gstreamer/gst/gstghostpad.c

1313 lines
36 KiB
C
Raw Normal View History

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
/* GStreamer
* Copyright (C) 1999,2000 Erik Walthinsen <omega@cse.ogi.edu>
* 2000 Wim Taymans <wtay@chello.be>
* 2005 Andy Wingo <wingo@pobox.com>
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
* 2006 Edward Hervey <bilboed@bilboed.com>
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
*
* gstghostpad.c: Proxy pads
*
* 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:gstghostpad
* @short_description: Pseudo link pads
* @see_also: #GstPad
*
* GhostPads are useful when organizing pipelines with #GstBin like elements.
* The idea here is to create hierarchical element graphs. The bin element
* contains a sub-graph. Now one would like to treat the bin-element like any
* other #GstElement. This is where GhostPads come into play. A GhostPad acts as
* a proxy for another pad. Thus the bin can have sink and source ghost-pads
* that are associated with sink and source pads of the child elements.
*
* If the target pad is known at creation time, gst_ghost_pad_new() is the
* function to use to get a ghost-pad. Otherwise one can use gst_ghost_pad_new_no_target()
* to create the ghost-pad and use gst_ghost_pad_set_target() to establish the
* association later on.
*
* Note that GhostPads add overhead to the data processing of a pipeline.
*
* Last reviewed on 2005-11-18 (0.9.5)
*/
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 "gst_private.h"
#include "gstinfo.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 "gstghostpad.h"
#include "gst.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
#define GST_CAT_DEFAULT GST_CAT_PADS
#define GST_PROXY_PAD_CAST(obj) ((GstProxyPad *)obj)
#define GST_PROXY_PAD_PRIVATE(obj) (GST_PROXY_PAD_CAST (obj)->priv)
#define GST_PROXY_PAD_TARGET(pad) (GST_PROXY_PAD_PRIVATE (pad)->target)
#define GST_PROXY_PAD_INTERNAL(pad) (GST_PROXY_PAD_PRIVATE (pad)->internal)
#define GST_PROXY_PAD_RETARGET(pad) (GST_PROXY_PAD_PRIVATE (pad)->retarget)
#define GST_PROXY_GET_LOCK(pad) (GST_PROXY_PAD_PRIVATE (pad)->proxy_lock)
#define GST_PROXY_LOCK(pad) (g_mutex_lock (GST_PROXY_GET_LOCK (pad)))
#define GST_PROXY_UNLOCK(pad) (g_mutex_unlock (GST_PROXY_GET_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
struct _GstProxyPadPrivate
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
{
/* with PROXY_LOCK */
GMutex *proxy_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
GstPad *target;
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
GstPad *internal;
gboolean retarget;
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_DEFINE_TYPE (GstProxyPad, gst_proxy_pad, GST_TYPE_PAD);
static GstPad *gst_proxy_pad_get_target (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 void gst_proxy_pad_dispose (GObject * object);
static void gst_proxy_pad_finalize (GObject * object);
/**
* gst_proxy_pad_query_type_default:
* @pad: a #GstPad.
*
* Invoke the default query type handler of the proxy pad.
*
* Returns: (transfer none) (array zero-terminated=1): a zero-terminated array
* of #GstQueryType.
*
* Since: 0.10.35
*/
const GstQueryType *
gst_proxy_pad_query_type_default (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 *target;
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
const GstQueryType *res = NULL;
g_return_val_if_fail (GST_IS_PROXY_PAD (pad), NULL);
if (!(target = gst_proxy_pad_get_target (pad)))
goto no_target;
res = gst_pad_get_query_types (target);
gst_object_unref (target);
return res;
/* ERRORS */
no_target:
{
GST_DEBUG_OBJECT (pad, "no target pad");
return FALSE;
}
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_proxy_pad_event_default:
* @pad: a #GstPad to push the event to.
* @event: (transfer full): the #GstEvent to send to the pad.
*
* Invoke the default event of the proxy pad.
*
* Returns: TRUE if the event was handled.
*
* Since: 0.10.35
*/
gboolean
gst_proxy_pad_event_default (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
{
gboolean res;
GstPad *internal;
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_PROXY_PAD (pad), FALSE);
g_return_val_if_fail (GST_IS_EVENT (event), FALSE);
internal = GST_PROXY_PAD_INTERNAL (pad);
res = gst_pad_push_event (internal, event);
return 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
}
/**
* gst_proxy_pad_query_default:
* @pad: a #GstPad to invoke the default query on.
* @query: (transfer none): the #GstQuery to perform.
*
* Invoke the default query function of the proxy pad.
*
* Returns: TRUE if the query could be performed.
*
* Since: 0.10.35
*/
gboolean
gst_proxy_pad_query_default (GstPad * pad, GstQuery * 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
{
gboolean res;
GstPad *target;
g_return_val_if_fail (GST_IS_PROXY_PAD (pad), FALSE);
g_return_val_if_fail (GST_IS_QUERY (query), FALSE);
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 (!(target = gst_proxy_pad_get_target (pad)))
goto no_target;
res = gst_pad_query (target, query);
gst_object_unref (target);
return res;
/* ERRORS */
no_target:
{
GST_DEBUG_OBJECT (pad, "no target pad");
return FALSE;
}
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_proyx_pad_iterate_internal_links_default:
* @pad: the #GstPad to get the internal links of.
