platform/upstream/gstreamer.git
19 years agodocs/design/part-activation.txt: Notes on how activation should work -- not quite...
Andy Wingo [Mon, 27 Jun 2005 13:25:44 +0000 (13:25 +0000)]
docs/design/part-activation.txt: Notes on how activation should work -- not quite implemented yet.

Original commit message from CVS:
2005-06-27  Andy Wingo  <wingo@pobox.com>

* docs/design/part-activation.txt: Notes on how activation should
work -- not quite implemented yet.

19 years agoremove stuff that isn't there anymore
Thomas Vander Stichele [Mon, 27 Jun 2005 08:54:16 +0000 (08:54 +0000)]
remove stuff that isn't there anymore

Original commit message from CVS:
remove stuff that isn't there anymore

19 years agomore doc and whitespace fixes
Thomas Vander Stichele [Mon, 27 Jun 2005 08:16:51 +0000 (08:16 +0000)]
more doc and whitespace fixes

Original commit message from CVS:
more doc and whitespace fixes

19 years agogst/gstghostpad.c: At least get the chain function correct, needs more fixing.
Wim Taymans [Sat, 25 Jun 2005 19:53:02 +0000 (19:53 +0000)]
gst/gstghostpad.c: At least get the chain function correct, needs more fixing.

Original commit message from CVS:
* gst/gstghostpad.c: (gst_proxy_pad_do_chain):
At least get the chain function correct, needs more
fixing.

19 years agogst/: Right, two problems here: ghostpads don't take locks and glib _rec_mutex_lock_f...
Wim Taymans [Sat, 25 Jun 2005 19:37:59 +0000 (19:37 +0000)]
gst/: Right, two problems here: ghostpads don't take locks and glib _rec_mutex_lock_full() with depth==0 still locks.

Original commit message from CVS:
* gst/base/gstbasesink.c: (gst_basesink_preroll_queue_empty),
(gst_basesink_handle_object), (gst_basesink_event),
(gst_basesink_do_sync), (gst_basesink_handle_event),
(gst_basesink_change_state):
* gst/gsttask.h:
Right, two problems here: ghostpads don't take locks and
glib _rec_mutex_lock_full() with depth==0 still locks.
Catch illegal locking and g_warn them.

19 years agocheck/states/sinks.c: Have to check for completion now...
Wim Taymans [Sat, 25 Jun 2005 19:14:51 +0000 (19:14 +0000)]
check/states/sinks.c: Have to check for completion now...

Original commit message from CVS:
* check/states/sinks.c: (START_TEST), (gst_object_suite):
Have to check for completion now...

19 years agogst/: Unlock STREAM_LOCK whatever the recursion was.
Wim Taymans [Sat, 25 Jun 2005 19:09:28 +0000 (19:09 +0000)]
gst/: Unlock STREAM_LOCK whatever the recursion was.

Original commit message from CVS:
* gst/base/gstbasesink.c: (gst_basesink_preroll_queue_empty),
(gst_basesink_handle_object), (gst_basesink_event),
(gst_basesink_do_sync), (gst_basesink_handle_event),
(gst_basesink_change_state):
* gst/gstpad.h:
Unlock STREAM_LOCK whatever the recursion was.

19 years agogst/base/gstbasesink.c: Reworked the base sink, handle event and buffer serialisation...
Wim Taymans [Sat, 25 Jun 2005 17:54:58 +0000 (17:54 +0000)]
gst/base/gstbasesink.c: Reworked the base sink, handle event and buffer serialisation correctly and removed possible ...

Original commit message from CVS:
* gst/base/gstbasesink.c: (gst_basesink_set_property),
(gst_basesink_preroll_queue_empty),
(gst_basesink_preroll_queue_flush), (gst_basesink_handle_object),
(gst_basesink_event), (gst_basesink_do_sync),
(gst_basesink_handle_event), (gst_basesink_handle_buffer),
(gst_basesink_chain), (gst_basesink_loop), (gst_basesink_activate),
(gst_basesink_change_state):
Reworked the base sink, handle event and buffer serialisation
correctly and removed possible deadlock.
Handle EOS correctly.

19 years agoAllow elements to post EOS in the state change function.
Wim Taymans [Sat, 25 Jun 2005 17:51:12 +0000 (17:51 +0000)]
Allow elements to post EOS in the state change function.

Original commit message from CVS:
* gst/gstpipeline.c: (is_eos), (pipeline_bus_handler),
(gst_pipeline_change_state):
* tools/gst-launch.c: (check_intr), (event_loop), (main):
Allow elements to post EOS in the state change function.
Fix up -launch, make it exit the poll loop when the
pipeline actually changed state.
Fix up warning parsing in -launch.

19 years agogst/elements/gsttee.c: Core takes STREAM_LOCK for us now.
Wim Taymans [Sat, 25 Jun 2005 17:44:39 +0000 (17:44 +0000)]
gst/elements/gsttee.c: Core takes STREAM_LOCK for us now.

Original commit message from CVS:
* gst/elements/gsttee.c: (gst_tee_chain), (gst_tee_loop),
(gst_tee_sink_activate):
Core takes STREAM_LOCK for us now.

19 years agogst/: Keep track of current target state while performing a state change so that...
Wim Taymans [Sat, 25 Jun 2005 17:42:17 +0000 (17:42 +0000)]
gst/: Keep track of current target state while performing a state change so that subclasses can do something interest...

