is
* possible.
gst/: Simplify pad activation. Original commit message from CVS: * gst/base/Makefile.am: * gst/base/README: * gst/base/gstbasesink.c: (gst_basesink_get_type), (gst_basesink_base_init), (gst_basesink_class_init), (gst_basesink_pad_getcaps), (gst_basesink_init), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesink.h: * gst/base/gstbasetransform.c: (gst_base_transform_get_type), (gst_base_transform_base_init), (gst_base_transform_finalize), (gst_base_transform_class_init), (gst_base_transform_init), (gst_base_transform_proxy_getcaps), (gst_base_transform_setcaps), (gst_base_transform_event), (gst_base_transform_getrange), (gst_base_transform_chain), (gst_base_transform_handle_buffer), (gst_base_transform_set_property), (gst_base_transform_get_property), (gst_base_transform_sink_activate), (gst_base_transform_src_activate), (gst_base_transform_change_state): * gst/base/gstbasetransform.h: * gst/elements/gstidentity.c: (gst_identity_finalize), (gst_identity_class_init), (gst_identity_init), (gst_identity_event), (gst_identity_check_perfect), (gst_identity_transform), (gst_identity_set_property), (gst_identity_get_property), (gst_identity_change_state): * gst/elements/gstidentity.h: * gst/gstelement.c: (gst_element_get_state_func), (gst_element_lost_state), (gst_element_pads_activate): * gst/gstpad.c: (gst_pad_set_active), (gst_pad_peer_set_active), (gst_pad_check_pull_range), (gst_pad_pull_range): * gst/gstpad.h: Simplify pad activation. Added function to check if pull_range can be performed. Error out when pulling inactive or flushing pads. Removed const from refcounted types as it does not make sense. Simplify pad templates in basesink Added base class for simple 1-to-1 transforms. Make identity subclass the base transform.
2005-03-29 16:18:12 +00:00
*
* Returns: a gboolean with the result.
*
* MT safe.
*/
gboolean
gst_pad_check_pull_range (GstPad * pad)
{
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
GstPad *peer;
gboolean ret;
GstPadCheckGetRangeFunction checkgetrangefunc;
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
g_return_val_if_fail (GST_IS_PAD (pad), FALSE);
GST_OBJECT_LOCK (pad);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
if (GST_PAD_DIRECTION (pad) != GST_PAD_SINK)
goto wrong_direction;
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
if (G_UNLIKELY ((peer = GST_PAD_PEER (pad)) == NULL))
goto not_connected;
gst_object_ref (peer);
GST_OBJECT_UNLOCK (pad);
gst/gstelement.c (gst_element_dispose): Protect against multiple invocations. Original commit message from CVS: 2004-02-24 Andy Wingo <wingo@pobox.com> * gst/gstelement.c (gst_element_dispose): Protect against multiple invocations. * gst/schedulers/gstoptimalscheduler.c I added a mess of prototypes at the top of the file by way of documentation. Some of the operations on chains and groups were re-organized. (create_group): Added a type argument so if the group is enabled, the setup_group_scheduler knows what to do. (group_elements): Added a type argument here, too, to be passed on to create_group. (group_element_set_enabled): If an unlinked PLAYING element is added to a bin, we have to create a new group to hold the element, and this function will be called before the group is added to the chain. Thus we have a valid case for group->chain==NULL. Instead of calling chain_group_set_enabled, just set the flag on the group (the chain's status will be set when the group is added to it). (gst_opt_scheduler_state_transition, chain_group_set_enabled): Setup the group scheduler when the group is enabled, not specifically when an element goes PAUSED->PLAYING. This means PLAYING elements can be added, linked, and scheduled into a PLAYING pipeline, as was intended. (add_to_group): Don't ref the group twice. I don't know when this double-ref got in here. Removing it has the potential to cause segfaults if other parts of the scheduler are buggy. If you find that the scheduler is segfaulting for you, put in an extra ref here and see if that hacks over the underlying issue. Of course, then find out what code is unreffing a group it doesn't own... (create_group): Make the extra refcount floating, and remove it after adding the element. This means that... (unref_group): Destroy when the refcount reaches 0, not 1, like every other refcounted object in the known universe. (remove_from_group): When a group becomes empty, set it to be not active, and remove it from its chain. Don't unref it again, there's no floating reference any more. (destroy_group): We have to remove the group from the chain in remove_from_group (rather than here) to break refcounting cycles (the chain always has a ref on the group). So assert that group->chain==NULL. (ref_group_by_count): Removed, it was commented out anyway. (merge_chains): Use the remove_from_chain and add_to_chain primitives to do the reparenting, instead of rolling our own implementation. (add_to_chain): The first non-disabled group in the chain's group list will be the entry point for the chain. Because buffers can accumulate in loop elements' peer bufpens, we preferentially schedule loop groups before get groups to avoid unnecessary execution of get-based groups when the bufpens are already full. (gst_opt_scheduler_schedule_run_queue): Debug fixes. (get_group_schedule_function): Ditto. (loop_group_schedule_function): Ditto. (gst_opt_scheduler_loop_wrapper): Ditto. (gst_opt_scheduler_iterate): Ditto. I understand the opt scheduler now, yippee! * gst/gstpad.c: All throughout, added FIXMEs to look at for 0.9. (gst_pad_get_name, gst_pad_set_chain_function) (gst_pad_set_get_function, gst_pad_set_event_function) (gst_pad_set_event_mask_function, gst_pad_get_event_masks) (gst_pad_get_event_masks_default, gst_pad_set_convert_function) (gst_pad_set_query_function, gst_pad_get_query_types) (gst_pad_get_query_types_default) (gst_pad_set_internal_link_function) (gst_pad_set_formats_function, gst_pad_set_link_function) (gst_pad_set_fixate_function, gst_pad_set_getcaps_function) (gst_pad_set_bufferalloc_function, gst_pad_unlink) (gst_pad_renegotiate, gst_pad_set_parent, gst_pad_get_parent) (gst_pad_add_ghost_pad, gst_pad_proxy_getcaps) (gst_pad_proxy_pad_link, gst_pad_proxy_fixate) (gst_pad_get_pad_template_caps, gst_pad_check_compatibility) (gst_pad_get_peer, gst_pad_get_allowed_caps) (gst_pad_alloc_buffer, gst_pad_push, gst_pad_pull) (gst_pad_selectv, gst_pad_select, gst_pad_template_get_caps) (gst_pad_event_default_dispatch, gst_pad_event_default) (gst_pad_dispatcher, gst_pad_send_event, gst_pad_convert_default) (gst_pad_convert, gst_pad_query_default, gst_pad_query) (gst_pad_get_formats_default, gst_pad_get_formats): Better argument checks, and some doc fixes. (gst_pad_custom_new_from_template): Um, does anyone use these functions? Actually make a custom pad instead of a normal one. (gst_pad_try_set_caps): Transpose some checks. (gst_pad_try_set_caps_nonfixed): Same, and use a macro to check if the pad is in negotiation. (gst_pad_try_relink_filtered): Use pad_link_prepare. * gst/gstelement.c: Remove prototypes also defined in gstclock.h. * gst/gstelement.h: * gst/gstclock.h: Un-deprecate the old clocking API, as discussed on the list.