*
* Invoke the default iterate internal links function of the proxy pad.
*
* Returns: a #GstIterator of #GstPad, or NULL if @pad has no parent. Unref each
* returned pad with gst_object_unref().
*
* Since: 0.10.35
*/
GstIterator *
gst_proxy_pad_iterate_internal_links_default (GstPad * pad)
{
GstIterator *res = NULL;
GstPad *internal;
GValue v = { 0, };
g_return_val_if_fail (GST_IS_PROXY_PAD (pad), NULL);
internal = GST_PROXY_PAD_INTERNAL (pad);
g_value_init (&v, GST_TYPE_PAD);
g_value_set_object (&v, internal);
res = gst_iterator_new_single (GST_TYPE_PAD, &v);
g_value_unset (&v);
gst_object_unref (internal);
return res;
}
/**
* gst_proxy_pad_chain_default:
* @pad: a sink #GstPad, returns GST_FLOW_ERROR if not.
* @buffer: (transfer full): the #GstBuffer to send, return GST_FLOW_ERROR
* if not.
*
* Invoke the default chain function of the proxy pad.
*
* Returns: a #GstFlowReturn from the pad.
*
* Since: 0.10.35
*/
GstFlowReturn
gst_proxy_pad_chain_default (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
{
GstFlowReturn res;
GstPad *internal;
g_return_val_if_fail (GST_IS_PROXY_PAD (pad), GST_FLOW_ERROR);
g_return_val_if_fail (GST_IS_BUFFER (buffer), GST_FLOW_ERROR);
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
internal = GST_PROXY_PAD_INTERNAL (pad);
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
res = gst_pad_push (internal, buffer);
return 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
}
/**
* gst_proxy_pad_chain_list_default:
* @pad: a sink #GstPad, returns GST_FLOW_ERROR if not.
* @list: (transfer full): the #GstBufferList to send, return GST_FLOW_ERROR
* if not.
*
* Invoke the default chain list function of the proxy pad.
*
* Returns: a #GstFlowReturn from the pad.
*
* Since: 0.10.35
*/
GstFlowReturn
gst_proxy_pad_chain_list_default (GstPad * pad, GstBufferList * list)
{
GstFlowReturn res;
GstPad *internal;
g_return_val_if_fail (GST_IS_PROXY_PAD (pad), GST_FLOW_ERROR);
g_return_val_if_fail (GST_IS_BUFFER_LIST (list), GST_FLOW_ERROR);
internal = GST_PROXY_PAD_INTERNAL (pad);
res = gst_pad_push_list (internal, list);
return res;
}
/**
* gst_proxy_pad_get_range_default:
* @pad: a src #GstPad, returns #GST_FLOW_ERROR if not.
* @offset: The start offset of the buffer
* @size: The length of the buffer
* @buffer: (out callee-allocates): a pointer to hold the #GstBuffer,
* returns #GST_FLOW_ERROR if %NULL.
*
* Invoke the default getrange function of the proxy pad.
*
* Returns: a #GstFlowReturn from the pad.
*
* Since: 0.10.35
*/
GstFlowReturn
gst_proxy_pad_getrange_default (GstPad * pad, guint64 offset, guint size,
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
GstBuffer ** buffer)
{
GstFlowReturn res;
GstPad *internal;
g_return_val_if_fail (GST_IS_PROXY_PAD (pad), GST_FLOW_ERROR);
g_return_val_if_fail (buffer != NULL, GST_FLOW_ERROR);
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
internal = GST_PROXY_PAD_INTERNAL (pad);
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
res = gst_pad_pull_range (internal, offset, size, buffer);
return 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
}
/**
* gst_proxy_pad_getcaps_default:
* @pad: a #GstPad to get the capabilities of.
* @filter: a #GstCaps filter.
*
* Invoke the default getcaps function of the proxy pad.
*
* Returns: (transfer full): the caps of the pad with incremented ref-count
*
* Since: 0.10.35
*/
GstCaps *
gst_proxy_pad_getcaps_default (GstPad * pad, GstCaps * filter)
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 *target;
GstCaps *res;
GstPadTemplate *templ;
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_PROXY_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
templ = GST_PAD_PAD_TEMPLATE (pad);
target = gst_proxy_pad_get_target (pad);
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
if (target) {
/* if we have a real target, proxy the call */
res = gst_pad_get_caps (target, filter);
GST_DEBUG_OBJECT (pad, "get caps of target %s:%s : %" GST_PTR_FORMAT,
GST_DEBUG_PAD_NAME (target), res);
gst_object_unref (target);
/* filter against the template */
if (templ && res) {
GstCaps *filt, *tmp;
filt = GST_PAD_TEMPLATE_CAPS (templ);
if (filt) {
tmp = gst_caps_intersect_full (res, filt, GST_CAPS_INTERSECT_FIRST);
gst_caps_unref (res);
res = tmp;
GST_DEBUG_OBJECT (pad,
"filtered against template gives %" GST_PTR_FORMAT, res);
}
}
} else {
/* else, if we have a template, use its caps. */
if (templ) {
res = GST_PAD_TEMPLATE_CAPS (templ);
GST_DEBUG_OBJECT (pad,
"using pad template %p with caps %p %" GST_PTR_FORMAT, templ, res,
res);
res = gst_caps_ref (res);
if (filter) {
GstCaps *intersection =
gst_caps_intersect_full (filter, res, GST_CAPS_INTERSECT_FIRST);
gst_caps_unref (res);
res = intersection;
}
goto done;
}
/* last resort, any caps */
GST_DEBUG_OBJECT (pad, "pad has no template, returning ANY");
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
res = gst_caps_new_any ();
}
done:
return 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
}
/**
* gst_proxy_pad_acceptcaps_default:
* @pad: a #GstPad to check
* @caps: a #GstCaps to check on the pad
*
* Invoke the default acceptcaps function of the proxy pad.