Original commit message from CVS:
* gst/gstelement.c: (gst_element_get_state_func),
(gst_element_set_state):
* gst/gstelement.h:
* gst/gstmessage.c: (gst_message_parse_error),
(gst_message_parse_warning):
Keep track of current target state while performing a state
change so that subclasses can do something interesting.
Fix parsing of warning/error messages when GError is NULL.

19 years agostart pulling in base classes and elements for docs
Thomas Vander Stichele [Fri, 24 Jun 2005 18:16:14 +0000 (18:16 +0000)]
start pulling in base classes and elements for docs

Original commit message from CVS:
start pulling in base classes and elements for docs

19 years agofixed make distcheck with gtk-doc 1.3
Stefan Kost [Fri, 24 Jun 2005 07:49:40 +0000 (07:49 +0000)]
fixed make distcheck with gtk-doc 1.3

Original commit message from CVS:
fixed make distcheck with gtk-doc 1.3

19 years agogst/gstelement.c: When the state did not change, also report NO_PREROLL when it matters.
Wim Taymans [Thu, 23 Jun 2005 17:11:49 +0000 (17:11 +0000)]
gst/gstelement.c: When the state did not change, also report NO_PREROLL when it matters.

Original commit message from CVS:
* gst/gstelement.c: (gst_element_get_state_func),
(gst_element_set_state), (gst_element_change_state):
When the state did not change, also report NO_PREROLL
when it matters.

19 years agogst/: No unsafe task pausing please.
Wim Taymans [Thu, 23 Jun 2005 17:09:21 +0000 (17:09 +0000)]
gst/: No unsafe task pausing please.

Original commit message from CVS:
* gst/gstpad.c: (gst_pad_event_default):
* gst/gstqueue.c: (gst_queue_loop):
No unsafe task pausing please.

19 years agogst/schedulers/threadscheduler.c: Ref the task before pushing it on the threadpool...
Wim Taymans [Thu, 23 Jun 2005 17:07:08 +0000 (17:07 +0000)]
gst/schedulers/threadscheduler.c: Ref the task before pushing it on the threadpool. This makes sure that we have a re...

Original commit message from CVS:
* gst/schedulers/threadscheduler.c:
(gst_thread_scheduler_task_start),
(gst_thread_scheduler_task_pause), (gst_thread_scheduler_func):
Ref the task before pushing it on the threadpool. This
makes sure that we have a ref when the threadfunction is
actually called.

19 years agogst/base/gstbasesrc.c (gst_basesrc_get_range): Check if the offset is greater than...
Andy Wingo [Thu, 23 Jun 2005 15:26:09 +0000 (15:26 +0000)]
gst/base/gstbasesrc.c (gst_basesrc_get_range): Check if the offset is greater than the file's size.

Original commit message from CVS:
2005-06-23  Andy Wingo  <wingo@pobox.com>

* gst/base/gstbasesrc.c (gst_basesrc_get_range): Check if the
offset is greater than the file's size.

19 years agogst/gstobject.h (GST_CLASS_LOCK, GST_CLASS_TRYLOCK) (GST_CLASS_UNLOCK, GST_CLASS_GET_...
Andy Wingo [Thu, 23 Jun 2005 15:04:48 +0000 (15:04 +0000)]
gst/gstobject.h (GST_CLASS_LOCK, GST_CLASS_TRYLOCK) (GST_CLASS_UNLOCK, GST_CLASS_GET_LOCK, GstObjectClass)

Original commit message from CVS:
2005-06-23  Andy Wingo  <wingo@pobox.com>

* gst/gstobject.h (GST_CLASS_LOCK, GST_CLASS_TRYLOCK)
(GST_CLASS_UNLOCK, GST_CLASS_GET_LOCK, GstObjectClass)
* gst/gstobject.c (gst_object_class_init): Make the class lock
recursive. Wim won't let me drop deep_notify. Decodebin works
again, whoopdy doo.

19 years agogst/gstghostpad.c (on_int_notify): Catches notify::caps on the internal pad, and...
Andy Wingo [Thu, 23 Jun 2005 14:18:15 +0000 (14:18 +0000)]
gst/gstghostpad.c (on_int_notify): Catches notify::caps on the internal pad, and hacks accordingly. Doesn't do it on ...

Original commit message from CVS:
2005-06-23  Andy Wingo  <wingo@pobox.com>

* gst/gstghostpad.c (on_int_notify): Catches notify::caps on the
internal pad, and hacks accordingly. Doesn't do it on the target
pad because we change its caps. Probably catches all cases of
interest tho.
(gst_ghost_pad_set_property): Connect to notify::caps as
appropritate.

19 years agotests/network-clock.scm (plot-simulation): Pipe data to the elite python skript.
Andy Wingo [Thu, 23 Jun 2005 13:20:44 +0000 (13:20 +0000)]
tests/network-clock.scm (plot-simulation): Pipe data to the elite python skript.

Original commit message from CVS:
2005-06-23  Andy Wingo  <wingo@pobox.com>

* tests/network-clock.scm (plot-simulation): Pipe data to the
elite python skript.

* tests/network-clock-utils.scm (define-parameter): New macro,
defines a parameter that can be set via the command line.
(set-parameter!, parse-parameter-arguments): Command line args
parser.

* tests/plot-data: Simple matplotlib-based plotter, takes input on
stdin.