2004-02-25 13:16:12 +00:00
/* see note in above function */
if (G_LIKELY ((checkgetrangefunc = peer->checkgetrangefunc) == NULL)) {
/* FIXME, kindoff ghetto */
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
ret = GST_PAD_GETRANGEFUNC (peer) != NULL;
} else {
GST_CAT_LOG_OBJECT (GST_CAT_SCHEDULING, pad,
"calling checkgetrangefunc %s of peer pad %s:%s",
GST_DEBUG_FUNCPTR_NAME (checkgetrangefunc), GST_DEBUG_PAD_NAME (peer));
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
ret = checkgetrangefunc (peer);
}
gst_object_unref (peer);
return ret;
/* ERROR recovery here */
wrong_direction:
{
GST_CAT_LOG_OBJECT (GST_CAT_SCHEDULING, pad,
"checking pull range, but pad must be a sinkpad");
GST_OBJECT_UNLOCK (pad);
return FALSE;
}
not_connected:
{
GST_CAT_LOG_OBJECT (GST_CAT_SCHEDULING, pad,
"checking pull range, but it was not linked");
GST_OBJECT_UNLOCK (pad);
return FALSE;
}
}
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
/**
* gst_pad_get_range:
* @pad: a src #GstPad, returns #GST_FLOW_ERROR if not.
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
* @offset: The start offset of the buffer
* @size: The length of the buffer
* @buffer: a pointer to hold the #GstBuffer, returns #GST_FLOW_ERROR if %NULL.
*
* When @pad is flushing this function returns #GST_FLOW_WRONG_STATE
* immediatly.
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
*
* Calls the getrange function of @pad, see #GstPadGetRangeFunction for a
* description of a getrange function. If @pad has no getrange function
* installed (see gst_pad_set_getrange_function()) this function returns
* #GST_FLOW_NOT_SUPPORTED.
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
*
* Returns: a #GstFlowReturn from the pad.
*
* MT safe.
*/
GstFlowReturn
gst_pad_get_range (GstPad * pad, guint64 offset, guint size,
GstBuffer ** buffer)
{
GstFlowReturn ret;
GstPadGetRangeFunction getrangefunc;
gboolean emit_signal;
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
g_return_val_if_fail (GST_IS_PAD (pad), GST_FLOW_ERROR);
g_return_val_if_fail (GST_PAD_DIRECTION (pad) == GST_PAD_SRC, GST_FLOW_ERROR);
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
g_return_val_if_fail (buffer != NULL, GST_FLOW_ERROR);
GST_PAD_STREAM_LOCK (pad);
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
GST_OBJECT_LOCK (pad);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
if (G_UNLIKELY (GST_PAD_IS_FLUSHING (pad)))
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
goto flushing;
emit_signal = GST_PAD_DO_BUFFER_SIGNALS (pad) > 0;
GST_OBJECT_UNLOCK (pad);
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
if (G_UNLIKELY ((getrangefunc = GST_PAD_GETRANGEFUNC (pad)) == NULL))
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
goto no_function;
GST_CAT_LOG_OBJECT (GST_CAT_SCHEDULING, pad,
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
"calling getrangefunc %s, offset %"
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
G_GUINT64_FORMAT ", size %u",
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_DEBUG_FUNCPTR_NAME (getrangefunc), offset, size);
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
ret = getrangefunc (pad, offset, size, buffer);
/* can only fire the signal if we have a valid buffer */
if (G_UNLIKELY (emit_signal) && (ret == GST_FLOW_OK)) {
if (!gst_pad_emit_have_data_signal (pad, GST_MINI_OBJECT (*buffer)))
goto dropping;
}
GST_PAD_STREAM_UNLOCK (pad);
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
return ret;
/* ERRORS */
flushing:
{
GST_CAT_LOG_OBJECT (GST_CAT_SCHEDULING, pad,
"pulling range, but pad was flushing");
GST_OBJECT_UNLOCK (pad);
GST_PAD_STREAM_UNLOCK (pad);
return GST_FLOW_WRONG_STATE;
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
}
no_function:
{
GST_ELEMENT_ERROR (GST_PAD_PARENT (pad), CORE, PAD, (NULL),
("pullrange on pad %s:%s but it has no getrangefunction",
GST_DEBUG_PAD_NAME (pad)));
GST_PAD_STREAM_UNLOCK (pad);
return GST_FLOW_NOT_SUPPORTED;
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
}
dropping:
{
GST_CAT_LOG_OBJECT (GST_CAT_SCHEDULING, pad,
"Dropping data after FALSE probe return");
GST_PAD_STREAM_UNLOCK (pad);
gst_buffer_unref (*buffer);
*buffer = NULL;
return GST_FLOW_UNEXPECTED;
}
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
}
/**
* gst_pad_pull_range:
* @pad: a sink #GstPad, returns GST_FLOW_ERROR if not.
* @offset: The start offset of the buffer
* @size: The length of the buffer
* @buffer: a pointer to hold the #GstBuffer, returns GST_FLOW_ERROR if %NULL.
*
* Pulls a buffer from the peer pad.
*
* This function will first trigger the pad block signal if it was
* installed.
*
* When @pad is not linked #GST_FLOW_NOT_LINKED is returned else this
* function returns the result of gst_pad_get_range() on the peer pad.
* See gst_pad_get_range() for a list of return values and for the
* semantics of the arguments of this function.
*
* Returns: a #GstFlowReturn from the peer pad.
* When this function returns #GST_FLOW_OK, @buffer will contain a valid
* #GstBuffer that should be freed with gst_buffer_unref() after usage.
* @buffer may not be used or freed when any other return value than
* #GST_FLOW_OK is returned.
*
* MT safe.
*/
GstFlowReturn
gst_pad_pull_range (GstPad * pad, guint64 offset, guint size,
GstBuffer ** buffer)
{
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
GstPad *peer;
GstFlowReturn ret;
gboolean emit_signal;
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
g_return_val_if_fail (GST_IS_PAD (pad), GST_FLOW_ERROR);
g_return_val_if_fail (GST_PAD_DIRECTION (pad) == GST_PAD_SINK,
GST_FLOW_ERROR);
g_return_val_if_fail (buffer != NULL, GST_FLOW_ERROR);
GST_OBJECT_LOCK (pad);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
while (G_UNLIKELY (GST_PAD_IS_BLOCKED (pad)))
handle_pad_block (pad);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
if (G_UNLIKELY ((peer = GST_PAD_PEER (pad)) == NULL))
goto not_connected;
/* signal emision for the pad, peer has chance to emit when
* we call _get_range() */
emit_signal = GST_PAD_DO_BUFFER_SIGNALS (pad) > 0;
gst_object_ref (peer);
GST_OBJECT_UNLOCK (pad);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
ret = gst_pad_get_range (peer, offset, size, buffer);
gst_object_unref (peer);
/* can only fire the signal if we have a valid buffer */
if (G_UNLIKELY (emit_signal) && (ret == GST_FLOW_OK)) {
if (!gst_pad_emit_have_data_signal (pad, GST_MINI_OBJECT (*buffer)))
goto dropping;
}
return ret;
/* ERROR recovery here */
not_connected:
{
GST_CAT_LOG_OBJECT (GST_CAT_SCHEDULING, pad,
"pulling range, but it was not linked");
GST_OBJECT_UNLOCK (pad);
return GST_FLOW_NOT_LINKED;
}
dropping:
{
GST_CAT_LOG_OBJECT (GST_CAT_SCHEDULING, pad,
"Dropping data after FALSE probe return");
gst_buffer_unref (*buffer);
*buffer = NULL;
return GST_FLOW_UNEXPECTED;
}
}
/**
* gst_pad_push_event:
* @pad: a #GstPad to push the event to.
* @event: the #GstEvent to send to the pad.
*
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
* Sends the event to the peer of the given pad. This function is
* mainly used by elements to send events to their peer
* elements.
*
* This function takes owership of the provided event so you should
* gst_event_ref() it if you want to reuse the event after this call.
*
* Returns: TRUE if the event was handled.
*
* MT safe.