*
* Returns: TRUE if the pad can accept the caps.
*
* Since: 0.10.35
*/
gboolean
gst_proxy_pad_acceptcaps_default (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 *target;
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
g_return_val_if_fail (GST_IS_PROXY_PAD (pad), FALSE);
g_return_val_if_fail (caps == NULL || GST_IS_CAPS (caps), FALSE);
target = gst_proxy_pad_get_target (pad);
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
if (target) {
res = gst_pad_accept_caps (target, caps);
gst_object_unref (target);
} else {
GST_DEBUG_OBJECT (pad, "no target");
/* We don't have a target, we return TRUE and we assume that any future
* target will be able to deal with any configured caps. */
res = TRUE;
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
}
return 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
}
/**
* gst_proxy_pad_fixatecaps_default:
* @pad: a #GstPad to fixate
* @caps: the #GstCaps to fixate
*
* Invoke the default fixatecaps function of the proxy pad.
*
* Since: 0.10.35
*/
void
gst_proxy_pad_fixatecaps_default (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 *target;
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_PROXY_PAD (pad));
g_return_if_fail (GST_IS_CAPS (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 (!(target = gst_proxy_pad_get_target (pad)))
goto no_target;
gst_pad_fixate_caps (target, caps);
gst_object_unref (target);
return;
/* ERRORS */
no_target:
{
GST_DEBUG_OBJECT (pad, "no target");
return;
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/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 gboolean
gst_proxy_pad_set_target_unlocked (GstPad * pad, GstPad * target)
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 *oldtarget;
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
if (target) {
GST_LOG_OBJECT (pad, "setting target %s:%s", GST_DEBUG_PAD_NAME (target));
if (G_UNLIKELY (GST_PAD_DIRECTION (pad) != GST_PAD_DIRECTION (target)))
goto wrong_direction;
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
} else
GST_LOG_OBJECT (pad, "clearing target");
/* clear old target */
if ((oldtarget = GST_PROXY_PAD_TARGET (pad)))
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_object_unref (oldtarget);
/* set and ref new target if any */
if (target)
GST_PROXY_PAD_TARGET (pad) = gst_object_ref (target);
else
GST_PROXY_PAD_TARGET (pad) = NULL;
return TRUE;
/* ERRORS */
wrong_direction:
{
GST_ERROR_OBJECT (pad,
"target pad doesn't have the same direction as ourself");
return FALSE;
}
}
static gboolean
gst_proxy_pad_set_target (GstPad * pad, GstPad * target)
{
gboolean result;
GST_PROXY_LOCK (pad);
result = gst_proxy_pad_set_target_unlocked (pad, target);
GST_PROXY_UNLOCK (pad);
return 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
}
static GstPad *
gst_proxy_pad_get_target (GstPad * pad)
{
GstPad *target;
GST_PROXY_LOCK (pad);
target = GST_PROXY_PAD_TARGET (pad);
if (target)
gst_object_ref (target);
GST_PROXY_UNLOCK (pad);
return target;
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_proxy_pad_get_internal:
* @pad: the #GstProxyPad
*
* Get the internal pad of @pad. Unref target pad after usage.
*
* The internal pad of a #GstGhostPad is the internally used
* pad of opposite direction, which is used to link to the target.
*
* Returns: (transfer full): the target #GstProxyPad, can be NULL.
* Unref target pad after usage.
*
* Since: 0.10.35
*/
GstProxyPad *
gst_proxy_pad_get_internal (GstProxyPad * pad)
{
GstPad *internal;
g_return_val_if_fail (GST_IS_PROXY_PAD (pad), NULL);
GST_PROXY_LOCK (pad);
internal = GST_PROXY_PAD_INTERNAL (pad);
if (internal)
gst_object_ref (internal);
GST_PROXY_UNLOCK (pad);
return GST_PROXY_PAD_CAST (internal);
}
/**
* gst_proxy_pad_unlink_default:
* @pad: a #GstPad to unlink
*
* Invoke the default unlink function of the proxy pad.
