gstreamer/plugins/elements/gstqueue.c

748 lines
23 KiB
C
Raw Normal View History

/* GStreamer
* Copyright (C) 1999,2000 Erik Walthinsen <omega@cse.ogi.edu>
* 2000 Wim Taymans <wtay@chello.be>
*
* gstqueue.c:
*
* 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.
*/
/* #define DEBUG_ENABLED */
/* #define STATUS_ENABLED */
#ifdef STATUS_ENABLED
#define STATUS(A) GST_DEBUG(GST_CAT_DATAFLOW, A, GST_ELEMENT_NAME(queue))
#else
#define STATUS(A)
#endif
#include "config.h"
#include "gst_private.h"
#include "gstqueue.h"
#include "gstscheduler.h"
#include "gstevent.h"
#include "gstlog.h"
GstElementDetails gst_queue_details = {
"Queue",
"Generic",
"LGPL",
"Simple data queue",
VERSION,
"Erik Walthinsen <omega@cse.ogi.edu>",
"(C) 1999",
};
/* Queue signals and args */
enum {
LOW_WATERMARK,
HIGH_WATERMARK,
LAST_SIGNAL
};
enum {
ARG_0,
ARG_LEVEL_BUFFERS,
ARG_LEVEL_BYTES,
ARG_LEVEL_TIME,
ARG_SIZE_BUFFERS,
ARG_SIZE_BYTES,
ARG_SIZE_TIME,
ARG_LEAKY,
ARG_LEVEL,
ARG_MAX_LEVEL,
This is an attempt at not segfaulting on errors but reporting some usefull info instead. Original commit message from CVS: This is an attempt at not segfaulting on errors but reporting some usefull info instead. - bin changes so errors can propagate. - changed the _FAST macros to _CAST because that is what they do. - removed all references to cothreads out of the core, they are really a scheduler issue, handler with a sched_private gpointer. - added a live buffer count, for debugging buffer leaks. - added error checking in gst_scheduler_state_transition this solves the "out of cothreads" problem. - GST_ELEMENT_NO_ENTRY == GST_ELEMENT_INFINITE_LOOP - added 2 private element flasg for use by the scheduler (_COTHREAD_STOPPING) is now - added scheduler entry points: - _yield : to create possible scheduling points. - _interrupt: to stop execution of an element. - _error: to signal en error condition to the scheduler. - improved error messages for pads. - signal gst_element_error where appropriate. - added the a new bin to the parent before entering it so one can reference its children. - queue memleak fixes on dispose. - added possible deadlock detection in queue (turned off be default) - GstBasicScheduler is a real class of its own now, hiding its internal variables. - GST_ELEMENT_IS_COTHREAD_STOPPING is gone. either call explicit _yield operations, or make a sane loop. - Better state change handling in filesrc. Better error reporting/recovery too. - updated core plugins. - detect non decoupled elements on scheduler boundries and error.
2001-12-22 21:18:17 +00:00
ARG_MAY_DEADLOCK,
ARG_BLOCK_TIMEOUT,
};
static void gst_queue_class_init (GstQueueClass *klass);
static void gst_queue_init (GstQueue *queue);
This is an attempt at not segfaulting on errors but reporting some usefull info instead. Original commit message from CVS: This is an attempt at not segfaulting on errors but reporting some usefull info instead. - bin changes so errors can propagate. - changed the _FAST macros to _CAST because that is what they do. - removed all references to cothreads out of the core, they are really a scheduler issue, handler with a sched_private gpointer. - added a live buffer count, for debugging buffer leaks. - added error checking in gst_scheduler_state_transition this solves the "out of cothreads" problem. - GST_ELEMENT_NO_ENTRY == GST_ELEMENT_INFINITE_LOOP - added 2 private element flasg for use by the scheduler (_COTHREAD_STOPPING) is now - added scheduler entry points: - _yield : to create possible scheduling points. - _interrupt: to stop execution of an element. - _error: to signal en error condition to the scheduler. - improved error messages for pads. - signal gst_element_error where appropriate. - added the a new bin to the parent before entering it so one can reference its children. - queue memleak fixes on dispose. - added possible deadlock detection in queue (turned off be default) - GstBasicScheduler is a real class of its own now, hiding its internal variables. - GST_ELEMENT_IS_COTHREAD_STOPPING is gone. either call explicit _yield operations, or make a sane loop. - Better state change handling in filesrc. Better error reporting/recovery too. - updated core plugins. - detect non decoupled elements on scheduler boundries and error.
2001-12-22 21:18:17 +00:00
static void gst_queue_dispose (GObject *object);
static void gst_queue_set_property (GObject *object, guint prop_id,
const GValue *value, GParamSpec *pspec);
static void gst_queue_get_property (GObject *object, guint prop_id,
GValue *value, GParamSpec *pspec);
static void gst_queue_chain (GstPad *pad, GstBuffer *buf);
static GstBuffer * gst_queue_get (GstPad *pad);
static GstBufferPool* gst_queue_get_bufferpool (GstPad *pad);
static gboolean gst_queue_handle_src_event (GstPad *pad, GstEvent *event);
static void gst_queue_locked_flush (GstQueue *queue);
static GstElementStateReturn gst_queue_change_state (GstElement *element);
static gboolean gst_queue_release_locks (GstElement *element);
#define GST_TYPE_QUEUE_LEAKY (queue_leaky_get_type())
static GType
queue_leaky_get_type(void) {
static GType queue_leaky_type = 0;
static GEnumValue queue_leaky[] = {
{ GST_QUEUE_NO_LEAK, "0", "Not Leaky" },
{ GST_QUEUE_LEAK_UPSTREAM, "1", "Leaky on Upstream" },
{ GST_QUEUE_LEAK_DOWNSTREAM, "2", "Leaky on Downstream" },
{ 0, NULL, NULL },
};
if (!queue_leaky_type) {
queue_leaky_type = g_enum_register_static("GstQueueLeaky", queue_leaky);
}
return queue_leaky_type;
}
static GstElementClass *parent_class = NULL;
/* static guint gst_queue_signals[LAST_SIGNAL] = { 0 }; */
GType
gst_queue_get_type(void)
{
static GType queue_type = 0;
if (!queue_type) {
static const GTypeInfo queue_info = {
sizeof(GstQueueClass),
NULL,
NULL,
(GClassInitFunc)gst_queue_class_init,
NULL,
NULL,
sizeof(GstQueue),
4,
(GInstanceInitFunc)gst_queue_init,
NULL
};
queue_type = g_type_register_static (GST_TYPE_ELEMENT, "GstQueue", &queue_info, 0);
}
return queue_type;
}
static void
gst_queue_class_init (GstQueueClass *klass)
{
GObjectClass *gobject_class;
GstElementClass *gstelement_class;
gobject_class = (GObjectClass*)klass;
gstelement_class = (GstElementClass*)klass;
parent_class = g_type_class_ref (GST_TYPE_ELEMENT);
This is an attempt at not segfaulting on errors but reporting some usefull info instead. Original commit message from CVS: This is an attempt at not segfaulting on errors but reporting some usefull info instead. - bin changes so errors can propagate. - changed the _FAST macros to _CAST because that is what they do. - removed all references to cothreads out of the core, they are really a scheduler issue, handler with a sched_private gpointer. - added a live buffer count, for debugging buffer leaks. - added error checking in gst_scheduler_state_transition this solves the "out of cothreads" problem. - GST_ELEMENT_NO_ENTRY == GST_ELEMENT_INFINITE_LOOP - added 2 private element flasg for use by the scheduler (_COTHREAD_STOPPING) is now - added scheduler entry points: - _yield : to create possible scheduling points. - _interrupt: to stop execution of an element. - _error: to signal en error condition to the scheduler. - improved error messages for pads. - signal gst_element_error where appropriate. - added the a new bin to the parent before entering it so one can reference its children. - queue memleak fixes on dispose. - added possible deadlock detection in queue (turned off be default) - GstBasicScheduler is a real class of its own now, hiding its internal variables. - GST_ELEMENT_IS_COTHREAD_STOPPING is gone. either call explicit _yield operations, or make a sane loop. - Better state change handling in filesrc. Better error reporting/recovery too. - updated core plugins. - detect non decoupled elements on scheduler boundries and error.