*/
gboolean
gst_pad_push_event (GstPad * pad, GstEvent * event)
{
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
GstPad *peerpad;
gboolean result;
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
g_return_val_if_fail (GST_IS_PAD (pad), FALSE);
g_return_val_if_fail (event != NULL, FALSE);
g_return_val_if_fail (GST_IS_EVENT (event), FALSE);
GST_LOG_OBJECT (pad, "event: %s", GST_EVENT_TYPE_NAME (event));
GST_OBJECT_LOCK (pad);
/* Two checks to be made:
* . (un)set the FLUSHING flag for flushing events,
* . handle pad blocking */
switch (GST_EVENT_TYPE (event)) {
case GST_EVENT_FLUSH_START:
GST_PAD_SET_FLUSHING (pad);
if (G_UNLIKELY (GST_PAD_IS_BLOCKED (pad))) {
/* flush start will have set the FLUSHING flag and will then
* unlock all threads doing a GCond wait on the blocking pad. This
* will typically unblock the STREAMING thread blocked on a pad. */
GST_LOG_OBJECT (pad, "Pad is blocked, not forwarding flush-start, "
"doing block signal.");
GST_PAD_BLOCK_BROADCAST (pad);
goto flushed;
}
break;
case GST_EVENT_FLUSH_STOP:
GST_PAD_UNSET_FLUSHING (pad);
/* if we are blocked, flush away the FLUSH_STOP event */
if (G_UNLIKELY (GST_PAD_IS_BLOCKED (pad))) {
GST_LOG_OBJECT (pad, "Pad is blocked, not forwarding flush-stop");
goto flushed;
}
break;
default:
while (G_UNLIKELY (GST_PAD_IS_BLOCKED (pad))) {
/* block the event as long as the pad is blocked */
if (handle_pad_block (pad) != GST_FLOW_OK)
goto flushed;
}
break;
}
if (G_UNLIKELY (GST_PAD_DO_EVENT_SIGNALS (pad) > 0)) {
GST_OBJECT_UNLOCK (pad);
if (!gst_pad_emit_have_data_signal (pad, GST_MINI_OBJECT (event)))
goto dropping;
GST_OBJECT_LOCK (pad);
}
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
peerpad = GST_PAD_PEER (pad);
if (peerpad == NULL)
goto not_linked;
GST_LOG_OBJECT (pad, "sending event %s to peerpad %" GST_PTR_FORMAT,
GST_EVENT_TYPE_NAME (event), peerpad);
gst_object_ref (peerpad);
GST_OBJECT_UNLOCK (pad);
result = gst_pad_send_event (peerpad, event);
/* Note: we gave away ownership of the event at this point */
GST_LOG_OBJECT (pad, "sent event to peerpad %" GST_PTR_FORMAT, peerpad);
gst_object_unref (peerpad);
return result;
/* ERROR handling */
dropping:
{
GST_DEBUG_OBJECT (pad, "Dropping event after FALSE probe return");
gst_event_unref (event);
return FALSE;
}
not_linked:
{
gst/gstbin.c: (de)activate src pads before calling state_change on the childs. Original commit message from CVS: * gst/gstbin.c: (activate_pads), (iterator_activate_fold_with_resync), (gst_bin_src_pads_activate), (gst_bin_change_state_func): (de)activate src pads before calling state_change on the childs. This is to avoid the case where a src ghostpad is blocked (holding the stream lock), which would block the deactivation of the ghostpad's target pad. * gst/gstghostpad.c: (gst_proxy_pad_do_query_type), (gst_proxy_pad_do_event), (gst_proxy_pad_do_query), (gst_proxy_pad_do_internal_link), (gst_proxy_pad_do_bufferalloc), (gst_proxy_pad_do_chain), (gst_proxy_pad_do_getrange), (gst_proxy_pad_do_checkgetrange), (gst_proxy_pad_do_getcaps), (gst_proxy_pad_do_acceptcaps), (gst_proxy_pad_do_fixatecaps), (gst_proxy_pad_do_setcaps), (gst_proxy_pad_set_target_unlocked), (gst_proxy_pad_set_target), (gst_proxy_pad_get_internal), (gst_proxy_pad_dispose), (gst_proxy_pad_init), (gst_ghost_pad_parent_set), (gst_ghost_pad_parent_unset), (gst_ghost_pad_class_init), (gst_ghost_pad_internal_do_activate_push), (gst_ghost_pad_internal_do_activate_pull), (gst_ghost_pad_do_activate_push), (gst_ghost_pad_do_activate_pull), (gst_ghost_pad_do_link), (gst_ghost_pad_do_unlink), (gst_ghost_pad_dispose), (gst_ghost_pad_new_no_target), (gst_ghost_pad_new), (gst_ghost_pad_set_target): GhostPads now create their internal GstProxyPad at creation (and not when they're linked, as it was being done previously). The internal and target pads are linked straight away. The data will also travel through the other pad in order to make pad blocking and probes non-hackish (the probe/block now really happens on the GhostPad and not on the target). * gst/gstpad.c: (gst_pad_set_blocked_async), (gst_pad_link_prepare), (gst_pad_push_event): Remove previous ghostpad cruft. * gst/gstutils.c: (gst_pad_add_data_probe), (gst_pad_add_event_probe), (gst_pad_add_buffer_probe), (gst_pad_remove_data_probe), (gst_pad_remove_event_probe), (gst_pad_remove_buffer_probe): Remove previous ghost pad cruft. Added more detailed debug statements. * tests/check/gst/gstghostpad.c: (GST_START_TEST): Fix the testsuite for refcounting changes. The comments about who has references were correct, but the refcount being checked wasn't the same (!?!).
2006-07-11 16:20:09 +00:00
GST_DEBUG_OBJECT (pad, "Dropping event because pad is not linked");
gst_event_unref (event);
GST_OBJECT_UNLOCK (pad);
return FALSE;
}
flushed:
{
GST_DEBUG_OBJECT (pad,
"Not forwarding event since we're flushing and blocking");
gst_event_unref (event);
GST_OBJECT_UNLOCK (pad);
return TRUE;
}
}
/**
* gst_pad_send_event:
* @pad: a #GstPad to send the event to.
* @event: the #GstEvent to send to the pad.
*
* Sends the event to the pad. This function can be used
* by applications to send events in the pipeline.
*
* If @pad is a source pad, @event should be an upstream event. If @pad is a
* sink pad, @event should be a downstream event. For example, you would not
* send a #GST_EVENT_EOS on a src pad; EOS events only propagate downstream.
* Furthermore, some downstream events have to be serialized with data flow,
* like EOS, while some can travel out-of-band, like #GST_EVENT_FLUSH_START. If
* the event needs to be serialized with data flow, this function will take the
* pad's stream lock while calling its event function.
*
* To find out whether an event type is upstream, downstream, or downstream and
* serialized, see #GstEventTypeFlags, gst_event_type_get_flags(),
* #GST_EVENT_IS_UPSTREAM, #GST_EVENT_IS_DOWNSTREAM, and
* #GST_EVENT_IS_SERIALIZED. Note that in practice that an application or plugin
* doesn't need to bother itself with this information; the core handles all
* necessary locks and checks.
*
* This function takes owership of the provided event so you should
* gst_event_ref() it if you want to reuse the event after this call.
*
* Returns: TRUE if the event was handled.