*
* Since: 0.10.35
*/
void
gst_proxy_pad_unlink_default (GstPad * pad)
{
GstPad *internal;
/* don't do anything if this unlink resulted from retargeting the pad
* controlled by the ghostpad. We only want to invalidate the target pad when
* the element suddenly unlinked with our internal pad. */
if (GST_PROXY_PAD_RETARGET (pad))
return;
internal = GST_PROXY_PAD_INTERNAL (pad);
GST_DEBUG_OBJECT (pad, "pad is unlinked");
gst_proxy_pad_set_target (internal, NULL);
}
static void
gst_proxy_pad_class_init (GstProxyPadClass * klass)
{
GObjectClass *gobject_class = (GObjectClass *) klass;
g_type_class_add_private (klass, sizeof (GstProxyPadPrivate));
gobject_class->dispose = gst_proxy_pad_dispose;
gobject_class->finalize = gst_proxy_pad_finalize;
/* Register common function pointer descriptions */
GST_DEBUG_REGISTER_FUNCPTR (gst_proxy_pad_query_type_default);
GST_DEBUG_REGISTER_FUNCPTR (gst_proxy_pad_event_default);
GST_DEBUG_REGISTER_FUNCPTR (gst_proxy_pad_query_default);
GST_DEBUG_REGISTER_FUNCPTR (gst_proxy_pad_iterate_internal_links_default);
GST_DEBUG_REGISTER_FUNCPTR (gst_proxy_pad_getcaps_default);
GST_DEBUG_REGISTER_FUNCPTR (gst_proxy_pad_acceptcaps_default);
GST_DEBUG_REGISTER_FUNCPTR (gst_proxy_pad_fixatecaps_default);
GST_DEBUG_REGISTER_FUNCPTR (gst_proxy_pad_unlink_default);
GST_DEBUG_REGISTER_FUNCPTR (gst_proxy_pad_chain_default);
GST_DEBUG_REGISTER_FUNCPTR (gst_proxy_pad_chain_list_default);
GST_DEBUG_REGISTER_FUNCPTR (gst_proxy_pad_getrange_default);
}
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_proxy_pad_dispose (GObject * object)
{
GstPad *pad = GST_PAD (object);
GstPad **target_p;
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_PROXY_LOCK (pad);
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
/* remove and unref the target */
target_p = &GST_PROXY_PAD_TARGET (pad);
gst_object_replace ((GstObject **) target_p, NULL);
/* The internal is only cleared by GstGhostPad::dispose, since it is the
* parent of non-ghost GstProxyPad and owns the refcount on the internal.
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_PROXY_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
G_OBJECT_CLASS (gst_proxy_pad_parent_class)->dispose (object);
}
static void
gst_proxy_pad_finalize (GObject * object)
{
GstProxyPad *pad = GST_PROXY_PAD (object);
g_mutex_free (GST_PROXY_GET_LOCK (pad));
GST_PROXY_GET_LOCK (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 (gst_proxy_pad_parent_class)->finalize (object);
}
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
static void
gst_proxy_pad_init (GstProxyPad * ppad)
{
GstPad *pad = (GstPad *) ppad;
GST_PROXY_PAD_PRIVATE (ppad) = G_TYPE_INSTANCE_GET_PRIVATE (ppad,
GST_TYPE_PROXY_PAD, GstProxyPadPrivate);
GST_PROXY_GET_LOCK (pad) = g_mutex_new ();
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_pad_set_query_type_function (pad, gst_proxy_pad_query_type_default);
gst_pad_set_event_function (pad, gst_proxy_pad_event_default);
gst_pad_set_query_function (pad, gst_proxy_pad_query_default);
gst_pad_set_iterate_internal_links_function (pad,
gst_proxy_pad_iterate_internal_links_default);
gst_pad_set_getcaps_function (pad, gst_proxy_pad_getcaps_default);
gst_pad_set_acceptcaps_function (pad, gst_proxy_pad_acceptcaps_default);
gst_pad_set_fixatecaps_function (pad, gst_proxy_pad_fixatecaps_default);
gst_pad_set_unlink_function (pad, gst_proxy_pad_unlink_default);
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/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
/***********************************************************************
* Ghost pads, implemented as a pair of proxy pads (sort of)
*/
#define GST_GHOST_PAD_PRIVATE(obj) (GST_GHOST_PAD_CAST (obj)->priv)
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
struct _GstGhostPadPrivate
{
/* with PROXY_LOCK */
gboolean constructed;
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_DEFINE_TYPE (GstGhostPad, gst_ghost_pad, GST_TYPE_PROXY_PAD);
static void gst_ghost_pad_dispose (GObject * object);
/**
* gst_ghost_pad_internal_activate_push_default:
* @pad: the #GstPad to activate or deactivate.
* @active: whether the pad should be active or not.
*
* Invoke the default activate push function of a proxy pad that is
* owned by a ghost pad.
*
* Returns: %TRUE if the operation was successful.
*
* Since: 0.10.35
*/
gboolean
gst_ghost_pad_internal_activate_push_default (GstPad * pad, gboolean 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
{
gboolean ret;
GstPad *other;
g_return_val_if_fail (GST_IS_PROXY_PAD (pad), FALSE);
GST_LOG_OBJECT (pad, "%sactivate push on %s:%s, we're ok",
(active ? "" : "de"), GST_DEBUG_PAD_NAME (pad));
/* in both cases (SRC and SINK) we activate just the internal pad. The targets
* will be activated later (or already in case of a ghost sinkpad). */
other = GST_PROXY_PAD_INTERNAL (pad);
ret = gst_pad_activate_push (other, 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
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_ghost_pad_internal_activate_pull_default:
* @pad: the #GstPad to activate or deactivate.
* @active: whether the pad should be active or not.
*
* Invoke the default activate pull function of a proxy pad that is
* owned by a ghost pad.
*
* Returns: %TRUE if the operation was successful.