2001-12-22 21:18:17 +00:00
g_object_class_install_property (G_OBJECT_CLASS (klass), ARG_LEAKY,
g_param_spec_enum ("leaky", "Leaky", "Where the queue leaks, if at all.",
GST_TYPE_QUEUE_LEAKY, GST_QUEUE_NO_LEAK, G_PARAM_READWRITE));
g_object_class_install_property (G_OBJECT_CLASS (klass), ARG_LEVEL,
g_param_spec_int ("level", "Level", "How many buffers are in the queue.",
0, G_MAXINT, 0, G_PARAM_READABLE));
g_object_class_install_property (G_OBJECT_CLASS (klass), ARG_MAX_LEVEL,
g_param_spec_int ("max_level", "Maximum Level", "How many buffers the queue holds.",
0, G_MAXINT, 100, G_PARAM_READWRITE));
g_object_class_install_property (G_OBJECT_CLASS (klass), ARG_MAY_DEADLOCK,
g_param_spec_boolean ("may_deadlock", "May Deadlock", "The queue may deadlock if it's full and not PLAYING",
TRUE, G_PARAM_READWRITE));
g_object_class_install_property (G_OBJECT_CLASS (klass), ARG_BLOCK_TIMEOUT,
g_param_spec_int ("block_timeout", "Timeout for Block",
"Microseconds until blocked queue times out and returns filler event. "
"Value of -1 disables timeout",
-1, G_MAXINT, -1, G_PARAM_READWRITE));
This is an attempt at not segfaulting on errors but reporting some usefull info instead. Original commit message from CVS: This is an attempt at not segfaulting on errors but reporting some usefull info instead. - bin changes so errors can propagate. - changed the _FAST macros to _CAST because that is what they do. - removed all references to cothreads out of the core, they are really a scheduler issue, handler with a sched_private gpointer. - added a live buffer count, for debugging buffer leaks. - added error checking in gst_scheduler_state_transition this solves the "out of cothreads" problem. - GST_ELEMENT_NO_ENTRY == GST_ELEMENT_INFINITE_LOOP - added 2 private element flasg for use by the scheduler (_COTHREAD_STOPPING) is now - added scheduler entry points: - _yield : to create possible scheduling points. - _interrupt: to stop execution of an element. - _error: to signal en error condition to the scheduler. - improved error messages for pads. - signal gst_element_error where appropriate. - added the a new bin to the parent before entering it so one can reference its children. - queue memleak fixes on dispose. - added possible deadlock detection in queue (turned off be default) - GstBasicScheduler is a real class of its own now, hiding its internal variables. - GST_ELEMENT_IS_COTHREAD_STOPPING is gone. either call explicit _yield operations, or make a sane loop. - Better state change handling in filesrc. Better error reporting/recovery too. - updated core plugins. - detect non decoupled elements on scheduler boundries and error.
2001-12-22 21:18:17 +00:00
gobject_class->dispose = GST_DEBUG_FUNCPTR (gst_queue_dispose);
gobject_class->set_property = GST_DEBUG_FUNCPTR (gst_queue_set_property);
gobject_class->get_property = GST_DEBUG_FUNCPTR (gst_queue_get_property);
gstelement_class->change_state = GST_DEBUG_FUNCPTR(gst_queue_change_state);
gstelement_class->release_locks = GST_DEBUG_FUNCPTR(gst_queue_release_locks);
}
static GstPadLinkReturn
gst_queue_link (GstPad *pad, GstCaps *caps)
{
GstQueue *queue = GST_QUEUE (gst_pad_get_parent (pad));
GstPad *otherpad;
if (pad == queue->srcpad)
otherpad = queue->sinkpad;
else
otherpad = queue->srcpad;
return gst_pad_proxy_link (otherpad, caps);
}
static GstCaps*
gst_queue_getcaps (GstPad *pad, GstCaps *caps)
{
GstQueue *queue = GST_QUEUE (gst_pad_get_parent (pad));
GstPad *otherpad;
if (pad == queue->srcpad)
otherpad = GST_PAD_PEER (queue->sinkpad);
else
otherpad = GST_PAD_PEER (queue->srcpad);
if (otherpad)
return gst_pad_get_caps (otherpad);
return NULL;
}
static void
gst_queue_init (GstQueue *queue)
{
/* scheduling on this kind of element is, well, interesting */
GST_FLAG_SET (queue, GST_ELEMENT_DECOUPLED);
GST_FLAG_SET (queue, GST_ELEMENT_EVENT_AWARE);
queue->sinkpad = gst_pad_new ("sink", GST_PAD_SINK);
gst_pad_set_chain_function (queue->sinkpad, GST_DEBUG_FUNCPTR (gst_queue_chain));
gst_element_add_pad (GST_ELEMENT (queue), queue->sinkpad);
gst_pad_set_bufferpool_function (queue->sinkpad, GST_DEBUG_FUNCPTR (gst_queue_get_bufferpool));
gst_pad_set_link_function (queue->sinkpad, GST_DEBUG_FUNCPTR (gst_queue_link));
gst_pad_set_getcaps_function (queue->sinkpad, GST_DEBUG_FUNCPTR (gst_queue_getcaps));
queue->srcpad = gst_pad_new ("src", GST_PAD_SRC);
gst_pad_set_get_function (queue->srcpad, GST_DEBUG_FUNCPTR (gst_queue_get));
gst_element_add_pad (GST_ELEMENT (queue), queue->srcpad);
gst_pad_set_link_function (queue->srcpad, GST_DEBUG_FUNCPTR (gst_queue_link));
gst_pad_set_getcaps_function (queue->srcpad, GST_DEBUG_FUNCPTR (gst_queue_getcaps));
gst_pad_set_event_function (queue->srcpad, GST_DEBUG_FUNCPTR (gst_queue_handle_src_event));
queue->leaky = GST_QUEUE_NO_LEAK;
queue->queue = NULL;
queue->level_buffers = 0;
queue->level_bytes = 0;
queue->level_time = G_GINT64_CONSTANT (0);
queue->size_buffers = 100; /* 100 buffers */
queue->size_bytes = 100 * 1024; /* 100KB */
queue->size_time = G_GINT64_CONSTANT (1000000000); /* 1sec */
This is an attempt at not segfaulting on errors but reporting some usefull info instead. Original commit message from CVS: This is an attempt at not segfaulting on errors but reporting some usefull info instead. - bin changes so errors can propagate. - changed the _FAST macros to _CAST because that is what they do. - removed all references to cothreads out of the core, they are really a scheduler issue, handler with a sched_private gpointer. - added a live buffer count, for debugging buffer leaks. - added error checking in gst_scheduler_state_transition this solves the "out of cothreads" problem. - GST_ELEMENT_NO_ENTRY == GST_ELEMENT_INFINITE_LOOP - added 2 private element flasg for use by the scheduler (_COTHREAD_STOPPING) is now - added scheduler entry points: - _yield : to create possible scheduling points. - _interrupt: to stop execution of an element. - _error: to signal en error condition to the scheduler. - improved error messages for pads. - signal gst_element_error where appropriate. - added the a new bin to the parent before entering it so one can reference its children. - queue memleak fixes on dispose. - added possible deadlock detection in queue (turned off be default) - GstBasicScheduler is a real class of its own now, hiding its internal variables. - GST_ELEMENT_IS_COTHREAD_STOPPING is gone. either call explicit _yield operations, or make a sane loop. - Better state change handling in filesrc. Better error reporting/recovery too. - updated core plugins. - detect non decoupled elements on scheduler boundries and error.
2001-12-22 21:18:17 +00:00
queue->may_deadlock = TRUE;
queue->block_timeout = -1;
queue->interrupt = FALSE;
queue->flush = FALSE;
queue->qlock = g_mutex_new ();
queue->not_empty = g_cond_new ();
queue->not_full = g_cond_new ();
queue->events = g_async_queue_new();
queue->queue = g_queue_new ();
GST_DEBUG_ELEMENT (GST_CAT_THREAD, queue, "initialized queue's not_empty & not_full conditions");
}
This is an attempt at not segfaulting on errors but reporting some usefull info instead. Original commit message from CVS: This is an attempt at not segfaulting on errors but reporting some usefull info instead. - bin changes so errors can propagate. - changed the _FAST macros to _CAST because that is what they do. - removed all references to cothreads out of the core, they are really a scheduler issue, handler with a sched_private gpointer. - added a live buffer count, for debugging buffer leaks. - added error checking in gst_scheduler_state_transition this solves the "out of cothreads" problem. - GST_ELEMENT_NO_ENTRY == GST_ELEMENT_INFINITE_LOOP - added 2 private element flasg for use by the scheduler (_COTHREAD_STOPPING) is now - added scheduler entry points: - _yield : to create possible scheduling points. - _interrupt: to stop execution of an element. - _error: to signal en error condition to the scheduler. - improved error messages for pads. - signal gst_element_error where appropriate. - added the a new bin to the parent before entering it so one can reference its children. - queue memleak fixes on dispose. - added possible deadlock detection in queue (turned off be default) - GstBasicScheduler is a real class of its own now, hiding its internal variables. - GST_ELEMENT_IS_COTHREAD_STOPPING is gone. either call explicit _yield operations, or make a sane loop. - Better state change handling in filesrc. Better error reporting/recovery too. - updated core plugins. - detect non decoupled elements on scheduler boundries and error.