19 years agoadd more important documentation
Thomas Vander Stichele [Thu, 23 Jun 2005 13:20:00 +0000 (13:20 +0000)]
add more important documentation

Original commit message from CVS:
add more important documentation

19 years agogst/elements/gsttypefindelement.c: Don't restart typefinding on a discont.
Jan Schmidt [Thu, 23 Jun 2005 11:43:39 +0000 (11:43 +0000)]
gst/elements/gsttypefindelement.c: Don't restart typefinding on a discont.

Original commit message from CVS:
2005-06-23  Jan Schmidt  <thaytan@mad.scientist.com>

* gst/elements/gsttypefindelement.c:
(gst_type_find_element_handle_event):
Don't restart typefinding on a discont.
* gst/gstelement.c: (gst_element_set_state):
Debug spelling fix.
* gst/gstpad.c: (gst_pad_set_active), (gst_pad_send_event):
Allow changing mode of an active pad.
Debug output fixes.
* gst/registries/gstlibxmlregistry.c: (load_feature):
Don't cast a static pad template to a normal pad template.

19 years agoremove gst_strtoll completely, since it didn't actually do anything more than what...
Thomas Vander Stichele [Thu, 23 Jun 2005 11:25:29 +0000 (11:25 +0000)]
remove gst_strtoll completely, since it didn't actually do anything more than what g_ascii_strtoull already does.

Original commit message from CVS:
* check/gst/gstvalue.c: (START_TEST), (gst_value_suite):
* gst/gstvalue.c: (gst_value_deserialize_int_helper):
remove gst_strtoll completely, since it didn't actually do
anything more than what g_ascii_strtoull already does.
check for range errors when deserializing
do a cast for the unsigned cases; but further fixing needs
a decision on what the interpretation of "(int)" and
deserialization should be for values that fall outside the
type's boundaries (ie, refuse, or interpret as casting)

19 years agoAdded support for live sources and other elements that cannot do preroll.
Wim Taymans [Thu, 23 Jun 2005 10:37:09 +0000 (10:37 +0000)]
Added support for live sources and other elements that cannot do preroll.

Original commit message from CVS:
Added support for live sources and other elements that
cannot do preroll.
Updated design docs, added live-source design doc.
Implemented live source functionality in basesrc
Fix error condition in _bin_get_state()
Implement live source handling in -launch.
Added check for live sources.
Fixed case in GstBin where elements were changed state
multiple times.

19 years agocheck/gst/gstpad.c (test_get_allowed_caps, test_refcount): Fix borken refcounting.
Andy Wingo [Thu, 23 Jun 2005 09:59:33 +0000 (09:59 +0000)]
check/gst/gstpad.c (test_get_allowed_caps, test_refcount): Fix borken refcounting.

Original commit message from CVS:
2005-06-23  Andy Wingo  <wingo@pobox.com>

* check/gst/gstpad.c (test_get_allowed_caps, test_refcount): Fix
borken refcounting.

19 years agocommit the file
Andy Wingo [Thu, 23 Jun 2005 09:41:41 +0000 (09:41 +0000)]
commit the file

Original commit message from CVS:
commit the file

19 years agogst/gstpad.c (gst_pad_set_caps): Remove needless refs, gst_caps_replace takes care...
Andy Wingo [Thu, 23 Jun 2005 09:41:09 +0000 (09:41 +0000)]
gst/gstpad.c (gst_pad_set_caps): Remove needless refs, gst_caps_replace takes care of this for us.

Original commit message from CVS:
2005-06-23  Andy Wingo  <wingo@pobox.com>

* gst/gstpad.c (gst_pad_set_caps): Remove needless refs,
gst_caps_replace takes care of this for us.

19 years agogst/gstghostpad.c (gst_proxy_pad_do_setcaps): Call the full gst_pad_set_caps on the...
Andy Wingo [Thu, 23 Jun 2005 09:28:27 +0000 (09:28 +0000)]
gst/gstghostpad.c (gst_proxy_pad_do_setcaps): Call the full gst_pad_set_caps on the target, not just its setcaps() fu...

Original commit message from CVS:
2005-06-23  Andy Wingo  <wingo@pobox.com>

* gst/gstghostpad.c (gst_proxy_pad_do_setcaps): Call the full
gst_pad_set_caps on the target, not just its setcaps() function.

19 years agotests/: A network clock simulator.
Andy Wingo [Thu, 23 Jun 2005 00:39:26 +0000 (00:39 +0000)]
tests/: A network clock simulator.

Original commit message from CVS:
2005-06-23  Andy Wingo  <wingo@pobox.com>

* tests/network-clock.scm:
* tests/network-clock-utils.scm: A network clock simulator.
Something of an algorithmic testbed before doing something in C.

19 years agomake sure capslist.h gets disted
Thomas Vander Stichele [Wed, 22 Jun 2005 19:57:12 +0000 (19:57 +0000)]
make sure capslist.h gets disted

Original commit message from CVS:
make sure capslist.h gets disted

19 years agofile fromstring.c was initially added on branch BRANCH-GSTREAMER-0_8.
Thomas Vander Stichele [Wed, 22 Jun 2005 19:48:54 +0000 (19:48 +0000)]
file fromstring.c was initially added on branch BRANCH-GSTREAMER-0_8.

Original commit message from CVS:
file fromstring.c was initially added on branch BRANCH-GSTREAMER-0_8.

19 years agocheck/: copy over from 0.8, and add two with bitmasks specified with (int) 0xFF...
Thomas Vander Stichele [Wed, 22 Jun 2005 19:22:34 +0000 (19:22 +0000)]
check/: copy over from 0.8, and add two with bitmasks specified with (int) 0xFF...