*
* Since: 0.10.35
*/
gboolean
gst_ghost_pad_internal_activate_pull_default (GstPad * pad, gboolean active)
{
gboolean ret;
GstPad *other;
g_return_val_if_fail (GST_IS_PROXY_PAD (pad), FALSE);
GST_LOG_OBJECT (pad, "%sactivate pull on %s:%s", (active ? "" : "de"),
GST_DEBUG_PAD_NAME (pad));
if (GST_PAD_DIRECTION (pad) == GST_PAD_SRC) {
/* we are activated in pull mode by our peer element, which is a sinkpad
* that wants to operate in pull mode. This activation has to propagate
* upstream through the pipeline. We call the internal activation function,
* which will trigger gst_ghost_pad_activate_pull_default, which propagates even
* further upstream */
GST_LOG_OBJECT (pad, "pad is src, activate internal");
other = GST_PROXY_PAD_INTERNAL (pad);
ret = gst_pad_activate_pull (other, active);
} else if (G_LIKELY ((other = gst_pad_get_peer (pad)))) {
/* We are SINK, the ghostpad is SRC, we propagate the activation upstream
* since we hold a pointer to the upstream peer. */
GST_LOG_OBJECT (pad, "activating peer");
ret = gst_pad_activate_pull (other, active);
gst_object_unref (other);
} else {
/* this is failure, we can't activate pull if there is no peer */
GST_LOG_OBJECT (pad, "not src and no peer, failing");
ret = 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
return ret;
}
/**
* gst_ghost_pad_activate_push_default:
* @pad: the #GstPad to activate or deactivate.
* @active: whether the pad should be active or not.
*
* Invoke the default activate push function of a ghost pad.
*
* Returns: %TRUE if the operation was successful.
*
* Since: 0.10.35
*/
gboolean
gst_ghost_pad_activate_push_default (GstPad * pad, gboolean active)
{
gboolean ret;
GstPad *other;
g_return_val_if_fail (GST_IS_GHOST_PAD (pad), FALSE);
GST_LOG_OBJECT (pad, "%sactivate push on %s:%s, proxy internal",
(active ? "" : "de"), GST_DEBUG_PAD_NAME (pad));
/* just activate the internal pad */
other = GST_PROXY_PAD_INTERNAL (pad);
ret = gst_pad_activate_push (other, active);
return ret;
}
/**
* gst_ghost_pad_activate_pull_default:
* @pad: the #GstPad to activate or deactivate.
* @active: whether the pad should be active or not.
*
* Invoke the default activate pull function of a ghost pad.
*
* Returns: %TRUE if the operation was successful.
*
* Since: 0.10.35
*/
gboolean
gst_ghost_pad_activate_pull_default (GstPad * pad, gboolean active)
{
gboolean ret;
GstPad *other;
g_return_val_if_fail (GST_IS_GHOST_PAD (pad), FALSE);
GST_LOG_OBJECT (pad, "%sactivate pull on %s:%s", (active ? "" : "de"),
GST_DEBUG_PAD_NAME (pad));
if (GST_PAD_DIRECTION (pad) == GST_PAD_SRC) {
/* the ghostpad is SRC and activated in pull mode by its peer, call the
* activation function of the internal pad to propagate the activation
* upstream */
GST_LOG_OBJECT (pad, "pad is src, activate internal");
other = GST_PROXY_PAD_INTERNAL (pad);
ret = gst_pad_activate_pull (other, active);
} else if (G_LIKELY ((other = gst_pad_get_peer (pad)))) {
/* We are SINK and activated by the internal pad, propagate activation
* upstream because we hold a ref to the upstream peer */
GST_LOG_OBJECT (pad, "activating peer");
ret = gst_pad_activate_pull (other, active);
gst_object_unref (other);
} else {
/* no peer, we fail */
GST_LOG_OBJECT (pad, "pad not src and no peer, failing");
ret = FALSE;
}
return ret;
}
/**
* gst_ghost_pad_link_default:
* @pad: the #GstPad to link.
* @peer: the #GstPad peer
*
* Invoke the default link function of a ghost pad.
*
* Returns: #GstPadLinkReturn of the operation
*
* Since: 0.10.35
*/
GstPadLinkReturn
gst_ghost_pad_link_default (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
{
GstPadLinkReturn ret;
GstPad *internal;
g_return_val_if_fail (GST_IS_GHOST_PAD (pad), GST_PAD_LINK_REFUSED);
g_return_val_if_fail (GST_IS_PAD (peer), GST_PAD_LINK_REFUSED);
GST_DEBUG_OBJECT (pad, "linking ghostpad");
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
internal = GST_PROXY_PAD_INTERNAL (pad);
if (!gst_proxy_pad_set_target (internal, peer))
goto target_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
ret = GST_PAD_LINK_OK;
/* if we are a source pad, we should call the peer link function
* if the peer has one, see design docs. */
if (GST_PAD_IS_SRC (pad)) {
if (GST_PAD_LINKFUNC (peer)) {
ret = GST_PAD_LINKFUNC (peer) (peer, pad);
if (ret != GST_PAD_LINK_OK)
goto link_failed;
}
}
return ret;
/* ERRORS */
target_failed:
{
GST_DEBUG_OBJECT (pad, "setting target failed");
return GST_PAD_LINK_REFUSED;
}
link_failed:
{
GST_DEBUG_OBJECT (pad, "linking failed");
/* clear target again */
gst_proxy_pad_set_target (internal, NULL);
return ret;
}
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_ghost_pad_unlink_default:
* @pad: the #GstPad to link.