2001-12-22 21:18:17 +00:00
static void
gst_queue_dispose (GObject *object)
{
GstQueue *queue = GST_QUEUE (object);
gst_element_set_state (GST_ELEMENT (queue), GST_STATE_NULL);
This is an attempt at not segfaulting on errors but reporting some usefull info instead. Original commit message from CVS: This is an attempt at not segfaulting on errors but reporting some usefull info instead. - bin changes so errors can propagate. - changed the _FAST macros to _CAST because that is what they do. - removed all references to cothreads out of the core, they are really a scheduler issue, handler with a sched_private gpointer. - added a live buffer count, for debugging buffer leaks. - added error checking in gst_scheduler_state_transition this solves the "out of cothreads" problem. - GST_ELEMENT_NO_ENTRY == GST_ELEMENT_INFINITE_LOOP - added 2 private element flasg for use by the scheduler (_COTHREAD_STOPPING) is now - added scheduler entry points: - _yield : to create possible scheduling points. - _interrupt: to stop execution of an element. - _error: to signal en error condition to the scheduler. - improved error messages for pads. - signal gst_element_error where appropriate. - added the a new bin to the parent before entering it so one can reference its children. - queue memleak fixes on dispose. - added possible deadlock detection in queue (turned off be default) - GstBasicScheduler is a real class of its own now, hiding its internal variables. - GST_ELEMENT_IS_COTHREAD_STOPPING is gone. either call explicit _yield operations, or make a sane loop. - Better state change handling in filesrc. Better error reporting/recovery too. - updated core plugins. - detect non decoupled elements on scheduler boundries and error.
2001-12-22 21:18:17 +00:00
g_mutex_free (queue->qlock);
g_cond_free (queue->not_empty);
g_cond_free (queue->not_full);
g_queue_free (queue->queue);
This is an attempt at not segfaulting on errors but reporting some usefull info instead. Original commit message from CVS: This is an attempt at not segfaulting on errors but reporting some usefull info instead. - bin changes so errors can propagate. - changed the _FAST macros to _CAST because that is what they do. - removed all references to cothreads out of the core, they are really a scheduler issue, handler with a sched_private gpointer. - added a live buffer count, for debugging buffer leaks. - added error checking in gst_scheduler_state_transition this solves the "out of cothreads" problem. - GST_ELEMENT_NO_ENTRY == GST_ELEMENT_INFINITE_LOOP - added 2 private element flasg for use by the scheduler (_COTHREAD_STOPPING) is now - added scheduler entry points: - _yield : to create possible scheduling points. - _interrupt: to stop execution of an element. - _error: to signal en error condition to the scheduler. - improved error messages for pads. - signal gst_element_error where appropriate. - added the a new bin to the parent before entering it so one can reference its children. - queue memleak fixes on dispose. - added possible deadlock detection in queue (turned off be default) - GstBasicScheduler is a real class of its own now, hiding its internal variables. - GST_ELEMENT_IS_COTHREAD_STOPPING is gone. either call explicit _yield operations, or make a sane loop. - Better state change handling in filesrc. Better error reporting/recovery too. - updated core plugins. - detect non decoupled elements on scheduler boundries and error.
2001-12-22 21:18:17 +00:00
g_async_queue_unref(queue->events);
This is an attempt at not segfaulting on errors but reporting some usefull info instead. Original commit message from CVS: This is an attempt at not segfaulting on errors but reporting some usefull info instead. - bin changes so errors can propagate. - changed the _FAST macros to _CAST because that is what they do. - removed all references to cothreads out of the core, they are really a scheduler issue, handler with a sched_private gpointer. - added a live buffer count, for debugging buffer leaks. - added error checking in gst_scheduler_state_transition this solves the "out of cothreads" problem. - GST_ELEMENT_NO_ENTRY == GST_ELEMENT_INFINITE_LOOP - added 2 private element flasg for use by the scheduler (_COTHREAD_STOPPING) is now - added scheduler entry points: - _yield : to create possible scheduling points. - _interrupt: to stop execution of an element. - _error: to signal en error condition to the scheduler. - improved error messages for pads. - signal gst_element_error where appropriate. - added the a new bin to the parent before entering it so one can reference its children. - queue memleak fixes on dispose. - added possible deadlock detection in queue (turned off be default) - GstBasicScheduler is a real class of its own now, hiding its internal variables. - GST_ELEMENT_IS_COTHREAD_STOPPING is gone. either call explicit _yield operations, or make a sane loop. - Better state change handling in filesrc. Better error reporting/recovery too. - updated core plugins. - detect non decoupled elements on scheduler boundries and error.
2001-12-22 21:18:17 +00:00
G_OBJECT_CLASS (parent_class)->dispose (object);
}
static GstBufferPool*
gst_queue_get_bufferpool (GstPad *pad)
{
GstQueue *queue;
queue = GST_QUEUE (GST_OBJECT_PARENT (pad));
return gst_pad_get_bufferpool (queue->srcpad);
}
static void
gst_queue_cleanup_data (gpointer data, const gpointer user_data)
{
GST_DEBUG_ELEMENT (GST_CAT_DATAFLOW, user_data, "cleaning buffer %p", data);
gst_data_unref (GST_DATA (data));
}
static void
gst_queue_locked_flush (GstQueue *queue)
{
gpointer data;
while ((data = g_queue_pop_head (queue->queue))) {
gst_queue_cleanup_data (data, (gpointer) queue);
}
queue->timeval = NULL;
queue->level_buffers = 0;
queue->level_bytes = 0;
queue->level_time = G_GINT64_CONSTANT (0);
/* make sure any pending buffers to be added are flushed too */
queue->flush = TRUE;
/* signal not_full, since we apparently aren't full anymore */
g_cond_signal (queue->not_full);
}
static void
gst_queue_chain (GstPad *pad, GstBuffer *buf)
{
GstQueue *queue;
g_return_if_fail (pad != NULL);
g_return_if_fail (GST_IS_PAD (pad));
g_return_if_fail (buf != NULL);
queue = GST_QUEUE (GST_OBJECT_PARENT (pad));
/* check for events to send upstream */
g_async_queue_lock(queue->events);
while (g_async_queue_length_unlocked(queue->events) > 0){
GstEvent *event = (GstEvent*)g_async_queue_pop_unlocked(queue->events);
GST_DEBUG_ELEMENT (GST_CAT_DATAFLOW, queue, "sending event upstream\n");
gst_pad_event_default (pad, event);
GST_DEBUG_ELEMENT (GST_CAT_DATAFLOW, queue, "event sent\n");
}
g_async_queue_unlock(queue->events);
- some fixes to int2float making automake 1.5 happy (gst now requires automake1.5). It's still not perfect but it bui... Original commit message from CVS: - added playondemand plugin by Leif Morgan Johnson <lmjohns3@eos.ncsu.edu> - some fixes to int2float - aplied a patch from wrobell <wrobell@ite.pl> that is a first attempt at making automake 1.5 happy (gst now requires automake1.5). It's still not perfect but it builds. - Made the schedulers plugable. The default scheduler now lives inside a plugin. - Added a new mpeg1/2 parser/demuxer. - Fixed some compiler warnings in the core libs. - substantial work to GstThread (hopefully less race conditions). simplified the code in GstThread a bit. A state change can now also happen in the thread context. - reworked the state semantics of a bin. it'll now automatically get the highest state of its children. - the autoplugger now nests the threads so that a state change failure of one thread doesn't make its upstream thread lock. - GstQueue refuses to go to PLAYING if the sinkpad is not connected. This way the queue will not wedge in the _get lock. - GstQueue unlocks its mutexes when going to PAUSED. - make sure that when all elements in a bin/thread go to PAUSED, the bin is set to PAUSED too. - make a parent bin wait for its children to PAUSE before ending the iteration with FALSE (EOS) - Some changes to GstPlay to deal with EOS. - aplied the latest patch from Zeenix to gstrtp. end result: GstPlay doesn't crash on EOS and the pipeline is now shut down properly.