Original commit message from CVS:
* check/Makefile.am:
* check/gst/capslist.h:
copy over from 0.8, and add two with bitmasks specified with
(int) 0xFF...
* check/gst/gstcaps.c: (START_TEST), (gst_caps_suite):
add test to parse everything from capslist.h
* check/gst/gststructure.c: (START_TEST), (gst_value_suite),
(main):
add test for structure deserialization
* check/gst/gstvalue.c: (START_TEST), (gst_value_suite):
add tests for deserialization of strings to int types
* gst/gststructure.c: (gst_structure_nth_field_name):
* gst/gststructure.h:
add a way to get the name of a field referenced by index
* gst/gstvalue.c: (gst_value_deserialize_int_helper):
instead of checking if the resulting long long lies between
min and max, we check if the long long would fit into
a number of bytes for the final type.
This fixes cases where a string represents 2^32 - 1, which
when cast to int would be the (valid) -1, but is bigger than
G_MAXINT

19 years agoadd a log line for type deserialization
Thomas Vander Stichele [Wed, 22 Jun 2005 11:02:57 +0000 (11:02 +0000)]
add a log line for type deserialization

Original commit message from CVS:
add a log line for type deserialization

19 years agoreturn long long, not int, so gint64 deserialization actually works. Is there any...
Thomas Vander Stichele [Wed, 22 Jun 2005 10:52:18 +0000 (10:52 +0000)]
return long long, not int, so gint64 deserialization actually works.  Is there any flag that makes the compiler check...

Original commit message from CVS:

* check/gst/gstvalue.c: (START_TEST):
* gst/gstvalue.c: (gst_value_deserialize):
return long long, not int, so gint64 deserialization actually
works.  Is there any flag that makes the compiler check this ?
Fixes #308559

19 years agogst/gstbuffer.h: Added convenience macros for setting buffers in GValue.
Wim Taymans [Wed, 22 Jun 2005 09:55:16 +0000 (09:55 +0000)]
gst/gstbuffer.h: Added convenience macros for setting buffers in GValue.

Original commit message from CVS:
* gst/gstbuffer.h:
Added convenience macros for setting buffers in GValue.

19 years agocheck/gst/: add a test deserializing int64, and comment part out because it fails...
Thomas Vander Stichele [Tue, 21 Jun 2005 17:41:35 +0000 (17:41 +0000)]
check/gst/: add a test deserializing int64, and comment part out because it fails, yay !

Original commit message from CVS:
* check/gst/.cvsignore:
* check/gst/gstvalue.c: (START_TEST), (gst_value_suite):
add a test deserializing int64, and comment part out because
it fails, yay !

19 years agocommit a file I forgot
Thomas Vander Stichele [Tue, 21 Jun 2005 16:53:14 +0000 (16:53 +0000)]
commit a file I forgot

Original commit message from CVS:
commit a file I forgot

19 years agomove over a value_serialize test
Thomas Vander Stichele [Tue, 21 Jun 2005 16:48:46 +0000 (16:48 +0000)]
move over a value_serialize test

Original commit message from CVS:
move over a value_serialize test

19 years agogst/gstpad.c: Small doc updates.
Wim Taymans [Mon, 20 Jun 2005 15:18:17 +0000 (15:18 +0000)]
gst/gstpad.c: Small doc updates.

Original commit message from CVS:
* gst/gstpad.c:
Small doc updates.

* gst/gstvalue.c: (gst_value_compare_buffer),
(gst_value_serialize_buffer), (gst_value_deserialize_buffer),
(gst_value_compare_flags), (gst_value_serialize_flags),
(gst_value_deserialize_flags), (_gst_value_initialize):
Fix serialisation of buffers, they are not boxed types anymore

19 years agocheck/gst/gstcaps.c: Testcase to show error in buffer-on-caps serialisation.
Wim Taymans [Mon, 20 Jun 2005 15:14:58 +0000 (15:14 +0000)]
check/gst/gstcaps.c: Testcase to show error in buffer-on-caps serialisation.

Original commit message from CVS:
* check/gst/gstcaps.c: (START_TEST), (gst_caps_suite):
Testcase to show error in buffer-on-caps serialisation.

19 years agodocs/random/wingo/porting-plugins-to-0.9: A pitiful document I will be adding to...
Andy Wingo [Mon, 20 Jun 2005 15:13:43 +0000 (15:13 +0000)]
docs/random/wingo/porting-plugins-to-0.9: A pitiful document I will be adding to later.

Original commit message from CVS:
2005-06-20  Andy Wingo  <wingo@pobox.com>

* docs/random/wingo/porting-plugins-to-0.9: A pitiful document I
will be adding to later.

19 years agogst/gstsystemclock.c (gst_system_clock_init): Unlock the clock if its socks fill...
Andy Wingo [Mon, 20 Jun 2005 11:41:17 +0000 (11:41 +0000)]
gst/gstsystemclock.c (gst_system_clock_init): Unlock the clock if its socks fill with rocks.

Original commit message from CVS:
2005-06-20  Andy Wingo  <wingo@pobox.com>

* gst/gstsystemclock.c (gst_system_clock_init): Unlock the clock
if its socks fill with rocks.
(gst_system_clock_obtain): Set the name on object construction.
Avoid double-checked locking.