*
* Invoke the default unlink function of a ghost pad.
*
* Since: 0.10.35
*/
void
gst_ghost_pad_unlink_default (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 *internal;
g_return_if_fail (GST_IS_GHOST_PAD (pad));
internal = GST_PROXY_PAD_INTERNAL (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_DEBUG_OBJECT (pad, "unlinking ghostpad");
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
/* The target of the internal pad is no longer valid */
gst_proxy_pad_set_target (internal, 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
}
static void
gst_ghost_pad_class_init (GstGhostPadClass * klass)
{
GObjectClass *gobject_class = (GObjectClass *) klass;
g_type_class_add_private (klass, sizeof (GstGhostPadPrivate));
gobject_class->dispose = gst_ghost_pad_dispose;
GST_DEBUG_REGISTER_FUNCPTR (gst_ghost_pad_activate_pull_default);
GST_DEBUG_REGISTER_FUNCPTR (gst_ghost_pad_activate_push_default);
GST_DEBUG_REGISTER_FUNCPTR (gst_ghost_pad_link_default);
}
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_ghost_pad_init (GstGhostPad * pad)
{
GST_GHOST_PAD_PRIVATE (pad) = G_TYPE_INSTANCE_GET_PRIVATE (pad,
GST_TYPE_GHOST_PAD, GstGhostPadPrivate);
gst_pad_set_activatepull_function (GST_PAD_CAST (pad),
gst_ghost_pad_activate_pull_default);
gst_pad_set_activatepush_function (GST_PAD_CAST (pad),
gst_ghost_pad_activate_push_default);
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_ghost_pad_dispose (GObject * object)
{
GstPad *pad;
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
GstPad *internal;
GstPad *peer;
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
pad = GST_PAD (object);
GST_DEBUG_OBJECT (pad, "dispose");
gst_ghost_pad_set_target (GST_GHOST_PAD (pad), NULL);
/* Unlink here so that gst_pad_dispose doesn't. That would lead to a call to
* gst_ghost_pad_unlink_default when the ghost pad is in an inconsistent state */
peer = gst_pad_get_peer (pad);
if (peer) {
if (GST_PAD_IS_SRC (pad))
gst_pad_unlink (pad, peer);
else
gst_pad_unlink (peer, pad);
gst_object_unref (peer);
}
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_PROXY_LOCK (pad);
internal = GST_PROXY_PAD_INTERNAL (pad);
gst_pad_set_activatepull_function (internal, NULL);
gst_pad_set_activatepush_function (internal, NULL);
/* disposes of the internal pad, since the ghostpad is the only possible object
* that has a refcount on the internal pad. */
gst_object_unparent (GST_OBJECT_CAST (internal));
GST_PROXY_PAD_INTERNAL (pad) = NULL;
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_PROXY_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
G_OBJECT_CLASS (gst_ghost_pad_parent_class)->dispose (object);
}
/**
* gst_ghost_pad_construct:
* @gpad: the newly allocated ghost pad
*
* Finish initialization of a newly allocated ghost pad.
*
* This function is most useful in language bindings and when subclassing
* #GstGhostPad; plugin and application developers normally will not call this
* function. Call this function directly after a call to g_object_new
* (GST_TYPE_GHOST_PAD, "direction", @dir, ..., NULL).
*
* Returns: %TRUE if the construction succeeds, %FALSE otherwise.
*
* Since: 0.10.22
*/
gboolean
gst_ghost_pad_construct (GstGhostPad * gpad)
{
GstPadDirection dir, otherdir;
GstPadTemplate *templ;
GstPad *pad, *internal;
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
g_return_val_if_fail (GST_IS_GHOST_PAD (gpad), FALSE);
g_return_val_if_fail (GST_GHOST_PAD_PRIVATE (gpad)->constructed == FALSE,
FALSE);
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
g_object_get (gpad, "direction", &dir, "template", &templ, NULL);
g_return_val_if_fail (dir != GST_PAD_UNKNOWN, FALSE);
pad = GST_PAD (gpad);
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
/* Set directional padfunctions for ghostpad */
if (dir == GST_PAD_SINK) {
gst_pad_set_chain_function (pad, gst_proxy_pad_chain_default);
gst_pad_set_chain_list_function (pad, gst_proxy_pad_chain_list_default);
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
} else {
gst_pad_set_getrange_function (pad, gst_proxy_pad_getrange_default);
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
}
/* link/unlink functions */
gst_pad_set_link_function (pad, gst_ghost_pad_link_default);
gst_pad_set_unlink_function (pad, gst_ghost_pad_unlink_default);
/* INTERNAL PAD, it always exists and is child of the ghostpad */
otherdir = (dir == GST_PAD_SRC) ? GST_PAD_SINK : GST_PAD_SRC;
if (templ) {
internal =
g_object_new (GST_TYPE_PROXY_PAD, "name", NULL,
"direction", otherdir, "template", templ, NULL);
/* release ref obtained via g_object_get */
gst_object_unref (templ);
} else {
internal =
g_object_new (GST_TYPE_PROXY_PAD, "name", NULL,
"direction", otherdir, NULL);
}
GST_PAD_UNSET_FLUSHING (internal);
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
/* Set directional padfunctions for internal pad */
if (dir == GST_PAD_SRC) {
gst_pad_set_chain_function (internal, gst_proxy_pad_chain_default);
gst_pad_set_chain_list_function (internal,
gst_proxy_pad_chain_list_default);
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
} else {
gst_pad_set_getrange_function (internal, gst_proxy_pad_getrange_default);
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_PROXY_LOCK (pad);
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
/* now make the ghostpad a parent of the internal pad */
if (!gst_object_set_parent (GST_OBJECT_CAST (internal),
GST_OBJECT_CAST (pad)))
goto parent_failed;
/* The ghostpad is the parent of the internal pad and is the only object that
* can have a refcount on the internal pad.