*/
gboolean
gst_pad_send_event (GstPad * pad, GstEvent * event)
{
gboolean result = FALSE;
GstPadEventFunction eventfunc;
gboolean serialized, need_unlock = FALSE;
g_return_val_if_fail (GST_IS_PAD (pad), FALSE);
gst/gstelement.c (gst_element_dispose): Protect against multiple invocations. Original commit message from CVS: 2004-02-24 Andy Wingo <wingo@pobox.com> * gst/gstelement.c (gst_element_dispose): Protect against multiple invocations. * gst/schedulers/gstoptimalscheduler.c I added a mess of prototypes at the top of the file by way of documentation. Some of the operations on chains and groups were re-organized. (create_group): Added a type argument so if the group is enabled, the setup_group_scheduler knows what to do. (group_elements): Added a type argument here, too, to be passed on to create_group. (group_element_set_enabled): If an unlinked PLAYING element is added to a bin, we have to create a new group to hold the element, and this function will be called before the group is added to the chain. Thus we have a valid case for group->chain==NULL. Instead of calling chain_group_set_enabled, just set the flag on the group (the chain's status will be set when the group is added to it). (gst_opt_scheduler_state_transition, chain_group_set_enabled): Setup the group scheduler when the group is enabled, not specifically when an element goes PAUSED->PLAYING. This means PLAYING elements can be added, linked, and scheduled into a PLAYING pipeline, as was intended. (add_to_group): Don't ref the group twice. I don't know when this double-ref got in here. Removing it has the potential to cause segfaults if other parts of the scheduler are buggy. If you find that the scheduler is segfaulting for you, put in an extra ref here and see if that hacks over the underlying issue. Of course, then find out what code is unreffing a group it doesn't own... (create_group): Make the extra refcount floating, and remove it after adding the element. This means that... (unref_group): Destroy when the refcount reaches 0, not 1, like every other refcounted object in the known universe. (remove_from_group): When a group becomes empty, set it to be not active, and remove it from its chain. Don't unref it again, there's no floating reference any more. (destroy_group): We have to remove the group from the chain in remove_from_group (rather than here) to break refcounting cycles (the chain always has a ref on the group). So assert that group->chain==NULL. (ref_group_by_count): Removed, it was commented out anyway. (merge_chains): Use the remove_from_chain and add_to_chain primitives to do the reparenting, instead of rolling our own implementation. (add_to_chain): The first non-disabled group in the chain's group list will be the entry point for the chain. Because buffers can accumulate in loop elements' peer bufpens, we preferentially schedule loop groups before get groups to avoid unnecessary execution of get-based groups when the bufpens are already full. (gst_opt_scheduler_schedule_run_queue): Debug fixes. (get_group_schedule_function): Ditto. (loop_group_schedule_function): Ditto. (gst_opt_scheduler_loop_wrapper): Ditto. (gst_opt_scheduler_iterate): Ditto. I understand the opt scheduler now, yippee! * gst/gstpad.c: All throughout, added FIXMEs to look at for 0.9. (gst_pad_get_name, gst_pad_set_chain_function) (gst_pad_set_get_function, gst_pad_set_event_function) (gst_pad_set_event_mask_function, gst_pad_get_event_masks) (gst_pad_get_event_masks_default, gst_pad_set_convert_function) (gst_pad_set_query_function, gst_pad_get_query_types) (gst_pad_get_query_types_default) (gst_pad_set_internal_link_function) (gst_pad_set_formats_function, gst_pad_set_link_function) (gst_pad_set_fixate_function, gst_pad_set_getcaps_function) (gst_pad_set_bufferalloc_function, gst_pad_unlink) (gst_pad_renegotiate, gst_pad_set_parent, gst_pad_get_parent) (gst_pad_add_ghost_pad, gst_pad_proxy_getcaps) (gst_pad_proxy_pad_link, gst_pad_proxy_fixate) (gst_pad_get_pad_template_caps, gst_pad_check_compatibility) (gst_pad_get_peer, gst_pad_get_allowed_caps) (gst_pad_alloc_buffer, gst_pad_push, gst_pad_pull) (gst_pad_selectv, gst_pad_select, gst_pad_template_get_caps) (gst_pad_event_default_dispatch, gst_pad_event_default) (gst_pad_dispatcher, gst_pad_send_event, gst_pad_convert_default) (gst_pad_convert, gst_pad_query_default, gst_pad_query) (gst_pad_get_formats_default, gst_pad_get_formats): Better argument checks, and some doc fixes. (gst_pad_custom_new_from_template): Um, does anyone use these functions? Actually make a custom pad instead of a normal one. (gst_pad_try_set_caps): Transpose some checks. (gst_pad_try_set_caps_nonfixed): Same, and use a macro to check if the pad is in negotiation. (gst_pad_try_relink_filtered): Use pad_link_prepare. * gst/gstelement.c: Remove prototypes also defined in gstclock.h. * gst/gstelement.h: * gst/gstclock.h: Un-deprecate the old clocking API, as discussed on the list.
2004-02-25 13:16:12 +00:00
g_return_val_if_fail (event != NULL, FALSE);
GST_OBJECT_LOCK (pad);
if (GST_PAD_IS_SINK (pad)) {
if (G_UNLIKELY (!GST_EVENT_IS_DOWNSTREAM (event)))
goto wrong_direction;
serialized = GST_EVENT_IS_SERIALIZED (event);
} else if (GST_PAD_IS_SRC (pad)) {
if (G_UNLIKELY (!GST_EVENT_IS_UPSTREAM (event)))
goto wrong_direction;
/* events on srcpad never are serialized */
serialized = FALSE;
} else
goto unknown_direction;
Some docs updates Original commit message from CVS: * CHANGES-0.9: * docs/design/part-TODO.txt: * docs/design/part-events.txt: Some docs updates * gst/base/gstbasesink.c: (gst_base_sink_handle_object), (gst_base_sink_event), (gst_base_sink_do_sync), (gst_base_sink_activate_push), (gst_base_sink_activate_pull): * gst/base/gstbasesrc.c: (gst_base_src_send_discont), (gst_base_src_do_seek), (gst_base_src_event_handler), (gst_base_src_loop): * gst/base/gstbasetransform.c: (gst_base_transform_transform_caps), (gst_base_transform_configure_caps), (gst_base_transform_setcaps), (gst_base_transform_get_size), (gst_base_transform_buffer_alloc), (gst_base_transform_event), (gst_base_transform_handle_buffer), (gst_base_transform_set_passthrough), (gst_base_transform_is_passthrough): * gst/elements/gstfakesink.c: (gst_fake_sink_event): * gst/elements/gstfilesink.c: (gst_file_sink_event): Event updates. * gst/gstbuffer.h: Use faster casts. * gst/gstelement.c: (gst_element_seek): * gst/gstelement.h: Update gst_element_seek. * gst/gstevent.c: (gst_event_finalize), (_gst_event_copy), (gst_event_new), (gst_event_new_custom), (gst_event_get_structure), (gst_event_new_flush_start), (gst_event_new_flush_stop), (gst_event_new_eos), (gst_event_new_newsegment), (gst_event_parse_newsegment), (gst_event_new_tag), (gst_event_parse_tag), (gst_event_new_filler), (gst_event_new_qos), (gst_event_parse_qos), (gst_event_new_seek), (gst_event_parse_seek), (gst_event_new_navigation): * gst/gstevent.h: Make GstEvent use GstStructure. Add parsing code, make sure the API is sufficiently generic. Mark possible directions of events and serialization. * gst/gstmessage.c: (gst_message_init), (gst_message_finalize), (_gst_message_copy), (gst_message_new_segment_start), (gst_message_new_segment_done), (gst_message_new_custom), (gst_message_parse_segment_start), (gst_message_parse_segment_done): Small cleanups. * gst/gstpad.c: (gst_pad_get_caps_unlocked), (gst_pad_accept_caps), (gst_pad_set_caps), (gst_pad_send_event): Update for new events. Catch events sent in wrong directions. * gst/gstqueue.c: (gst_queue_link_src), (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_handle_src_query): Event updates. * gst/gsttag.c: * gst/gsttag.h: Remove event code from this file. * libs/gst/dataprotocol/dataprotocol.c: (gst_dp_packet_from_event), (gst_dp_event_from_packet): Event updates.