19 years agogst/gsturi.c: Fix potential endless loop.
Tim-Philipp Müller [Mon, 20 Jun 2005 11:32:14 +0000 (11:32 +0000)]
gst/gsturi.c: Fix potential endless loop.

Original commit message from CVS:
* gst/gsturi.c: (gst_element_make_from_uri):
Fix potential endless loop.

19 years agoremove another test that's obsolete
Thomas Vander Stichele [Mon, 20 Jun 2005 11:27:36 +0000 (11:27 +0000)]
remove another test that's obsolete

Original commit message from CVS:
remove another test that's obsolete

19 years agoremove clock testsuite, important stuff already moved to check
Thomas Vander Stichele [Mon, 20 Jun 2005 11:23:59 +0000 (11:23 +0000)]
remove clock testsuite, important stuff already moved to check

Original commit message from CVS:
remove clock testsuite, important stuff already moved to check

19 years agoremove test that was already moved to check
Thomas Vander Stichele [Mon, 20 Jun 2005 11:18:40 +0000 (11:18 +0000)]
remove test that was already moved to check

Original commit message from CVS:
remove test that was already moved to check

19 years agocheck/Makefile.am: add gsttag
Thomas Vander Stichele [Sun, 19 Jun 2005 11:32:42 +0000 (11:32 +0000)]
check/Makefile.am: add gsttag

Original commit message from CVS:

* check/Makefile.am:
add gsttag
* check/gst/gsttag.c: (check_tags), (START_TEST), (gst_tag_suite),
(main):
move over from testsuite dir and clean up
* configure.ac:
* gst/gsttag.c:
* testsuite/Makefile.am:
* testsuite/tags/.cvsignore:
* testsuite/tags/Makefile.am:
* testsuite/tags/merge.c:
remove testsuite/tags

19 years agosome more docs cleanup
Thomas Vander Stichele [Sun, 19 Jun 2005 10:54:12 +0000 (10:54 +0000)]
some more docs cleanup

Original commit message from CVS:
some more docs cleanup

19 years agoremove bufspeed and spidey_bench
Thomas Vander Stichele [Sun, 19 Jun 2005 10:31:42 +0000 (10:31 +0000)]
remove bufspeed and spidey_bench

Original commit message from CVS:
remove bufspeed and spidey_bench

19 years agoclean up docs a little
Thomas Vander Stichele [Sun, 19 Jun 2005 10:22:47 +0000 (10:22 +0000)]
clean up docs a little

Original commit message from CVS:
clean up docs a little

19 years agocheck/gstcheck.h: add macros for checking refcounts on objects and caps
Thomas Vander Stichele [Sun, 19 Jun 2005 00:52:31 +0000 (00:52 +0000)]
check/gstcheck.h: add macros for checking refcounts on objects and caps

Original commit message from CVS:

* check/gstcheck.h:
add macros for checking refcounts on objects and caps
* check/gst/gstpad.c: (START_TEST), (gst_pad_suite):
add some more unit tests
* gst/gstpad.c: (gst_pad_link_check_compatible_unlocked),
(gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_allowed_caps):
fix leaked refcounts (I hope :)) so unittest works
* gst/gstpad.h:
whitespace removal

19 years agoback to head
Thomas Vander Stichele [Sat, 18 Jun 2005 22:33:14 +0000 (22:33 +0000)]
back to head

Original commit message from CVS:
back to head

19 years agochangelog
Andy Wingo [Fri, 17 Jun 2005 12:00:35 +0000 (12:00 +0000)]
changelog

Original commit message from CVS:
changelog

19 years agogst/base/gstbasesink.c (gst_basesink_chain): Remove bogus assert; it's always possibl...
Andy Wingo [Fri, 17 Jun 2005 11:58:48 +0000 (11:58 +0000)]
gst/base/gstbasesink.c (gst_basesink_chain): Remove bogus assert; it's always possible that the pad gets deactivated ...

Original commit message from CVS:
2005-06-17  Andy Wingo  <wingo@pobox.com>

* gst/base/gstbasesink.c (gst_basesink_chain): Remove bogus
assert; it's always possible that the pad gets deactivated in
between the checks in gstpad.c and the implementation. Rely on
finish_preroll() to return a FLUSHING or similar instead of on the
assert.

19 years agogst/base/gstbasesink.c (gst_basesink_event): Only wait for the clock and post an...
Andy Wingo [Fri, 17 Jun 2005 11:33:27 +0000 (11:33 +0000)]
gst/base/gstbasesink.c (gst_basesink_event): Only wait for the clock and post an EOS message if we come out of finish...

Original commit message from CVS:
2005-06-17  Andy Wingo  <wingo@pobox.com>

* gst/base/gstbasesink.c (gst_basesink_event): Only wait for the
clock and post an EOS message if we come out of finish_preroll in
the playing state.

19 years agoincorporate plugins stuff and uninstalled stuff
Thomas Vander Stichele [Fri, 17 Jun 2005 09:58:38 +0000 (09:58 +0000)]
incorporate plugins stuff and uninstalled stuff

Original commit message from CVS:
incorporate plugins stuff and uninstalled stuff

19 years agoignore more
Thomas Vander Stichele [Fri, 17 Jun 2005 09:32:19 +0000 (09:32 +0000)]
ignore more

Original commit message from CVS:
ignore more

19 years agoignore more
Thomas Vander Stichele [Fri, 17 Jun 2005 09:12:33 +0000 (09:12 +0000)]
ignore more

Original commit message from CVS:
ignore more

19 years agoignore more; fix README
Thomas Vander Stichele [Fri, 17 Jun 2005 08:59:41 +0000 (08:59 +0000)]
ignore more; fix README

Original commit message from CVS:
ignore more; fix README

19 years agogst/elements/gstcapsfilter.c: Allow NULL as possible value for filter_caps property...
David Schleef [Thu, 16 Jun 2005 17:50:16 +0000 (17:50 +0000)]
gst/elements/gstcapsfilter.c: Allow NULL as possible value for filter_caps property, indicating GST_CAPS_ANY.