2001-12-04 22:12:50 +00:00
restart:
/* we have to lock the queue since we span threads */
GST_DEBUG_ELEMENT (GST_CAT_DATAFLOW, queue, "locking t:%p", g_thread_self ());
g_mutex_lock (queue->qlock);
GST_DEBUG_ELEMENT (GST_CAT_DATAFLOW, queue, "locked t:%p", g_thread_self ());
/* assume don't need to flush this buffer when the queue is filled */
queue->flush = FALSE;
if (GST_IS_EVENT (buf)) {
switch (GST_EVENT_TYPE (buf)) {
case GST_EVENT_FLUSH:
GST_DEBUG_ELEMENT (GST_CAT_DATAFLOW, queue, "FLUSH event, flushing queue\n");
gst_queue_locked_flush (queue);
break;
case GST_EVENT_EOS:
- some fixes to int2float making automake 1.5 happy (gst now requires automake1.5). It's still not perfect but it bui... Original commit message from CVS: - added playondemand plugin by Leif Morgan Johnson <lmjohns3@eos.ncsu.edu> - some fixes to int2float - aplied a patch from wrobell <wrobell@ite.pl> that is a first attempt at making automake 1.5 happy (gst now requires automake1.5). It's still not perfect but it builds. - Made the schedulers plugable. The default scheduler now lives inside a plugin. - Added a new mpeg1/2 parser/demuxer. - Fixed some compiler warnings in the core libs. - substantial work to GstThread (hopefully less race conditions). simplified the code in GstThread a bit. A state change can now also happen in the thread context. - reworked the state semantics of a bin. it'll now automatically get the highest state of its children. - the autoplugger now nests the threads so that a state change failure of one thread doesn't make its upstream thread lock. - GstQueue refuses to go to PLAYING if the sinkpad is not connected. This way the queue will not wedge in the _get lock. - GstQueue unlocks its mutexes when going to PAUSED. - make sure that when all elements in a bin/thread go to PAUSED, the bin is set to PAUSED too. - make a parent bin wait for its children to PAUSE before ending the iteration with FALSE (EOS) - Some changes to GstPlay to deal with EOS. - aplied the latest patch from Zeenix to gstrtp. end result: GstPlay doesn't crash on EOS and the pipeline is now shut down properly.
2001-12-04 22:12:50 +00:00
GST_DEBUG_ELEMENT (GST_CAT_DATAFLOW, queue, "eos in on %s %d\n",
GST_ELEMENT_NAME (queue), queue->level_buffers);
break;
default:
/* we put the event in the queue, we don't have to act ourselves */
break;
}
}
GST_DEBUG_ELEMENT (GST_CAT_DATAFLOW, queue, "adding buffer %p of size %d",buf,GST_BUFFER_SIZE(buf));
if (queue->level_buffers == queue->size_buffers) {
/* if this is a leaky queue... */
if (queue->leaky) {
/* FIXME don't want to leak events! */
/* if we leak on the upstream side, drop the current buffer */
if (queue->leaky == GST_QUEUE_LEAK_UPSTREAM) {
GST_DEBUG_ELEMENT (GST_CAT_DATAFLOW, queue, "queue is full, leaking buffer on upstream end");
if (GST_IS_EVENT (buf))
fprintf(stderr, "Error: queue [%s] leaked an event, type:%d\n",
GST_ELEMENT_NAME(GST_ELEMENT(queue)),
GST_EVENT_TYPE(GST_EVENT(buf)));
GST_DEBUG_ELEMENT (GST_CAT_DATAFLOW, queue, "queue is full, leaking buffer on upstream end");
/* now we have to clean up and exit right away */
g_mutex_unlock (queue->qlock);
goto out_unref;
}
/* otherwise we have to push a buffer off the other end */
else {
gpointer front;
GstBuffer *leakbuf;
GST_DEBUG_ELEMENT (GST_CAT_DATAFLOW, queue, "queue is full, leaking buffer on downstream end");
front = g_queue_pop_head (queue->queue);
leakbuf = (GstBuffer *)(front);
if (GST_IS_EVENT (leakbuf)) {
fprintf(stderr, "Error: queue [%s] leaked an event, type:%d\n",
GST_ELEMENT_NAME(GST_ELEMENT(queue)),
GST_EVENT_TYPE(GST_EVENT(leakbuf)));
}
queue->level_buffers--;
queue->level_bytes -= GST_BUFFER_SIZE(leakbuf);
gst_data_unref (GST_DATA (leakbuf));
}
}
GST_DEBUG_ELEMENT (GST_CAT_DATAFLOW, queue, "pre full wait, level:%d/%d",
queue->level_buffers, queue->size_buffers);
while (queue->level_buffers == queue->size_buffers) {
/* if there's a pending state change for this queue or its manager, switch */
/* back to iterator so bottom half of state change executes */
if (queue->interrupt) {
GST_DEBUG_ELEMENT (GST_CAT_DATAFLOW, queue, "interrupted!!");
g_mutex_unlock (queue->qlock);
if (gst_scheduler_interrupt (gst_pad_get_scheduler (queue->sinkpad), GST_ELEMENT (queue)))
goto out_unref;
/* if we got here bacause we were unlocked after a flush, we don't need
* to add the buffer to the queue again */
if (queue->flush) {
GST_DEBUG_ELEMENT (GST_CAT_DATAFLOW, queue, "not adding pending buffer after flush");
goto out_unref;
}
GST_DEBUG_ELEMENT (GST_CAT_DATAFLOW, queue, "adding pending buffer after interrupt");
- some fixes to int2float making automake 1.5 happy (gst now requires automake1.5). It's still not perfect but it bui... Original commit message from CVS: - added playondemand plugin by Leif Morgan Johnson <lmjohns3@eos.ncsu.edu> - some fixes to int2float - aplied a patch from wrobell <wrobell@ite.pl> that is a first attempt at making automake 1.5 happy (gst now requires automake1.5). It's still not perfect but it builds. - Made the schedulers plugable. The default scheduler now lives inside a plugin. - Added a new mpeg1/2 parser/demuxer. - Fixed some compiler warnings in the core libs. - substantial work to GstThread (hopefully less race conditions). simplified the code in GstThread a bit. A state change can now also happen in the thread context. - reworked the state semantics of a bin. it'll now automatically get the highest state of its children. - the autoplugger now nests the threads so that a state change failure of one thread doesn't make its upstream thread lock. - GstQueue refuses to go to PLAYING if the sinkpad is not connected. This way the queue will not wedge in the _get lock. - GstQueue unlocks its mutexes when going to PAUSED. - make sure that when all elements in a bin/thread go to PAUSED, the bin is set to PAUSED too. - make a parent bin wait for its children to PAUSE before ending the iteration with FALSE (EOS) - Some changes to GstPlay to deal with EOS. - aplied the latest patch from Zeenix to gstrtp. end result: GstPlay doesn't crash on EOS and the pipeline is now shut down properly.
2001-12-04 22:12:50 +00:00
goto restart;
}
This is an attempt at not segfaulting on errors but reporting some usefull info instead. Original commit message from CVS: This is an attempt at not segfaulting on errors but reporting some usefull info instead. - bin changes so errors can propagate. - changed the _FAST macros to _CAST because that is what they do. - removed all references to cothreads out of the core, they are really a scheduler issue, handler with a sched_private gpointer. - added a live buffer count, for debugging buffer leaks. - added error checking in gst_scheduler_state_transition this solves the "out of cothreads" problem. - GST_ELEMENT_NO_ENTRY == GST_ELEMENT_INFINITE_LOOP - added 2 private element flasg for use by the scheduler (_COTHREAD_STOPPING) is now - added scheduler entry points: - _yield : to create possible scheduling points. - _interrupt: to stop execution of an element. - _error: to signal en error condition to the scheduler. - improved error messages for pads. - signal gst_element_error where appropriate. - added the a new bin to the parent before entering it so one can reference its children. - queue memleak fixes on dispose. - added possible deadlock detection in queue (turned off be default) - GstBasicScheduler is a real class of its own now, hiding its internal variables. - GST_ELEMENT_IS_COTHREAD_STOPPING is gone. either call explicit _yield operations, or make a sane loop. - Better state change handling in filesrc. Better error reporting/recovery too. - updated core plugins. - detect non decoupled elements on scheduler boundries and error.
2001-12-22 21:18:17 +00:00
if (GST_STATE (queue) != GST_STATE_PLAYING) {
/* this means the other end is shut down */
/* try to signal to resolve the error */
if (!queue->may_deadlock) {
g_mutex_unlock (queue->qlock);
gst_data_unref (GST_DATA (buf));
This is an attempt at not segfaulting on errors but reporting some usefull info instead. Original commit message from CVS: This is an attempt at not segfaulting on errors but reporting some usefull info instead. - bin changes so errors can propagate. - changed the _FAST macros to _CAST because that is what they do. - removed all references to cothreads out of the core, they are really a scheduler issue, handler with a sched_private gpointer. - added a live buffer count, for debugging buffer leaks. - added error checking in gst_scheduler_state_transition this solves the "out of cothreads" problem. - GST_ELEMENT_NO_ENTRY == GST_ELEMENT_INFINITE_LOOP - added 2 private element flasg for use by the scheduler (_COTHREAD_STOPPING) is now - added scheduler entry points: - _yield : to create possible scheduling points. - _interrupt: to stop execution of an element. - _error: to signal en error condition to the scheduler. - improved error messages for pads. - signal gst_element_error where appropriate. - added the a new bin to the parent before entering it so one can reference its children. - queue memleak fixes on dispose. - added possible deadlock detection in queue (turned off be default) - GstBasicScheduler is a real class of its own now, hiding its internal variables. - GST_ELEMENT_IS_COTHREAD_STOPPING is gone. either call explicit _yield operations, or make a sane loop. - Better state change handling in filesrc. Better error reporting/recovery too. - updated core plugins. - detect non decoupled elements on scheduler boundries and error.