2005-07-27 18:33:03 +00:00
if (G_UNLIKELY (GST_EVENT_SRC (event) == NULL)) {
GST_LOG_OBJECT (pad, "event had no source, setting pad as event source");
GST_EVENT_SRC (event) = gst_object_ref (pad);
}
/* pad signals */
if (G_UNLIKELY (GST_PAD_DO_EVENT_SIGNALS (pad) > 0)) {
GST_OBJECT_UNLOCK (pad);
if (!gst_pad_emit_have_data_signal (pad, GST_MINI_OBJECT_CAST (event)))
goto dropping;
GST_OBJECT_LOCK (pad);
}
switch (GST_EVENT_TYPE (event)) {
Some docs updates Original commit message from CVS: * CHANGES-0.9: * docs/design/part-TODO.txt: * docs/design/part-events.txt: Some docs updates * gst/base/gstbasesink.c: (gst_base_sink_handle_object), (gst_base_sink_event), (gst_base_sink_do_sync), (gst_base_sink_activate_push), (gst_base_sink_activate_pull): * gst/base/gstbasesrc.c: (gst_base_src_send_discont), (gst_base_src_do_seek), (gst_base_src_event_handler), (gst_base_src_loop): * gst/base/gstbasetransform.c: (gst_base_transform_transform_caps), (gst_base_transform_configure_caps), (gst_base_transform_setcaps), (gst_base_transform_get_size), (gst_base_transform_buffer_alloc), (gst_base_transform_event), (gst_base_transform_handle_buffer), (gst_base_transform_set_passthrough), (gst_base_transform_is_passthrough): * gst/elements/gstfakesink.c: (gst_fake_sink_event): * gst/elements/gstfilesink.c: (gst_file_sink_event): Event updates. * gst/gstbuffer.h: Use faster casts. * gst/gstelement.c: (gst_element_seek): * gst/gstelement.h: Update gst_element_seek. * gst/gstevent.c: (gst_event_finalize), (_gst_event_copy), (gst_event_new), (gst_event_new_custom), (gst_event_get_structure), (gst_event_new_flush_start), (gst_event_new_flush_stop), (gst_event_new_eos), (gst_event_new_newsegment), (gst_event_parse_newsegment), (gst_event_new_tag), (gst_event_parse_tag), (gst_event_new_filler), (gst_event_new_qos), (gst_event_parse_qos), (gst_event_new_seek), (gst_event_parse_seek), (gst_event_new_navigation): * gst/gstevent.h: Make GstEvent use GstStructure. Add parsing code, make sure the API is sufficiently generic. Mark possible directions of events and serialization. * gst/gstmessage.c: (gst_message_init), (gst_message_finalize), (_gst_message_copy), (gst_message_new_segment_start), (gst_message_new_segment_done), (gst_message_new_custom), (gst_message_parse_segment_start), (gst_message_parse_segment_done): Small cleanups. * gst/gstpad.c: (gst_pad_get_caps_unlocked), (gst_pad_accept_caps), (gst_pad_set_caps), (gst_pad_send_event): Update for new events. Catch events sent in wrong directions. * gst/gstqueue.c: (gst_queue_link_src), (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_handle_src_query): Event updates. * gst/gsttag.c: * gst/gsttag.h: Remove event code from this file. * libs/gst/dataprotocol/dataprotocol.c: (gst_dp_packet_from_event), (gst_dp_event_from_packet): Event updates.
2005-07-27 18:33:03 +00:00
case GST_EVENT_FLUSH_START:
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_DEBUG_OBJECT (GST_CAT_EVENT, pad,
"have event type %d (FLUSH_START)", GST_EVENT_TYPE (event));
Some docs updates Original commit message from CVS: * CHANGES-0.9: * docs/design/part-TODO.txt: * docs/design/part-events.txt: Some docs updates * gst/base/gstbasesink.c: (gst_base_sink_handle_object), (gst_base_sink_event), (gst_base_sink_do_sync), (gst_base_sink_activate_push), (gst_base_sink_activate_pull): * gst/base/gstbasesrc.c: (gst_base_src_send_discont), (gst_base_src_do_seek), (gst_base_src_event_handler), (gst_base_src_loop): * gst/base/gstbasetransform.c: (gst_base_transform_transform_caps), (gst_base_transform_configure_caps), (gst_base_transform_setcaps), (gst_base_transform_get_size), (gst_base_transform_buffer_alloc), (gst_base_transform_event), (gst_base_transform_handle_buffer), (gst_base_transform_set_passthrough), (gst_base_transform_is_passthrough): * gst/elements/gstfakesink.c: (gst_fake_sink_event): * gst/elements/gstfilesink.c: (gst_file_sink_event): Event updates. * gst/gstbuffer.h: Use faster casts. * gst/gstelement.c: (gst_element_seek): * gst/gstelement.h: Update gst_element_seek. * gst/gstevent.c: (gst_event_finalize), (_gst_event_copy), (gst_event_new), (gst_event_new_custom), (gst_event_get_structure), (gst_event_new_flush_start), (gst_event_new_flush_stop), (gst_event_new_eos), (gst_event_new_newsegment), (gst_event_parse_newsegment), (gst_event_new_tag), (gst_event_parse_tag), (gst_event_new_filler), (gst_event_new_qos), (gst_event_parse_qos), (gst_event_new_seek), (gst_event_parse_seek), (gst_event_new_navigation): * gst/gstevent.h: Make GstEvent use GstStructure. Add parsing code, make sure the API is sufficiently generic. Mark possible directions of events and serialization. * gst/gstmessage.c: (gst_message_init), (gst_message_finalize), (_gst_message_copy), (gst_message_new_segment_start), (gst_message_new_segment_done), (gst_message_new_custom), (gst_message_parse_segment_start), (gst_message_parse_segment_done): Small cleanups. * gst/gstpad.c: (gst_pad_get_caps_unlocked), (gst_pad_accept_caps), (gst_pad_set_caps), (gst_pad_send_event): Update for new events. Catch events sent in wrong directions. * gst/gstqueue.c: (gst_queue_link_src), (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_handle_src_query): Event updates. * gst/gsttag.c: * gst/gsttag.h: Remove event code from this file. * libs/gst/dataprotocol/dataprotocol.c: (gst_dp_packet_from_event), (gst_dp_event_from_packet): Event updates.
2005-07-27 18:33:03 +00:00
/* can't even accept a flush begin event when flushing */
if (GST_PAD_IS_FLUSHING (pad))
goto flushing;
GST_PAD_SET_FLUSHING (pad);
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_DEBUG_OBJECT (GST_CAT_EVENT, pad, "set flush flag");
Some docs updates Original commit message from CVS: * CHANGES-0.9: * docs/design/part-TODO.txt: * docs/design/part-events.txt: Some docs updates * gst/base/gstbasesink.c: (gst_base_sink_handle_object), (gst_base_sink_event), (gst_base_sink_do_sync), (gst_base_sink_activate_push), (gst_base_sink_activate_pull): * gst/base/gstbasesrc.c: (gst_base_src_send_discont), (gst_base_src_do_seek), (gst_base_src_event_handler), (gst_base_src_loop): * gst/base/gstbasetransform.c: (gst_base_transform_transform_caps), (gst_base_transform_configure_caps), (gst_base_transform_setcaps), (gst_base_transform_get_size), (gst_base_transform_buffer_alloc), (gst_base_transform_event), (gst_base_transform_handle_buffer), (gst_base_transform_set_passthrough), (gst_base_transform_is_passthrough): * gst/elements/gstfakesink.c: (gst_fake_sink_event): * gst/elements/gstfilesink.c: (gst_file_sink_event): Event updates. * gst/gstbuffer.h: Use faster casts. * gst/gstelement.c: (gst_element_seek): * gst/gstelement.h: Update gst_element_seek. * gst/gstevent.c: (gst_event_finalize), (_gst_event_copy), (gst_event_new), (gst_event_new_custom), (gst_event_get_structure), (gst_event_new_flush_start), (gst_event_new_flush_stop), (gst_event_new_eos), (gst_event_new_newsegment), (gst_event_parse_newsegment), (gst_event_new_tag), (gst_event_parse_tag), (gst_event_new_filler), (gst_event_new_qos), (gst_event_parse_qos), (gst_event_new_seek), (gst_event_parse_seek), (gst_event_new_navigation): * gst/gstevent.h: Make GstEvent use GstStructure. Add parsing code, make sure the API is sufficiently generic. Mark possible directions of events and serialization. * gst/gstmessage.c: (gst_message_init), (gst_message_finalize), (_gst_message_copy), (gst_message_new_segment_start), (gst_message_new_segment_done), (gst_message_new_custom), (gst_message_parse_segment_start), (gst_message_parse_segment_done): Small cleanups. * gst/gstpad.c: (gst_pad_get_caps_unlocked), (gst_pad_accept_caps), (gst_pad_set_caps), (gst_pad_send_event): Update for new events. Catch events sent in wrong directions. * gst/gstqueue.c: (gst_queue_link_src), (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_handle_src_query): Event updates. * gst/gsttag.c: * gst/gsttag.h: Remove event code from this file. * libs/gst/dataprotocol/dataprotocol.c: (gst_dp_packet_from_event), (gst_dp_event_from_packet): Event updates.