Original commit message from CVS:
* gst/elements/gstcapsfilter.c: (gst_capsfilter_class_init),
(gst_capsfilter_set_property): Allow NULL as possible value
for filter_caps property, indicating GST_CAPS_ANY.

19 years agogst/elements/gstfakesrc.c: fix debug output
Thomas Vander Stichele [Thu, 9 Jun 2005 13:33:00 +0000 (13:33 +0000)]
gst/elements/gstfakesrc.c: fix debug output

Original commit message from CVS:

* gst/elements/gstfakesrc.c: (gst_fakesrc_create):
fix debug output
* gst/schedulers/Makefile.am:
use libgst prefix
* gstreamer.spec.in:
fix spec for it

19 years agogstreamer.spec.in: clean up
Thomas Vander Stichele [Thu, 9 Jun 2005 12:23:05 +0000 (12:23 +0000)]
gstreamer.spec.in: clean up

Original commit message from CVS:

* gstreamer.spec.in:
clean up

19 years agogstreamer.spec.in: clean up
Thomas Vander Stichele [Thu, 9 Jun 2005 12:09:56 +0000 (12:09 +0000)]
gstreamer.spec.in: clean up

Original commit message from CVS:

* gstreamer.spec.in:
clean up

19 years agohave a real TODO, move old TODO
Thomas Vander Stichele [Thu, 9 Jun 2005 12:03:04 +0000 (12:03 +0000)]
have a real TODO, move old TODO

Original commit message from CVS:
have a real TODO, move old TODO

19 years agopo updates
Thomas Vander Stichele [Thu, 9 Jun 2005 12:00:27 +0000 (12:00 +0000)]
po updates

Original commit message from CVS:
po updates

19 years agoupdate translations
Thomas Vander Stichele [Thu, 9 Jun 2005 11:12:57 +0000 (11:12 +0000)]
update translations

Original commit message from CVS:
update translations

19 years agogst/gstutils.c: RPAD fixes all around.
Andy Wingo [Wed, 8 Jun 2005 22:16:27 +0000 (22:16 +0000)]
gst/gstutils.c: RPAD fixes all around.

Original commit message from CVS:
2005-06-08  Andy Wingo  <wingo@pobox.com>

* gst/gstutils.c: RPAD fixes all around.
(gst_element_link_pads): Refcounting fixes.

* tools/gst-inspect.c:
* tools/gst-xmlinspect.c:
* parse/grammar.y:
* gst/base/gsttypefindhelper.c:
* gst/base/gstbasesink.c:
* gst/gstqueue.c: RPAD fixes.

* gst/gstghostpad.h:
* gst/gstghostpad.c: New ghost pad implementation as full proxy
pads. The tricky thing is they provide both source and sink
interfaces, since they proxy the internal pad for the external
pad, and vice versa. Implement with lower-level ProxyPad objects,
with the interior proxy pad as a child of the exterior ghost pad.
Should write a doc on this.

* gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/.
(gst_pad_set_name, gst_pad_set_parent): Macros removed, use
gst_object API.

* gst/gstpad.c: Big changes. No more stub base GstPad, now all
pads are real pads. No ghost pads in this file. Not documenting
the myriad s/RPAD/PAD/ and REALIZE fixes.
(gst_pad_class_init): Add properties for "direction" and
"template". Both are construct-only, so they can't change during
the life of the pad. Fixes properly deriving from GstPad.
(gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For
derived objects, just set properties when creating the objects via
g_object_new.
(gst_pad_get_parent): Implement as a function, return NULL if the
parent is not an element.
(gst_pad_get_real_parent, gst_pad_add_ghost_pad)
(gst_pad_remove_ghost_pad, gst_pad_realize): Removed.

* gst/gstobject.c (gst_object_class_init): Make name a construct
property. Don't set it in the object init.

* gst/gstelement.c (gst_element_add_pad): Don't allow adding pads
with UNKNOWN direction.
(gst_element_add_ghost_pad): Remove non-orthogonal API. Replace
with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)).
(gst_element_remove_pad): Remove ghost-pad special cases.
(gst_element_pads_activate): Remove rpad cruft.

* gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to
catch the pad's-parent-not-an-element case.

* gst/gst.h: Include gstghostpad.h.

* gst/gst.c (init_post): No more real, ghost pads.

* gst/Makefile.am: Add gstghostpad.[ch].

* check/Makefile.am:
* check/gst/gstbin.c:
* check/gst/gstghostpad.c (test_ghost_pads): Check that linking
into a bin creates ghost pads, and that the refcounts are right.
Partly moved from gstbin.c.