2001-12-22 21:18:17 +00:00
gst_element_error (GST_ELEMENT (queue), "deadlock found, source pad elements are shut down");
/* we don't want to goto out_unref here, since we want to clean up before calling gst_element_error */
This is an attempt at not segfaulting on errors but reporting some usefull info instead. Original commit message from CVS: This is an attempt at not segfaulting on errors but reporting some usefull info instead. - bin changes so errors can propagate. - changed the _FAST macros to _CAST because that is what they do. - removed all references to cothreads out of the core, they are really a scheduler issue, handler with a sched_private gpointer. - added a live buffer count, for debugging buffer leaks. - added error checking in gst_scheduler_state_transition this solves the "out of cothreads" problem. - GST_ELEMENT_NO_ENTRY == GST_ELEMENT_INFINITE_LOOP - added 2 private element flasg for use by the scheduler (_COTHREAD_STOPPING) is now - added scheduler entry points: - _yield : to create possible scheduling points. - _interrupt: to stop execution of an element. - _error: to signal en error condition to the scheduler. - improved error messages for pads. - signal gst_element_error where appropriate. - added the a new bin to the parent before entering it so one can reference its children. - queue memleak fixes on dispose. - added possible deadlock detection in queue (turned off be default) - GstBasicScheduler is a real class of its own now, hiding its internal variables. - GST_ELEMENT_IS_COTHREAD_STOPPING is gone. either call explicit _yield operations, or make a sane loop. - Better state change handling in filesrc. Better error reporting/recovery too. - updated core plugins. - detect non decoupled elements on scheduler boundries and error.
2001-12-22 21:18:17 +00:00
return;
}
else {
g_print ("%s: waiting for the app to restart source pad elements\n", GST_ELEMENT_NAME (queue));
This is an attempt at not segfaulting on errors but reporting some usefull info instead. Original commit message from CVS: This is an attempt at not segfaulting on errors but reporting some usefull info instead. - bin changes so errors can propagate. - changed the _FAST macros to _CAST because that is what they do. - removed all references to cothreads out of the core, they are really a scheduler issue, handler with a sched_private gpointer. - added a live buffer count, for debugging buffer leaks. - added error checking in gst_scheduler_state_transition this solves the "out of cothreads" problem. - GST_ELEMENT_NO_ENTRY == GST_ELEMENT_INFINITE_LOOP - added 2 private element flasg for use by the scheduler (_COTHREAD_STOPPING) is now - added scheduler entry points: - _yield : to create possible scheduling points. - _interrupt: to stop execution of an element. - _error: to signal en error condition to the scheduler. - improved error messages for pads. - signal gst_element_error where appropriate. - added the a new bin to the parent before entering it so one can reference its children. - queue memleak fixes on dispose. - added possible deadlock detection in queue (turned off be default) - GstBasicScheduler is a real class of its own now, hiding its internal variables. - GST_ELEMENT_IS_COTHREAD_STOPPING is gone. either call explicit _yield operations, or make a sane loop. - Better state change handling in filesrc. Better error reporting/recovery too. - updated core plugins. - detect non decoupled elements on scheduler boundries and error.
2001-12-22 21:18:17 +00:00
}
}
GST_DEBUG_ELEMENT (GST_CAT_DATAFLOW, queue, "waiting for not_full, level:%d/%d",
queue->level_buffers, queue->size_buffers);
g_cond_wait (queue->not_full, queue->qlock);
GST_DEBUG_ELEMENT (GST_CAT_DATAFLOW, queue, "got not_full signal");
}
GST_DEBUG_ELEMENT (GST_CAT_DATAFLOW, queue, "post full wait, level:%d/%d",
queue->level_buffers, queue->size_buffers);
}
/* put the buffer on the tail of the list */
g_queue_push_tail (queue->queue, buf);
queue->level_buffers++;
queue->level_bytes += GST_BUFFER_SIZE(buf);
/* this assertion _has_ to hold */
g_assert (queue->queue->length == queue->level_buffers);
GST_DEBUG_ELEMENT (GST_CAT_DATAFLOW, queue, "(%s:%s)+ level:%d/%d",
GST_DEBUG_PAD_NAME(pad),
queue->level_buffers, queue->size_buffers);
GST_DEBUG_ELEMENT (GST_CAT_DATAFLOW, queue, "signalling not_empty");
g_cond_signal (queue->not_empty);
g_mutex_unlock (queue->qlock);
return;
out_unref:
gst_data_unref (GST_DATA (buf));
return;
}
static GstBuffer *
gst_queue_get (GstPad *pad)
{
GstQueue *queue;
GstBuffer *buf = NULL;
gpointer front;
g_assert(pad != NULL);
g_assert(GST_IS_PAD(pad));
g_return_val_if_fail (pad != NULL, NULL);
g_return_val_if_fail (GST_IS_PAD (pad), NULL);
queue = GST_QUEUE (GST_OBJECT_PARENT (pad));
- some fixes to int2float making automake 1.5 happy (gst now requires automake1.5). It's still not perfect but it bui... Original commit message from CVS: - added playondemand plugin by Leif Morgan Johnson <lmjohns3@eos.ncsu.edu> - some fixes to int2float - aplied a patch from wrobell <wrobell@ite.pl> that is a first attempt at making automake 1.5 happy (gst now requires automake1.5). It's still not perfect but it builds. - Made the schedulers plugable. The default scheduler now lives inside a plugin. - Added a new mpeg1/2 parser/demuxer. - Fixed some compiler warnings in the core libs. - substantial work to GstThread (hopefully less race conditions). simplified the code in GstThread a bit. A state change can now also happen in the thread context. - reworked the state semantics of a bin. it'll now automatically get the highest state of its children. - the autoplugger now nests the threads so that a state change failure of one thread doesn't make its upstream thread lock. - GstQueue refuses to go to PLAYING if the sinkpad is not connected. This way the queue will not wedge in the _get lock. - GstQueue unlocks its mutexes when going to PAUSED. - make sure that when all elements in a bin/thread go to PAUSED, the bin is set to PAUSED too. - make a parent bin wait for its children to PAUSE before ending the iteration with FALSE (EOS) - Some changes to GstPlay to deal with EOS. - aplied the latest patch from Zeenix to gstrtp. end result: GstPlay doesn't crash on EOS and the pipeline is now shut down properly.
2001-12-04 22:12:50 +00:00
restart:
/* have to lock for thread-safety */
GST_DEBUG_ELEMENT (GST_CAT_DATAFLOW, queue, "locking t:%p", g_thread_self ());
g_mutex_lock (queue->qlock);
GST_DEBUG_ELEMENT (GST_CAT_DATAFLOW, queue, "locked t:%p %p", g_thread_self (), queue->not_empty);
GST_DEBUG_ELEMENT (GST_CAT_DATAFLOW, queue, "pre empty wait, level:%d/%d", queue->level_buffers, queue->size_buffers);
while (queue->level_buffers == 0) {
/* if there's a pending state change for this queue or its manager, switch
* back to iterator so bottom half of state change executes
*/
//while (GST_STATE_PENDING (queue) != GST_STATE_VOID_PENDING) {
if (queue->interrupt) {
GST_DEBUG_ELEMENT (GST_CAT_DATAFLOW, queue, "interrupted!!");
g_mutex_unlock (queue->qlock);
if (gst_scheduler_interrupt (gst_pad_get_scheduler (queue->srcpad), GST_ELEMENT (queue)))
return GST_BUFFER (gst_event_new (GST_EVENT_INTERRUPT));
- some fixes to int2float making automake 1.5 happy (gst now requires automake1.5). It's still not perfect but it bui... Original commit message from CVS: - added playondemand plugin by Leif Morgan Johnson <lmjohns3@eos.ncsu.edu> - some fixes to int2float - aplied a patch from wrobell <wrobell@ite.pl> that is a first attempt at making automake 1.5 happy (gst now requires automake1.5). It's still not perfect but it builds. - Made the schedulers plugable. The default scheduler now lives inside a plugin. - Added a new mpeg1/2 parser/demuxer. - Fixed some compiler warnings in the core libs. - substantial work to GstThread (hopefully less race conditions). simplified the code in GstThread a bit. A state change can now also happen in the thread context. - reworked the state semantics of a bin. it'll now automatically get the highest state of its children. - the autoplugger now nests the threads so that a state change failure of one thread doesn't make its upstream thread lock. - GstQueue refuses to go to PLAYING if the sinkpad is not connected. This way the queue will not wedge in the _get lock. - GstQueue unlocks its mutexes when going to PAUSED. - make sure that when all elements in a bin/thread go to PAUSED, the bin is set to PAUSED too. - make a parent bin wait for its children to PAUSE before ending the iteration with FALSE (EOS) - Some changes to GstPlay to deal with EOS. - aplied the latest patch from Zeenix to gstrtp. end result: GstPlay doesn't crash on EOS and the pipeline is now shut down properly.