2005-07-27 18:33:03 +00:00
break;
case GST_EVENT_FLUSH_STOP:
GST_PAD_UNSET_FLUSHING (pad);
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_DEBUG_OBJECT (GST_CAT_EVENT, pad, "cleared flush flag");
GST_OBJECT_UNLOCK (pad);
/* grab stream lock */
GST_PAD_STREAM_LOCK (pad);
need_unlock = TRUE;
GST_OBJECT_LOCK (pad);
break;
default:
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_DEBUG_OBJECT (GST_CAT_EVENT, pad, "have event type %s",
GST_EVENT_TYPE_NAME (event));
/* make this a little faster, no point in grabbing the lock
* if the pad is allready flushing. */
if (G_UNLIKELY (GST_PAD_IS_FLUSHING (pad)))
goto flushing;
if (serialized) {
/* lock order: STREAM_LOCK, LOCK, recheck flushing. */
GST_OBJECT_UNLOCK (pad);
GST_PAD_STREAM_LOCK (pad);
need_unlock = TRUE;
GST_OBJECT_LOCK (pad);
if (G_UNLIKELY (GST_PAD_IS_FLUSHING (pad)))
goto flushing;
}
break;
}
if (G_UNLIKELY ((eventfunc = GST_PAD_EVENTFUNC (pad)) == NULL))
goto no_function;
GST_OBJECT_UNLOCK (pad);
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
result = eventfunc (pad, event);
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
if (need_unlock)
GST_PAD_STREAM_UNLOCK (pad);
return result;
/* ERROR handling */
Some docs updates Original commit message from CVS: * CHANGES-0.9: * docs/design/part-TODO.txt: * docs/design/part-events.txt: Some docs updates * gst/base/gstbasesink.c: (gst_base_sink_handle_object), (gst_base_sink_event), (gst_base_sink_do_sync), (gst_base_sink_activate_push), (gst_base_sink_activate_pull): * gst/base/gstbasesrc.c: (gst_base_src_send_discont), (gst_base_src_do_seek), (gst_base_src_event_handler), (gst_base_src_loop): * gst/base/gstbasetransform.c: (gst_base_transform_transform_caps), (gst_base_transform_configure_caps), (gst_base_transform_setcaps), (gst_base_transform_get_size), (gst_base_transform_buffer_alloc), (gst_base_transform_event), (gst_base_transform_handle_buffer), (gst_base_transform_set_passthrough), (gst_base_transform_is_passthrough): * gst/elements/gstfakesink.c: (gst_fake_sink_event): * gst/elements/gstfilesink.c: (gst_file_sink_event): Event updates. * gst/gstbuffer.h: Use faster casts. * gst/gstelement.c: (gst_element_seek): * gst/gstelement.h: Update gst_element_seek. * gst/gstevent.c: (gst_event_finalize), (_gst_event_copy), (gst_event_new), (gst_event_new_custom), (gst_event_get_structure), (gst_event_new_flush_start), (gst_event_new_flush_stop), (gst_event_new_eos), (gst_event_new_newsegment), (gst_event_parse_newsegment), (gst_event_new_tag), (gst_event_parse_tag), (gst_event_new_filler), (gst_event_new_qos), (gst_event_parse_qos), (gst_event_new_seek), (gst_event_parse_seek), (gst_event_new_navigation): * gst/gstevent.h: Make GstEvent use GstStructure. Add parsing code, make sure the API is sufficiently generic. Mark possible directions of events and serialization. * gst/gstmessage.c: (gst_message_init), (gst_message_finalize), (_gst_message_copy), (gst_message_new_segment_start), (gst_message_new_segment_done), (gst_message_new_custom), (gst_message_parse_segment_start), (gst_message_parse_segment_done): Small cleanups. * gst/gstpad.c: (gst_pad_get_caps_unlocked), (gst_pad_accept_caps), (gst_pad_set_caps), (gst_pad_send_event): Update for new events. Catch events sent in wrong directions. * gst/gstqueue.c: (gst_queue_link_src), (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_handle_src_query): Event updates. * gst/gsttag.c: * gst/gsttag.h: Remove event code from this file. * libs/gst/dataprotocol/dataprotocol.c: (gst_dp_packet_from_event), (gst_dp_event_from_packet): Event updates.
2005-07-27 18:33:03 +00:00
wrong_direction:
{
g_warning ("pad %s:%s sending %s event in wrong direction",
GST_DEBUG_PAD_NAME (pad), GST_EVENT_TYPE_NAME (event));
GST_OBJECT_UNLOCK (pad);
Some docs updates Original commit message from CVS: * CHANGES-0.9: * docs/design/part-TODO.txt: * docs/design/part-events.txt: Some docs updates * gst/base/gstbasesink.c: (gst_base_sink_handle_object), (gst_base_sink_event), (gst_base_sink_do_sync), (gst_base_sink_activate_push), (gst_base_sink_activate_pull): * gst/base/gstbasesrc.c: (gst_base_src_send_discont), (gst_base_src_do_seek), (gst_base_src_event_handler), (gst_base_src_loop): * gst/base/gstbasetransform.c: (gst_base_transform_transform_caps), (gst_base_transform_configure_caps), (gst_base_transform_setcaps), (gst_base_transform_get_size), (gst_base_transform_buffer_alloc), (gst_base_transform_event), (gst_base_transform_handle_buffer), (gst_base_transform_set_passthrough), (gst_base_transform_is_passthrough): * gst/elements/gstfakesink.c: (gst_fake_sink_event): * gst/elements/gstfilesink.c: (gst_file_sink_event): Event updates. * gst/gstbuffer.h: Use faster casts. * gst/gstelement.c: (gst_element_seek): * gst/gstelement.h: Update gst_element_seek. * gst/gstevent.c: (gst_event_finalize), (_gst_event_copy), (gst_event_new), (gst_event_new_custom), (gst_event_get_structure), (gst_event_new_flush_start), (gst_event_new_flush_stop), (gst_event_new_eos), (gst_event_new_newsegment), (gst_event_parse_newsegment), (gst_event_new_tag), (gst_event_parse_tag), (gst_event_new_filler), (gst_event_new_qos), (gst_event_parse_qos), (gst_event_new_seek), (gst_event_parse_seek), (gst_event_new_navigation): * gst/gstevent.h: Make GstEvent use GstStructure. Add parsing code, make sure the API is sufficiently generic. Mark possible directions of events and serialization. * gst/gstmessage.c: (gst_message_init), (gst_message_finalize), (_gst_message_copy), (gst_message_new_segment_start), (gst_message_new_segment_done), (gst_message_new_custom), (gst_message_parse_segment_start), (gst_message_parse_segment_done): Small cleanups. * gst/gstpad.c: (gst_pad_get_caps_unlocked), (gst_pad_accept_caps), (gst_pad_set_caps), (gst_pad_send_event): Update for new events. Catch events sent in wrong directions. * gst/gstqueue.c: (gst_queue_link_src), (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_handle_src_query): Event updates. * gst/gsttag.c: * gst/gsttag.h: Remove event code from this file. * libs/gst/dataprotocol/dataprotocol.c: (gst_dp_packet_from_event), (gst_dp_event_from_packet): Event updates.