19 years agocheck/: ignore more
Thomas Vander Stichele [Wed, 8 Jun 2005 14:00:21 +0000 (14:00 +0000)]
check/: ignore more

Original commit message from CVS:
* check/gst-libs/.cvsignore:
* check/gst/.cvsignore:
* check/pipelines/.cvsignore:
ignore more
* check/pipelines/cleanup.c: (setup_pipeline), (run_pipeline),
(START_TEST), (cleanup_suite), (main):
add some tests related to cleanup after running pipelines

19 years agoadd a GstBuffer unit test
Thomas Vander Stichele [Wed, 8 Jun 2005 13:57:13 +0000 (13:57 +0000)]
add a GstBuffer unit test

Original commit message from CVS:
add a GstBuffer unit test

19 years agoprevious commit accidentally also added refcount defines for gstminiobject, logging...
Thomas Vander Stichele [Wed, 8 Jun 2005 13:45:26 +0000 (13:45 +0000)]
previous commit accidentally also added refcount defines for gstminiobject, logging that now

Original commit message from CVS:
previous commit accidentally also added refcount defines for gstminiobject, logging that now

19 years agoadd a 'plugins' dir to the PLUGIN_PATH in the uninstalled script to drop random other...
Thomas Vander Stichele [Wed, 8 Jun 2005 13:42:52 +0000 (13:42 +0000)]
add a 'plugins' dir to the PLUGIN_PATH in the uninstalled script to drop random other plugin-having projects in

Original commit message from CVS:
add a 'plugins' dir to the PLUGIN_PATH in the uninstalled script to drop random other plugin-having projects in

19 years agowe did a prerelease
Thomas Vander Stichele [Wed, 8 Jun 2005 13:41:48 +0000 (13:41 +0000)]
we did a prerelease

Original commit message from CVS:
we did a prerelease

19 years agoOBJECT acts on obj not caps
Thomas Vander Stichele [Wed, 8 Jun 2005 13:41:31 +0000 (13:41 +0000)]
OBJECT acts on obj not caps

Original commit message from CVS:
OBJECT acts on obj not caps

19 years agoadd a debug line
Thomas Vander Stichele [Wed, 8 Jun 2005 13:41:05 +0000 (13:41 +0000)]
add a debug line

Original commit message from CVS:
add a debug line

19 years agowhite space fixes
Thomas Vander Stichele [Wed, 8 Jun 2005 13:40:46 +0000 (13:40 +0000)]
white space fixes

Original commit message from CVS:
white space fixes

19 years agoadded support for html unit test coverage reports
Stefan Kost [Fri, 3 Jun 2005 18:26:59 +0000 (18:26 +0000)]
added support for html unit test coverage reports

Original commit message from CVS:
added support for html unit test coverage reports

19 years agogst/elements/gstcapsfilter.c: Free existing caps if the capsfilter changes. Add a...
Jan Schmidt [Thu, 2 Jun 2005 15:45:16 +0000 (15:45 +0000)]
gst/elements/gstcapsfilter.c: Free existing caps if the capsfilter changes. Add a FIXME about setting those caps on t...

Original commit message from CVS:
* gst/elements/gstcapsfilter.c: (gst_capsfilter_set_property):
Free existing caps if the capsfilter changes. Add a FIXME about
setting those caps on the pads.

* gst/gstutils.c: (gst_element_get_compatible_pad), (ghost_up):
Before adding a ghost pad to a parent bin, check that there isn't
already one for the element on the bin. Prevents infinite recursion
when using decodebin in parse pipelines. Andy says he'll rewrite the
way this works anyway, so ignore the hack.

19 years agogst/elements/gsttypefindelement.c (do_pull_typefind): Query the file size, pass it...
Andy Wingo [Thu, 2 Jun 2005 11:12:34 +0000 (11:12 +0000)]
gst/elements/gsttypefindelement.c (do_pull_typefind): Query the file size, pass it on to the type find helper.

Original commit message from CVS:
2005-06-02  Andy Wingo  <wingo@pobox.com>

* gst/elements/gsttypefindelement.c (do_pull_typefind): Query the
file size, pass it on to the type find helper.

* gst/base/gstbasesrc.c (gst_basesrc_do_seek): Set the
segment_start and segment_end properly according to the seek
method. Segment_end is still a bit flaky because offset can be
negative for CUR and END cases, but it takes -1 as an "unset"
value.

19 years agogst/: Bufferalloc: return GstFlowReturn to more accuratly report why allocation failed.
Wim Taymans [Thu, 2 Jun 2005 09:42:02 +0000 (09:42 +0000)]
gst/: Bufferalloc: return GstFlowReturn to more accuratly report why allocation failed.

Original commit message from CVS:
* gst/base/gstbasesink.c: (gst_basesink_pad_buffer_alloc),
(gst_base_sink_buffer_alloc), (gst_basesink_preroll_queue_push),
(gst_basesink_activate):
* gst/base/gstbasesink.h:
* gst/gstpad.c: (gst_pad_set_active), (gst_pad_link_prepare),
(gst_pad_link), (gst_pad_accept_caps), (gst_pad_alloc_buffer),
(gst_pad_query), (gst_pad_start_task):
* gst/gstpad.h:
* gst/gstqueue.c: (gst_queue_bufferalloc),
(gst_queue_handle_sink_event), (gst_queue_chain):
Bufferalloc: return GstFlowReturn to more accuratly report
why allocation failed.

19 years agogst/gstpipeline.c: Take snapshot of state without blocking.
Wim Taymans [Thu, 2 Jun 2005 09:39:21 +0000 (09:39 +0000)]
gst/gstpipeline.c: Take snapshot of state without blocking.

Original commit message from CVS:
* gst/gstpipeline.c: (gst_pipeline_send_event):
Take snapshot of state without blocking.