2001-12-04 22:12:50 +00:00
goto restart;
}
This is an attempt at not segfaulting on errors but reporting some usefull info instead. Original commit message from CVS: This is an attempt at not segfaulting on errors but reporting some usefull info instead. - bin changes so errors can propagate. - changed the _FAST macros to _CAST because that is what they do. - removed all references to cothreads out of the core, they are really a scheduler issue, handler with a sched_private gpointer. - added a live buffer count, for debugging buffer leaks. - added error checking in gst_scheduler_state_transition this solves the "out of cothreads" problem. - GST_ELEMENT_NO_ENTRY == GST_ELEMENT_INFINITE_LOOP - added 2 private element flasg for use by the scheduler (_COTHREAD_STOPPING) is now - added scheduler entry points: - _yield : to create possible scheduling points. - _interrupt: to stop execution of an element. - _error: to signal en error condition to the scheduler. - improved error messages for pads. - signal gst_element_error where appropriate. - added the a new bin to the parent before entering it so one can reference its children. - queue memleak fixes on dispose. - added possible deadlock detection in queue (turned off be default) - GstBasicScheduler is a real class of its own now, hiding its internal variables. - GST_ELEMENT_IS_COTHREAD_STOPPING is gone. either call explicit _yield operations, or make a sane loop. - Better state change handling in filesrc. Better error reporting/recovery too. - updated core plugins. - detect non decoupled elements on scheduler boundries and error.
2001-12-22 21:18:17 +00:00
if (GST_STATE (queue) != GST_STATE_PLAYING) {
/* this means the other end is shut down */
if (!queue->may_deadlock) {
g_mutex_unlock (queue->qlock);
gst_element_error (GST_ELEMENT (queue), "deadlock found, sink pad elements are shut down");
goto restart;
This is an attempt at not segfaulting on errors but reporting some usefull info instead. Original commit message from CVS: This is an attempt at not segfaulting on errors but reporting some usefull info instead. - bin changes so errors can propagate. - changed the _FAST macros to _CAST because that is what they do. - removed all references to cothreads out of the core, they are really a scheduler issue, handler with a sched_private gpointer. - added a live buffer count, for debugging buffer leaks. - added error checking in gst_scheduler_state_transition this solves the "out of cothreads" problem. - GST_ELEMENT_NO_ENTRY == GST_ELEMENT_INFINITE_LOOP - added 2 private element flasg for use by the scheduler (_COTHREAD_STOPPING) is now - added scheduler entry points: - _yield : to create possible scheduling points. - _interrupt: to stop execution of an element. - _error: to signal en error condition to the scheduler. - improved error messages for pads. - signal gst_element_error where appropriate. - added the a new bin to the parent before entering it so one can reference its children. - queue memleak fixes on dispose. - added possible deadlock detection in queue (turned off be default) - GstBasicScheduler is a real class of its own now, hiding its internal variables. - GST_ELEMENT_IS_COTHREAD_STOPPING is gone. either call explicit _yield operations, or make a sane loop. - Better state change handling in filesrc. Better error reporting/recovery too. - updated core plugins. - detect non decoupled elements on scheduler boundries and error.
2001-12-22 21:18:17 +00:00
}
else {
g_print ("%s: waiting for the app to restart source pad elements\n", GST_ELEMENT_NAME (queue));
This is an attempt at not segfaulting on errors but reporting some usefull info instead. Original commit message from CVS: This is an attempt at not segfaulting on errors but reporting some usefull info instead. - bin changes so errors can propagate. - changed the _FAST macros to _CAST because that is what they do. - removed all references to cothreads out of the core, they are really a scheduler issue, handler with a sched_private gpointer. - added a live buffer count, for debugging buffer leaks. - added error checking in gst_scheduler_state_transition this solves the "out of cothreads" problem. - GST_ELEMENT_NO_ENTRY == GST_ELEMENT_INFINITE_LOOP - added 2 private element flasg for use by the scheduler (_COTHREAD_STOPPING) is now - added scheduler entry points: - _yield : to create possible scheduling points. - _interrupt: to stop execution of an element. - _error: to signal en error condition to the scheduler. - improved error messages for pads. - signal gst_element_error where appropriate. - added the a new bin to the parent before entering it so one can reference its children. - queue memleak fixes on dispose. - added possible deadlock detection in queue (turned off be default) - GstBasicScheduler is a real class of its own now, hiding its internal variables. - GST_ELEMENT_IS_COTHREAD_STOPPING is gone. either call explicit _yield operations, or make a sane loop. - Better state change handling in filesrc. Better error reporting/recovery too. - updated core plugins. - detect non decoupled elements on scheduler boundries and error.
2001-12-22 21:18:17 +00:00
}
}
GST_DEBUG_ELEMENT (GST_CAT_DATAFLOW, queue, "waiting for not_empty, level:%d/%d", queue->level_buffers, queue->size_buffers);
/* if (queue->block_timeout > -1){ */
if (FALSE) {
GTimeVal timeout;
g_get_current_time(&timeout);
g_time_val_add(&timeout, queue->block_timeout);
if (!g_cond_timed_wait (queue->not_empty, queue->qlock, &timeout)){
g_mutex_unlock (queue->qlock);
g_warning ("filler");
return GST_BUFFER(gst_event_new_filler());
}
}
else {
g_cond_wait (queue->not_empty, queue->qlock);
}
GST_DEBUG_ELEMENT (GST_CAT_DATAFLOW, queue, "got not_empty signal");
}
GST_DEBUG_ELEMENT (GST_CAT_DATAFLOW, queue, "post empty wait, level:%d/%d", queue->level_buffers, queue->size_buffers);
front = g_queue_pop_head (queue->queue);
buf = (GstBuffer *)(front);
GST_DEBUG_ELEMENT (GST_CAT_DATAFLOW, queue, "retrieved buffer %p from queue", buf);
queue->level_buffers--;
queue->level_bytes -= GST_BUFFER_SIZE(buf);
GST_DEBUG_ELEMENT (GST_CAT_DATAFLOW, queue, "(%s:%s)- level:%d/%d",
GST_DEBUG_PAD_NAME(pad),
queue->level_buffers, queue->size_buffers);
/* this assertion _has_ to hold */
g_assert (queue->queue->length == queue->level_buffers);
GST_DEBUG_ELEMENT (GST_CAT_DATAFLOW, queue, "signalling not_full");
g_cond_signal (queue->not_full);
g_mutex_unlock (queue->qlock);
/* FIXME where should this be? locked? */
if (GST_IS_EVENT(buf)) {
GstEvent *event = GST_EVENT(buf);
switch (GST_EVENT_TYPE(event)) {
case GST_EVENT_EOS:
GST_DEBUG_ELEMENT (GST_CAT_DATAFLOW, queue, "queue \"%s\" eos", GST_ELEMENT_NAME (queue));
gst_element_set_eos (GST_ELEMENT (queue));
break;
default:
break;
}
}
return buf;
}
static gboolean
gst_queue_handle_src_event (GstPad *pad, GstEvent *event)
{
GstQueue *queue;
gboolean res;
gint event_type;
gint flag_flush = 0;
queue = GST_QUEUE (GST_OBJECT_PARENT (pad));
g_mutex_lock (queue->qlock);
if (gst_element_get_state (GST_ELEMENT (queue)) == GST_STATE_PLAYING) {
/* push the event to the queue for upstream consumption */
g_async_queue_push(queue->events, event);
g_warning ("FIXME: sending event in a running queue");
/* FIXME wait for delivery of the event here, then return the result
* instead of FALSE */
res = FALSE;
goto done;
}
event_type = GST_EVENT_TYPE (event);
if (event_type == GST_EVENT_SEEK)
flag_flush = GST_EVENT_SEEK_FLAGS (event) & GST_SEEK_FLAG_FLUSH;
res = gst_pad_event_default (pad, event);
switch (event_type) {
case GST_EVENT_FLUSH:
GST_DEBUG_ELEMENT (GST_CAT_DATAFLOW, queue, "FLUSH event, flushing queue\n");
gst_queue_locked_flush (queue);
break;
case GST_EVENT_SEEK:
if (flag_flush) {
gst_queue_locked_flush (queue);
}
default:
break;
}
done:
g_mutex_unlock (queue->qlock);
/* we have to claim success, but we don't really know */
return res;
}
static gboolean
gst_queue_release_locks (GstElement *element)
{
GstQueue *queue;
queue = GST_QUEUE (element);
g_mutex_lock (queue->qlock);
queue->interrupt = TRUE;
g_cond_signal (queue->not_full);
g_cond_signal (queue->not_empty);
g_mutex_unlock (queue->qlock);
return TRUE;
}
static GstElementStateReturn
gst_queue_change_state (GstElement *element)
{
GstQueue *queue;
GstElementStateReturn ret;
queue = GST_QUEUE (element);
- some fixes to int2float making automake 1.5 happy (gst now requires automake1.5). It's still not perfect but it bui... Original commit message from CVS: - added playondemand plugin by Leif Morgan Johnson <lmjohns3@eos.ncsu.edu> - some fixes to int2float - aplied a patch from wrobell <wrobell@ite.pl> that is a first attempt at making automake 1.5 happy (gst now requires automake1.5). It's still not perfect but it builds. - Made the schedulers plugable. The default scheduler now lives inside a plugin. - Added a new mpeg1/2 parser/demuxer. - Fixed some compiler warnings in the core libs. - substantial work to GstThread (hopefully less race conditions). simplified the code in GstThread a bit. A state change can now also happen in the thread context. - reworked the state semantics of a bin. it'll now automatically get the highest state of its children. - the autoplugger now nests the threads so that a state change failure of one thread doesn't make its upstream thread lock. - GstQueue refuses to go to PLAYING if the sinkpad is not connected. This way the queue will not wedge in the _get lock. - GstQueue unlocks its mutexes when going to PAUSED. - make sure that when all elements in a bin/thread go to PAUSED, the bin is set to PAUSED too. - make a parent bin wait for its children to PAUSE before ending the iteration with FALSE (EOS) - Some changes to GstPlay to deal with EOS. - aplied the latest patch from Zeenix to gstrtp. end result: GstPlay doesn't crash on EOS and the pipeline is now shut down properly.