2005-07-27 18:33:03 +00:00
gst_event_unref (event);
return FALSE;
}
unknown_direction:
{
g_warning ("pad %s:%s has invalid direction", GST_DEBUG_PAD_NAME (pad));
GST_OBJECT_UNLOCK (pad);
gst_event_unref (event);
return FALSE;
}
no_function:
{
g_warning ("pad %s:%s has no event handler, file a bug.",
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
GST_DEBUG_PAD_NAME (pad));
GST_OBJECT_UNLOCK (pad);
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
gst_event_unref (event);
return FALSE;
}
flushing:
{
GST_OBJECT_UNLOCK (pad);
if (need_unlock)
GST_PAD_STREAM_UNLOCK (pad);
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_CAT_INFO_OBJECT (GST_CAT_EVENT, pad,
"Received event on flushing pad. Discarding");
gst_event_unref (event);
return FALSE;
}
dropping:
{
gst/gstpad.c: Use _DEBUG_OBJECT when it makes sense. Original commit message from CVS: * gst/gstpad.c: (gst_pad_class_init), (gst_pad_dispose), (gst_pad_get_property), (gst_pad_activate_pull), (gst_pad_activate_push), (gst_pad_set_blocked_async), (gst_pad_set_activate_function), (gst_pad_set_activatepull_function), (gst_pad_set_activatepush_function), (gst_pad_set_chain_function), (gst_pad_set_getrange_function), (gst_pad_set_checkgetrange_function), (gst_pad_set_event_function), (gst_pad_set_query_function), (gst_pad_set_query_type_function), (gst_pad_set_internal_link_function), (gst_pad_set_link_function), (gst_pad_set_unlink_function), (gst_pad_set_getcaps_function), (gst_pad_set_acceptcaps_function), (gst_pad_set_fixatecaps_function), (gst_pad_set_setcaps_function), (gst_pad_set_bufferalloc_function), (gst_pad_link_check_hierarchy), (gst_pad_get_caps_unlocked), (gst_pad_get_caps), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_peer_accept_caps), (gst_pad_set_caps), (gst_pad_configure_sink), (gst_pad_configure_src), (gst_pad_get_allowed_caps), (gst_pad_get_negotiated_caps), (gst_pad_buffer_alloc_unchecked), (gst_pad_alloc_buffer_full), (gst_pad_query), (gst_pad_load_and_link), (handle_pad_block), (gst_pad_chain_unchecked), (gst_pad_push), (gst_pad_get_range), (gst_pad_send_event): Use _DEBUG_OBJECT when it makes sense. Protect GST_PAD_CAPS with the OBJECT_LOCK. Small cleanups and code reflows. Avoid caps refcounting in _accept_caps. Refactor alloc_buffer so that the code performed on the peer is in a separate function. Also if the pad does not implement a buffer alloc function, we should still check if the pad is flushing before falling back to the default allocator.
2006-07-31 14:21:10 +00:00
GST_DEBUG_OBJECT (pad, "Dropping event after FALSE probe return");
gst_event_unref (event);
return FALSE;
}
}
/**
* gst_pad_set_element_private:
* @pad: the #GstPad to set the private data of.
* @priv: The private data to attach to the pad.
*
* Set the given private data gpointer on the pad.
* This function can only be used by the element that owns the pad.
* No locking is performed in this function.
*/
void
gst_pad_set_element_private (GstPad * pad, gpointer priv)
{
pad->element_private = priv;
}
/**
* gst_pad_get_element_private:
* @pad: the #GstPad to get the private data of.
*
* Gets the private data of a pad.
* No locking is performed in this function.
*
* Returns: a #gpointer to the private data.
*/
gpointer
gst_pad_get_element_private (GstPad * pad)
{
return pad->element_private;
}
/**
* gst_pad_start_task:
* @pad: the #GstPad to start the task of
* @func: the task function to call
* @data: data passed to the task function
*
* Starts a task that repeadedly calls @func with @data. This function
* is nostly used in the pad activation function to start the
* dataflow. This function will automatically acquire the #GST_PAD_STREAM_LOCK
* of the pad before calling @func.
*
* Returns: a %TRUE if the task could be started.
*/
gboolean
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
gst_pad_start_task (GstPad * pad, GstTaskFunction func, gpointer data)
{
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
GstTask *task;
g_return_val_if_fail (GST_IS_PAD (pad), FALSE);
g_return_val_if_fail (func != NULL, FALSE);
GST_DEBUG_OBJECT (pad, "start task");
GST_OBJECT_LOCK (pad);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
task = GST_PAD_TASK (pad);
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
if (task == NULL) {
task = gst_task_create (func, data);
gst_task_set_lock (task, GST_PAD_GET_STREAM_LOCK (pad));
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
GST_PAD_TASK (pad) = task;
GST_DEBUG_OBJECT (pad, "created task");
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
}
gst_task_start (task);
GST_OBJECT_UNLOCK (pad);
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
return TRUE;
}
/**
* gst_pad_pause_task:
* @pad: the #GstPad to pause the task of
*
* Pause the task of @pad. This function will also make sure that the
* function executed by the task will effectively stop.
*
* Returns: a TRUE if the task could be paused or FALSE when the pad
* has no task.
*/
gboolean
gst_pad_pause_task (GstPad * pad)
{
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
GstTask *task;
g_return_val_if_fail (GST_IS_PAD (pad), FALSE);
GST_DEBUG_OBJECT (pad, "pause task");
GST_OBJECT_LOCK (pad);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
task = GST_PAD_TASK (pad);
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
if (task == NULL)
goto no_task;
gst_task_pause (task);
GST_OBJECT_UNLOCK (pad);
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
GST_PAD_STREAM_LOCK (pad);
GST_PAD_STREAM_UNLOCK (pad);
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
return TRUE;
no_task:
{
GST_DEBUG_OBJECT (pad, "pad has no task");
GST_OBJECT_UNLOCK (pad);
return FALSE;
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
}
}
/**
* gst_pad_stop_task:
* @pad: the #GstPad to stop the task of
*
* Stop the task of @pad. This function will also make sure that the
* function executed by the task will effectively stop if not called
* from the GstTaskFunction.
*
* This function will deadlock if called from the GstTaskFunction of
* the task. Use gst_task_pause() instead.
*
* Regardless of whether the pad has a task, the stream lock is acquired and
* released so as to ensure that streaming through this pad has finished.
*
* Returns: a TRUE if the task could be stopped or FALSE on error.
*/
gst/gstquery.h Original commit message from CVS: 2005-05-06 Andy Wingo <wingo@pobox.com> * gst/gstquery.h * gst/gstquery.c (_gst_query_initialize): Extend GstQuery from GstData, init a memchunk. (standard_definitions): Add a few query types, deprecate a few. (gst_query_get_type): New proc. (_gst_query_copy, _gst_query_free, gst_query_new): GstData implementation. (gst_query_new_application, gst_query_get_structure): New public procs. * docs/design/draft-query.txt: Removed LINKS from the query types, because all the rest can be dispatched to other pads -- seemed ugly to have a query that couldn't be dispatched. internal_links is fine as a pad method. * gst/gstpad.h: Add query2 as a pad method, add the new functions in gstpad.c, but maintain binary compatibility for the moment. Will fix before 0.9 is out. * gst/gstqueryutils.c: * gst/gstqueryutils.h: New files, implement 3 methods for each query type: parse_query, parse_response, and set. Probably need an allocator as well. * gst/gst.h: Add gstquery.h and gstqueryutils.h to the list. * gst/elements/gstfilesink.c (gst_filesink_query2): * gst/base/gstbasesrc.c (gst_basesrc_query2): Replace old query, query_types, and formats methods. * gst/gstpad.c (gst_pad_query2, gst_pad_query2_default) (gst_pad_set_query2_function): New functions. (gst_real_pad_init): Set query2_default as the default query2 function. Basically just dispatches to internally linked pads. Needs review! * gst/gstdata_private.h (_GST_DATA_INIT): Set data->refcount to 1 without using the atomic operations. Only one thread can possibly be accessing the data at this point. Changed so as to avoid gst_atomic operations.