* At this point, the GstGhostPad has a refcount of 1, and the internal pad has
* a refcount of 1.
* When the refcount of the GstGhostPad drops to 0, the ghostpad will dispose
2009-08-31 15:56:22 +00:00
* its refcount on the internal pad in the dispose method by un-parenting it.
* This is why we don't take extra refcounts in the assignments below
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_PROXY_PAD_INTERNAL (pad) = internal;
GST_PROXY_PAD_INTERNAL (internal) = pad;
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
/* special activation functions for the internal pad */
gst_pad_set_activatepull_function (internal,
gst_ghost_pad_internal_activate_pull_default);
gst_pad_set_activatepush_function (internal,
gst_ghost_pad_internal_activate_push_default);
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_PROXY_UNLOCK (pad);
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_GHOST_PAD_PRIVATE (gpad)->constructed = TRUE;
return TRUE;
/* ERRORS */
parent_failed:
{
GST_WARNING_OBJECT (gpad, "Could not set internal pad %s:%s",
GST_DEBUG_PAD_NAME (internal));
g_critical ("Could not set internal pad %s:%s",
GST_DEBUG_PAD_NAME (internal));
GST_PROXY_UNLOCK (pad);
gst_object_unref (internal);
return FALSE;
}
}
static GstPad *
gst_ghost_pad_new_full (const gchar * name, GstPadDirection dir,
GstPadTemplate * templ)
{
GstGhostPad *ret;
g_return_val_if_fail (dir != GST_PAD_UNKNOWN, NULL);
/* OBJECT CREATION */
if (templ) {
ret = g_object_new (GST_TYPE_GHOST_PAD, "name", name,
"direction", dir, "template", templ, NULL);
} else {
ret = g_object_new (GST_TYPE_GHOST_PAD, "name", name,
"direction", dir, NULL);
}
if (!gst_ghost_pad_construct (ret))
goto construct_failed;
return GST_PAD_CAST (ret);
construct_failed:
/* already logged */
gst_object_unref (ret);
return NULL;
}
/**
* gst_ghost_pad_new_no_target:
* @name: (allow-none): the name of the new pad, or NULL to assign a default name.
* @dir: the direction of the ghostpad
*
* Create a new ghostpad without a target with the given direction.
* A target can be set on the ghostpad later with the
* gst_ghost_pad_set_target() function.
*
* The created ghostpad will not have a padtemplate.
*
* Returns: (transfer full): a new #GstPad, or NULL in case of an error.
*/
GstPad *
gst_ghost_pad_new_no_target (const gchar * name, GstPadDirection dir)
{
GstPad *ret;
g_return_val_if_fail (dir != GST_PAD_UNKNOWN, NULL);
GST_LOG ("name:%s, direction:%d", GST_STR_NULL (name), dir);
ret = gst_ghost_pad_new_full (name, dir, NULL);
return ret;
}
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_ghost_pad_new:
* @name: (allow-none): the name of the new pad, or NULL to assign a default name
* @target: (transfer none): the pad to ghost.
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
*
* Create a new ghostpad with @target as the target. The direction will be taken
* from the target pad. @target must be 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
* Will ref the target.
*
* Returns: (transfer full): a new #GstPad, or NULL in case of an error.
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 *
gst_ghost_pad_new (const gchar * name, GstPad * target)
{
GstPad *ret;
g_return_val_if_fail (GST_IS_PAD (target), NULL);
g_return_val_if_fail (!gst_pad_is_linked (target), 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_LOG ("name:%s, target:%s:%s", GST_STR_NULL (name),
GST_DEBUG_PAD_NAME (target));
if ((ret = gst_ghost_pad_new_no_target (name, GST_PAD_DIRECTION (target))))
if (!gst_ghost_pad_set_target (GST_GHOST_PAD_CAST (ret), target))
goto set_target_failed;
return ret;
/* ERRORS */
set_target_failed:
{
GST_WARNING_OBJECT (ret, "failed to set target %s:%s",
GST_DEBUG_PAD_NAME (target));
gst_object_unref (ret);
return NULL;
}
}
/**
* gst_ghost_pad_new_from_template:
* @name: (allow-none): the name of the new pad, or NULL to assign a default name.
* @target: (transfer none): the pad to ghost.
* @templ: (transfer none): the #GstPadTemplate to use on the ghostpad.
*
* Create a new ghostpad with @target as the target. The direction will be taken
* from the target pad. The template used on the ghostpad will be @template.
*
* Will ref the target.
*
* Returns: (transfer full): a new #GstPad, or NULL in case of an error.