2001-12-04 22:12:50 +00:00
GST_DEBUG_ENTER("('%s')", GST_ELEMENT_NAME (element));
/* lock the queue so another thread (not in sync with this thread's state)
* can't call this queue's _get (or whatever)
*/
- some fixes to int2float making automake 1.5 happy (gst now requires automake1.5). It's still not perfect but it bui... Original commit message from CVS: - added playondemand plugin by Leif Morgan Johnson <lmjohns3@eos.ncsu.edu> - some fixes to int2float - aplied a patch from wrobell <wrobell@ite.pl> that is a first attempt at making automake 1.5 happy (gst now requires automake1.5). It's still not perfect but it builds. - Made the schedulers plugable. The default scheduler now lives inside a plugin. - Added a new mpeg1/2 parser/demuxer. - Fixed some compiler warnings in the core libs. - substantial work to GstThread (hopefully less race conditions). simplified the code in GstThread a bit. A state change can now also happen in the thread context. - reworked the state semantics of a bin. it'll now automatically get the highest state of its children. - the autoplugger now nests the threads so that a state change failure of one thread doesn't make its upstream thread lock. - GstQueue refuses to go to PLAYING if the sinkpad is not connected. This way the queue will not wedge in the _get lock. - GstQueue unlocks its mutexes when going to PAUSED. - make sure that when all elements in a bin/thread go to PAUSED, the bin is set to PAUSED too. - make a parent bin wait for its children to PAUSE before ending the iteration with FALSE (EOS) - Some changes to GstPlay to deal with EOS. - aplied the latest patch from Zeenix to gstrtp. end result: GstPlay doesn't crash on EOS and the pipeline is now shut down properly.
2001-12-04 22:12:50 +00:00
g_mutex_lock (queue->qlock);
switch (GST_STATE_TRANSITION (element)) {
case GST_STATE_NULL_TO_READY:
gst_queue_locked_flush (queue);
break;
case GST_STATE_READY_TO_PAUSED:
break;
case GST_STATE_PAUSED_TO_PLAYING:
if (!GST_PAD_IS_LINKED (queue->sinkpad)) {
GST_DEBUG_ELEMENT (GST_CAT_STATES, queue, "queue %s is not linked", GST_ELEMENT_NAME (queue));
/* FIXME can this be? */
g_cond_signal (queue->not_empty);
ret = GST_STATE_FAILURE;
goto error;
}
else {
GstScheduler *src_sched, *sink_sched;
src_sched = gst_pad_get_scheduler (GST_PAD_CAST (queue->srcpad));
sink_sched = gst_pad_get_scheduler (GST_PAD_CAST (queue->sinkpad));
if (src_sched == sink_sched) {
GST_DEBUG_ELEMENT (GST_CAT_STATES, queue, "queue %s does not connect different schedulers",
GST_ELEMENT_NAME (queue));
- some fixes to int2float making automake 1.5 happy (gst now requires automake1.5). It's still not perfect but it bui... Original commit message from CVS: - added playondemand plugin by Leif Morgan Johnson <lmjohns3@eos.ncsu.edu> - some fixes to int2float - aplied a patch from wrobell <wrobell@ite.pl> that is a first attempt at making automake 1.5 happy (gst now requires automake1.5). It's still not perfect but it builds. - Made the schedulers plugable. The default scheduler now lives inside a plugin. - Added a new mpeg1/2 parser/demuxer. - Fixed some compiler warnings in the core libs. - substantial work to GstThread (hopefully less race conditions). simplified the code in GstThread a bit. A state change can now also happen in the thread context. - reworked the state semantics of a bin. it'll now automatically get the highest state of its children. - the autoplugger now nests the threads so that a state change failure of one thread doesn't make its upstream thread lock. - GstQueue refuses to go to PLAYING if the sinkpad is not connected. This way the queue will not wedge in the _get lock. - GstQueue unlocks its mutexes when going to PAUSED. - make sure that when all elements in a bin/thread go to PAUSED, the bin is set to PAUSED too. - make a parent bin wait for its children to PAUSE before ending the iteration with FALSE (EOS) - Some changes to GstPlay to deal with EOS. - aplied the latest patch from Zeenix to gstrtp. end result: GstPlay doesn't crash on EOS and the pipeline is now shut down properly.
2001-12-04 22:12:50 +00:00
g_warning ("queue %s does not connect different schedulers",
GST_ELEMENT_NAME (queue));
ret = GST_STATE_FAILURE;
goto error;
}
}
queue->interrupt = FALSE;
break;
case GST_STATE_PLAYING_TO_PAUSED:
break;
case GST_STATE_PAUSED_TO_READY:
gst_queue_locked_flush (queue);
break;
case GST_STATE_READY_TO_NULL:
break;
- some fixes to int2float making automake 1.5 happy (gst now requires automake1.5). It's still not perfect but it bui... Original commit message from CVS: - added playondemand plugin by Leif Morgan Johnson <lmjohns3@eos.ncsu.edu> - some fixes to int2float - aplied a patch from wrobell <wrobell@ite.pl> that is a first attempt at making automake 1.5 happy (gst now requires automake1.5). It's still not perfect but it builds. - Made the schedulers plugable. The default scheduler now lives inside a plugin. - Added a new mpeg1/2 parser/demuxer. - Fixed some compiler warnings in the core libs. - substantial work to GstThread (hopefully less race conditions). simplified the code in GstThread a bit. A state change can now also happen in the thread context. - reworked the state semantics of a bin. it'll now automatically get the highest state of its children. - the autoplugger now nests the threads so that a state change failure of one thread doesn't make its upstream thread lock. - GstQueue refuses to go to PLAYING if the sinkpad is not connected. This way the queue will not wedge in the _get lock. - GstQueue unlocks its mutexes when going to PAUSED. - make sure that when all elements in a bin/thread go to PAUSED, the bin is set to PAUSED too. - make a parent bin wait for its children to PAUSE before ending the iteration with FALSE (EOS) - Some changes to GstPlay to deal with EOS. - aplied the latest patch from Zeenix to gstrtp. end result: GstPlay doesn't crash on EOS and the pipeline is now shut down properly.