2005-05-06 21:41:22 +00:00
gboolean
gst_pad_stop_task (GstPad * pad)
gst/gstquery.h Original commit message from CVS: 2005-05-06 Andy Wingo <wingo@pobox.com> * gst/gstquery.h * gst/gstquery.c (_gst_query_initialize): Extend GstQuery from GstData, init a memchunk. (standard_definitions): Add a few query types, deprecate a few. (gst_query_get_type): New proc. (_gst_query_copy, _gst_query_free, gst_query_new): GstData implementation. (gst_query_new_application, gst_query_get_structure): New public procs. * docs/design/draft-query.txt: Removed LINKS from the query types, because all the rest can be dispatched to other pads -- seemed ugly to have a query that couldn't be dispatched. internal_links is fine as a pad method. * gst/gstpad.h: Add query2 as a pad method, add the new functions in gstpad.c, but maintain binary compatibility for the moment. Will fix before 0.9 is out. * gst/gstqueryutils.c: * gst/gstqueryutils.h: New files, implement 3 methods for each query type: parse_query, parse_response, and set. Probably need an allocator as well. * gst/gst.h: Add gstquery.h and gstqueryutils.h to the list. * gst/elements/gstfilesink.c (gst_filesink_query2): * gst/base/gstbasesrc.c (gst_basesrc_query2): Replace old query, query_types, and formats methods. * gst/gstpad.c (gst_pad_query2, gst_pad_query2_default) (gst_pad_set_query2_function): New functions. (gst_real_pad_init): Set query2_default as the default query2 function. Basically just dispatches to internally linked pads. Needs review! * gst/gstdata_private.h (_GST_DATA_INIT): Set data->refcount to 1 without using the atomic operations. Only one thread can possibly be accessing the data at this point. Changed so as to avoid gst_atomic operations.
2005-05-06 21:41:22 +00:00
{
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
GstTask *task;
g_return_val_if_fail (GST_IS_PAD (pad), FALSE);
GST_DEBUG_OBJECT (pad, "stop task");
GST_OBJECT_LOCK (pad);
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
task = GST_PAD_TASK (pad);
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
if (task == NULL)
goto no_task;
gst/gstutils.c: RPAD fixes all around. Original commit message from CVS: 2005-06-08 Andy Wingo <wingo@pobox.com> * gst/gstutils.c: RPAD fixes all around. (gst_element_link_pads): Refcounting fixes. * tools/gst-inspect.c: * tools/gst-xmlinspect.c: * parse/grammar.y: * gst/base/gsttypefindhelper.c: * gst/base/gstbasesink.c: * gst/gstqueue.c: RPAD fixes. * gst/gstghostpad.h: * gst/gstghostpad.c: New ghost pad implementation as full proxy pads. The tricky thing is they provide both source and sink interfaces, since they proxy the internal pad for the external pad, and vice versa. Implement with lower-level ProxyPad objects, with the interior proxy pad as a child of the exterior ghost pad. Should write a doc on this. * gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/. (gst_pad_set_name, gst_pad_set_parent): Macros removed, use gst_object API. * gst/gstpad.c: Big changes. No more stub base GstPad, now all pads are real pads. No ghost pads in this file. Not documenting the myriad s/RPAD/PAD/ and REALIZE fixes. (gst_pad_class_init): Add properties for "direction" and "template". Both are construct-only, so they can't change during the life of the pad. Fixes properly deriving from GstPad. (gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For derived objects, just set properties when creating the objects via g_object_new. (gst_pad_get_parent): Implement as a function, return NULL if the parent is not an element. (gst_pad_get_real_parent, gst_pad_add_ghost_pad) (gst_pad_remove_ghost_pad, gst_pad_realize): Removed. * gst/gstobject.c (gst_object_class_init): Make name a construct property. Don't set it in the object init. * gst/gstelement.c (gst_element_add_pad): Don't allow adding pads with UNKNOWN direction. (gst_element_add_ghost_pad): Remove non-orthogonal API. Replace with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)). (gst_element_remove_pad): Remove ghost-pad special cases. (gst_element_pads_activate): Remove rpad cruft. * gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to catch the pad's-parent-not-an-element case. * gst/gst.h: Include gstghostpad.h. * gst/gst.c (init_post): No more real, ghost pads. * gst/Makefile.am: Add gstghostpad.[ch]. * check/Makefile.am: * check/gst/gstbin.c: * check/gst/gstghostpad.c (test_ghost_pads): Check that linking into a bin creates ghost pads, and that the refcounts are right. Partly moved from gstbin.c.
2005-06-08 22:16:27 +00:00
GST_PAD_TASK (pad) = NULL;
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
gst_task_stop (task);
GST_OBJECT_UNLOCK (pad);
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
GST_PAD_STREAM_LOCK (pad);
GST_PAD_STREAM_UNLOCK (pad);
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
if (!gst_task_join (task))
goto join_failed;
gst_object_unref (task);
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
return TRUE;
no_task:
{
GST_DEBUG_OBJECT (pad, "no task");
GST_OBJECT_UNLOCK (pad);
GST_PAD_STREAM_LOCK (pad);
GST_PAD_STREAM_UNLOCK (pad);
/* this is not an error */
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Original commit message from CVS: * gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush): * gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push), (gst_basesink_finish_preroll), (gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate), (gst_basesink_change_state): * gst/base/gstbasesrc.c: (gst_basesrc_do_seek), (gst_basesrc_get_range), (gst_basesrc_loop), (gst_basesrc_activate): * gst/elements/gsttee.c: (gst_tee_sink_activate): * gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init), (gst_real_pad_init), (gst_real_pad_set_property), (gst_real_pad_get_property), (gst_pad_set_active), (gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink), (gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent), (gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps), (gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize), (gst_pad_event_default_dispatch), (gst_pad_event_default), (gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose), (gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain), (gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range), (gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task), (gst_pad_stop_task): * gst/gstpad.h: * gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop), (gst_queue_src_activate): * gst/gsttask.c: (gst_task_init), (gst_task_set_lock), (gst_task_get_state): * gst/gsttask.h: * gst/schedulers/threadscheduler.c: (gst_thread_scheduler_task_start), (gst_thread_scheduler_func): Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function. Remove ACTIVE pad flag, use FLUSHING everywhere Added _pad_chain(), _pad_get_range() to call chain/getrange functions. Add locks around IS_FLUSHING when reading. Take STREAM lock in chain(), get_range() functions so plugins don't need to take it anymore.
2005-05-25 11:50:11 +00:00
return TRUE;
}
join_failed:
{
/* this is bad, possibly the application tried to join the task from
* the task's thread. We install the task again so that it will be stopped
* again from the right thread next time hopefully. */
GST_OBJECT_LOCK (pad);
GST_DEBUG_OBJECT (pad, "join failed");
/* we can only install this task if there was no other task */
if (GST_PAD_TASK (pad) == NULL)
GST_PAD_TASK (pad) = task;
GST_OBJECT_UNLOCK (pad);
return FALSE;
}
}