*
* Since: 0.10.10
*/
GstPad *
gst_ghost_pad_new_from_template (const gchar * name, GstPad * target,
GstPadTemplate * templ)
{
GstPad *ret;
g_return_val_if_fail (GST_IS_PAD (target), NULL);
g_return_val_if_fail (!gst_pad_is_linked (target), NULL);
g_return_val_if_fail (templ != NULL, NULL);
g_return_val_if_fail (GST_PAD_TEMPLATE_DIRECTION (templ) ==
GST_PAD_DIRECTION (target), NULL);
GST_LOG ("name:%s, target:%s:%s, templ:%p", GST_STR_NULL (name),
GST_DEBUG_PAD_NAME (target), templ);
if ((ret = gst_ghost_pad_new_full (name, GST_PAD_DIRECTION (target), templ)))
if (!gst_ghost_pad_set_target (GST_GHOST_PAD_CAST (ret), target))
goto set_target_failed;
return ret;
/* ERRORS */
set_target_failed:
{
GST_WARNING_OBJECT (ret, "failed to set target %s:%s",
GST_DEBUG_PAD_NAME (target));
gst_object_unref (ret);
return NULL;
}
}
/**
* gst_ghost_pad_new_no_target_from_template:
* @name: (allow-none): the name of the new pad, or NULL to assign a default name
* @templ: (transfer none): the #GstPadTemplate to create the ghostpad from.
*
* Create a new ghostpad based on @templ, without setting a target. The
* direction will be taken from the @templ.
*
* Returns: (transfer full): a new #GstPad, or NULL in case of an error.
*
* Since: 0.10.10
*/
GstPad *
gst_ghost_pad_new_no_target_from_template (const gchar * name,
GstPadTemplate * templ)
{
GstPad *ret;
g_return_val_if_fail (templ != NULL, NULL);
ret =
gst_ghost_pad_new_full (name, GST_PAD_TEMPLATE_DIRECTION (templ), templ);
return ret;
}
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_ghost_pad_get_target:
* @gpad: the #GstGhostPad
*
* Get the target pad of @gpad. Unref target pad after usage.
*
* Returns: (transfer full): the target #GstPad, can be NULL if the ghostpad
* has no target set. Unref target pad after usage.
*/
GstPad *
gst_ghost_pad_get_target (GstGhostPad * gpad)
{
GstPad *ret;
g_return_val_if_fail (GST_IS_GHOST_PAD (gpad), 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
ret = gst_proxy_pad_get_target (GST_PAD_CAST (gpad));
GST_DEBUG_OBJECT (gpad, "get target %s:%s", GST_DEBUG_PAD_NAME (ret));
return ret;
}
/**
* gst_ghost_pad_set_target:
* @gpad: the #GstGhostPad
* @newtarget: (transfer none) (allow-none): the new pad target
*
* Set the new target of the ghostpad @gpad. Any existing target
* is unlinked and links to the new target are established. if @newtarget is
* NULL the target will be cleared.
*
* Returns: (transfer full): TRUE if the new target could be set. This function
* can return FALSE when the internal pads could not be linked.
*/
gboolean
gst_ghost_pad_set_target (GstGhostPad * gpad, GstPad * newtarget)
{
GstPad *internal;
GstPad *oldtarget;
gboolean result;
GstPadLinkReturn lret;
g_return_val_if_fail (GST_IS_GHOST_PAD (gpad), FALSE);
g_return_val_if_fail (GST_PAD_CAST (gpad) != newtarget, FALSE);
g_return_val_if_fail (newtarget != GST_PROXY_PAD_INTERNAL (gpad), FALSE);
/* no need for locking, the internal pad's lifecycle is directly linked to the
* ghostpad's */
internal = GST_PROXY_PAD_INTERNAL (gpad);
if (newtarget)
GST_DEBUG_OBJECT (gpad, "set target %s:%s", GST_DEBUG_PAD_NAME (newtarget));
else
GST_DEBUG_OBJECT (gpad, "clearing target");
/* clear old target */
GST_PROXY_LOCK (gpad);
if ((oldtarget = GST_PROXY_PAD_TARGET (gpad))) {
GST_PROXY_PAD_RETARGET (internal) = TRUE;
/* unlink internal pad */
if (GST_PAD_IS_SRC (internal))
gst_pad_unlink (internal, oldtarget);
else
gst_pad_unlink (oldtarget, internal);
GST_PROXY_PAD_RETARGET (internal) = FALSE;
}
result = gst_proxy_pad_set_target_unlocked (GST_PAD_CAST (gpad), newtarget);
GST_PROXY_UNLOCK (gpad);
if (result && newtarget) {
/* and link to internal pad without any checks */
GST_DEBUG_OBJECT (gpad, "connecting internal pad to target");
if (GST_PAD_IS_SRC (internal))
lret =
gst_pad_link_full (internal, newtarget, GST_PAD_LINK_CHECK_NOTHING);
else
lret =
gst_pad_link_full (newtarget, internal, GST_PAD_LINK_CHECK_NOTHING);
if (lret != GST_PAD_LINK_OK)
goto link_failed;
}
return result;
/* ERRORS */
link_failed:
{
GST_WARNING_OBJECT (gpad, "could not link internal and target, reason:%d",
lret);
/* and unset target again */
2009-11-05 13:02:28 +00:00
GST_PROXY_LOCK (gpad);
gst_proxy_pad_set_target_unlocked (GST_PAD_CAST (gpad), NULL);
2009-11-05 13:02:28 +00:00
GST_PROXY_UNLOCK (gpad);
return FALSE;
}
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
}