2001-12-04 22:12:50 +00:00
}
ret = GST_ELEMENT_CLASS (parent_class)->change_state (element);
error:
- some fixes to int2float making automake 1.5 happy (gst now requires automake1.5). It's still not perfect but it bui... Original commit message from CVS: - added playondemand plugin by Leif Morgan Johnson <lmjohns3@eos.ncsu.edu> - some fixes to int2float - aplied a patch from wrobell <wrobell@ite.pl> that is a first attempt at making automake 1.5 happy (gst now requires automake1.5). It's still not perfect but it builds. - Made the schedulers plugable. The default scheduler now lives inside a plugin. - Added a new mpeg1/2 parser/demuxer. - Fixed some compiler warnings in the core libs. - substantial work to GstThread (hopefully less race conditions). simplified the code in GstThread a bit. A state change can now also happen in the thread context. - reworked the state semantics of a bin. it'll now automatically get the highest state of its children. - the autoplugger now nests the threads so that a state change failure of one thread doesn't make its upstream thread lock. - GstQueue refuses to go to PLAYING if the sinkpad is not connected. This way the queue will not wedge in the _get lock. - GstQueue unlocks its mutexes when going to PAUSED. - make sure that when all elements in a bin/thread go to PAUSED, the bin is set to PAUSED too. - make a parent bin wait for its children to PAUSE before ending the iteration with FALSE (EOS) - Some changes to GstPlay to deal with EOS. - aplied the latest patch from Zeenix to gstrtp. end result: GstPlay doesn't crash on EOS and the pipeline is now shut down properly.
2001-12-04 22:12:50 +00:00
g_mutex_unlock (queue->qlock);
- some fixes to int2float making automake 1.5 happy (gst now requires automake1.5). It's still not perfect but it bui... Original commit message from CVS: - added playondemand plugin by Leif Morgan Johnson <lmjohns3@eos.ncsu.edu> - some fixes to int2float - aplied a patch from wrobell <wrobell@ite.pl> that is a first attempt at making automake 1.5 happy (gst now requires automake1.5). It's still not perfect but it builds. - Made the schedulers plugable. The default scheduler now lives inside a plugin. - Added a new mpeg1/2 parser/demuxer. - Fixed some compiler warnings in the core libs. - substantial work to GstThread (hopefully less race conditions). simplified the code in GstThread a bit. A state change can now also happen in the thread context. - reworked the state semantics of a bin. it'll now automatically get the highest state of its children. - the autoplugger now nests the threads so that a state change failure of one thread doesn't make its upstream thread lock. - GstQueue refuses to go to PLAYING if the sinkpad is not connected. This way the queue will not wedge in the _get lock. - GstQueue unlocks its mutexes when going to PAUSED. - make sure that when all elements in a bin/thread go to PAUSED, the bin is set to PAUSED too. - make a parent bin wait for its children to PAUSE before ending the iteration with FALSE (EOS) - Some changes to GstPlay to deal with EOS. - aplied the latest patch from Zeenix to gstrtp. end result: GstPlay doesn't crash on EOS and the pipeline is now shut down properly.
2001-12-04 22:12:50 +00:00
GST_DEBUG_LEAVE("('%s')", GST_ELEMENT_NAME (element));
return ret;
}
static void
gst_queue_set_property (GObject *object, guint prop_id, const GValue *value, GParamSpec *pspec)
{
GstQueue *queue;
/* it's not null if we got it, but it might not be ours */
g_return_if_fail (GST_IS_QUEUE (object));
queue = GST_QUEUE (object);
switch (prop_id) {
case ARG_LEAKY:
queue->leaky = g_value_get_enum (value);
break;
case ARG_MAX_LEVEL:
This is an attempt at not segfaulting on errors but reporting some usefull info instead. Original commit message from CVS: This is an attempt at not segfaulting on errors but reporting some usefull info instead. - bin changes so errors can propagate. - changed the _FAST macros to _CAST because that is what they do. - removed all references to cothreads out of the core, they are really a scheduler issue, handler with a sched_private gpointer. - added a live buffer count, for debugging buffer leaks. - added error checking in gst_scheduler_state_transition this solves the "out of cothreads" problem. - GST_ELEMENT_NO_ENTRY == GST_ELEMENT_INFINITE_LOOP - added 2 private element flasg for use by the scheduler (_COTHREAD_STOPPING) is now - added scheduler entry points: - _yield : to create possible scheduling points. - _interrupt: to stop execution of an element. - _error: to signal en error condition to the scheduler. - improved error messages for pads. - signal gst_element_error where appropriate. - added the a new bin to the parent before entering it so one can reference its children. - queue memleak fixes on dispose. - added possible deadlock detection in queue (turned off be default) - GstBasicScheduler is a real class of its own now, hiding its internal variables. - GST_ELEMENT_IS_COTHREAD_STOPPING is gone. either call explicit _yield operations, or make a sane loop. - Better state change handling in filesrc. Better error reporting/recovery too. - updated core plugins. - detect non decoupled elements on scheduler boundries and error.
2001-12-22 21:18:17 +00:00
queue->size_buffers = g_value_get_int (value);
break;
case ARG_MAY_DEADLOCK:
queue->may_deadlock = g_value_get_boolean (value);
break;
case ARG_BLOCK_TIMEOUT:
queue->block_timeout = g_value_get_int (value);
break;
default:
G_OBJECT_WARN_INVALID_PROPERTY_ID (object, prop_id, pspec);
break;
}
}
static void
gst_queue_get_property (GObject *object, guint prop_id, GValue *value, GParamSpec *pspec)
{
GstQueue *queue;
/* it's not null if we got it, but it might not be ours */
g_return_if_fail (GST_IS_QUEUE (object));
queue = GST_QUEUE (object);
switch (prop_id) {
case ARG_LEAKY:
g_value_set_enum (value, queue->leaky);
break;
case ARG_LEVEL:
This is an attempt at not segfaulting on errors but reporting some usefull info instead. Original commit message from CVS: This is an attempt at not segfaulting on errors but reporting some usefull info instead. - bin changes so errors can propagate. - changed the _FAST macros to _CAST because that is what they do. - removed all references to cothreads out of the core, they are really a scheduler issue, handler with a sched_private gpointer. - added a live buffer count, for debugging buffer leaks. - added error checking in gst_scheduler_state_transition this solves the "out of cothreads" problem. - GST_ELEMENT_NO_ENTRY == GST_ELEMENT_INFINITE_LOOP - added 2 private element flasg for use by the scheduler (_COTHREAD_STOPPING) is now - added scheduler entry points: - _yield : to create possible scheduling points. - _interrupt: to stop execution of an element. - _error: to signal en error condition to the scheduler. - improved error messages for pads. - signal gst_element_error where appropriate. - added the a new bin to the parent before entering it so one can reference its children. - queue memleak fixes on dispose. - added possible deadlock detection in queue (turned off be default) - GstBasicScheduler is a real class of its own now, hiding its internal variables. - GST_ELEMENT_IS_COTHREAD_STOPPING is gone. either call explicit _yield operations, or make a sane loop. - Better state change handling in filesrc. Better error reporting/recovery too. - updated core plugins. - detect non decoupled elements on scheduler boundries and error.
2001-12-22 21:18:17 +00:00
g_value_set_int (value, queue->level_buffers);
break;
case ARG_MAX_LEVEL:
This is an attempt at not segfaulting on errors but reporting some usefull info instead. Original commit message from CVS: This is an attempt at not segfaulting on errors but reporting some usefull info instead. - bin changes so errors can propagate. - changed the _FAST macros to _CAST because that is what they do. - removed all references to cothreads out of the core, they are really a scheduler issue, handler with a sched_private gpointer. - added a live buffer count, for debugging buffer leaks. - added error checking in gst_scheduler_state_transition this solves the "out of cothreads" problem. - GST_ELEMENT_NO_ENTRY == GST_ELEMENT_INFINITE_LOOP - added 2 private element flasg for use by the scheduler (_COTHREAD_STOPPING) is now - added scheduler entry points: - _yield : to create possible scheduling points. - _interrupt: to stop execution of an element. - _error: to signal en error condition to the scheduler. - improved error messages for pads. - signal gst_element_error where appropriate. - added the a new bin to the parent before entering it so one can reference its children. - queue memleak fixes on dispose. - added possible deadlock detection in queue (turned off be default) - GstBasicScheduler is a real class of its own now, hiding its internal variables. - GST_ELEMENT_IS_COTHREAD_STOPPING is gone. either call explicit _yield operations, or make a sane loop. - Better state change handling in filesrc. Better error reporting/recovery too. - updated core plugins. - detect non decoupled elements on scheduler boundries and error.
2001-12-22 21:18:17 +00:00
g_value_set_int (value, queue->size_buffers);
break;
case ARG_MAY_DEADLOCK:
g_value_set_boolean (value, queue->may_deadlock);
break;
case ARG_BLOCK_TIMEOUT:
g_value_set_int (value, queue->block_timeout);
break;
default:
G_OBJECT_WARN_INVALID_PROPERTY_ID (object, prop_id, pspec);
break;
}
}