19 years agodocs/design/: Small doc updates
Wim Taymans [Thu, 2 Jun 2005 08:26:58 +0000 (08:26 +0000)]
docs/design/: Small doc updates

Original commit message from CVS:
* docs/design/part-TODO.txt:
* docs/design/part-caps.txt:
* docs/design/part-clocks.txt:
* docs/design/part-negotiation.txt:
* docs/design/part-preroll.txt:
Small doc updates

19 years agogst/elements/gstidentity.c: Protect last_message property as it is accessed from...
Wim Taymans [Mon, 30 May 2005 16:28:41 +0000 (16:28 +0000)]
gst/elements/gstidentity.c: Protect last_message property as it is accessed from multiple threads.

Original commit message from CVS:
* gst/elements/gstidentity.c: (gst_identity_event),
(gst_identity_transform), (gst_identity_get_property):
Protect last_message property as it is accessed from
multiple threads.

19 years agogst/gstelement.c: Slicker pad activation code.
Wim Taymans [Mon, 30 May 2005 15:53:04 +0000 (15:53 +0000)]
gst/gstelement.c: Slicker pad activation code.

Original commit message from CVS:
* gst/gstelement.c: (gst_element_init),
(gst_element_pads_activate), (gst_element_change_state):
Slicker pad activation code.

19 years agogst/: Move elementfactory methods to separate .h file.
Wim Taymans [Mon, 30 May 2005 15:51:40 +0000 (15:51 +0000)]
gst/: Move elementfactory methods to separate .h file.

Original commit message from CVS:
* gst/Makefile.am:
* gst/gstelement.h:
* gst/gstelementfactory.h:
* gst/gsttypes.h:
Move elementfactory methods to separate .h file.

19 years agoSmall typo fixes, doc updates.
Wim Taymans [Mon, 30 May 2005 15:48:45 +0000 (15:48 +0000)]
Small typo fixes, doc updates.

Original commit message from CVS:
* docs/design/part-overview.txt:
* gst/gstsystemclock.h:
Small typo fixes, doc updates.

19 years agogst/gst.c: Remove cpu-opt flag.
Wim Taymans [Mon, 30 May 2005 15:46:15 +0000 (15:46 +0000)]
gst/gst.c: Remove cpu-opt flag.

Original commit message from CVS:
* gst/gst.c: (gst_init_get_popt_table), (init_post),
(init_popt_callback):
Remove cpu-opt flag.

19 years agogst/gstbuffer.*: Avoid typechecking in places where not needed.
Wim Taymans [Mon, 30 May 2005 15:44:50 +0000 (15:44 +0000)]
gst/gstbuffer.*: Avoid typechecking in places where not needed.

Original commit message from CVS:
* gst/gstbuffer.c: (gst_subbuffer_finalize),
(gst_buffer_create_sub), (gst_buffer_is_span_fast):
* gst/gstbuffer.h:
Avoid typechecking in places where not needed.
Added accessor for malloc_data.

19 years agogst/gstpad.c: Propagate errors from _set_caps() in configure_src/sink functions inste...
Wim Taymans [Mon, 30 May 2005 15:41:54 +0000 (15:41 +0000)]
gst/gstpad.c: Propagate errors from _set_caps() in configure_src/sink functions instead of returning TRUE.

Original commit message from CVS:
* gst/gstpad.c: (gst_real_pad_init), (gst_pad_set_active),
(gst_pad_link_prepare), (gst_pad_link), (gst_pad_accept_caps),
(gst_pad_configure_sink), (gst_pad_configure_src),
(gst_pad_alloc_buffer), (gst_pad_query), (gst_pad_send_event),
(gst_pad_start_task):
Propagate errors from _set_caps() in configure_src/sink
functions instead of returning TRUE.
FLUSH events can travel up and downstream

19 years agogst/base/gstbasesink.c: Handle EOS in preroll.
Wim Taymans [Mon, 30 May 2005 15:36:09 +0000 (15:36 +0000)]
gst/base/gstbasesink.c: Handle EOS in preroll.

Original commit message from CVS:
* gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push),
(gst_basesink_activate):
Handle EOS in preroll.

19 years agogst/gstqueue.c: Remove old pieces of code
Wim Taymans [Mon, 30 May 2005 15:34:13 +0000 (15:34 +0000)]
gst/gstqueue.c: Remove old pieces of code

Original commit message from CVS:
* gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain),
(gst_queue_loop), (gst_queue_handle_src_event):
Remove old pieces of code
Flushing the queue in an upstream event is a very bad idea.

19 years agofile gstsimplescheduler.c was initially added on branch BRANCH-COMPANY.
Benjamin Otte [Sun, 29 May 2005 13:56:55 +0000 (13:56 +0000)]
file gstsimplescheduler.c was initially added on branch BRANCH-COMPANY.

Original commit message from CVS:
file gstsimplescheduler.c was initially added on branch BRANCH-COMPANY.

19 years agoremove conflict doobers
Andy Wingo [Fri, 27 May 2005 09:28:05 +0000 (09:28 +0000)]
remove conflict doobers

Original commit message from CVS:
remove conflict doobers

19 years agogst/gstminiobject.c (gst_value_mini_object_collect): Use gst_value_set_mini_object...
Andy Wingo [Fri, 27 May 2005 09:27:35 +0000 (09:27 +0000)]
gst/gstminiobject.c (gst_value_mini_object_collect): Use gst_value_set_mini_object so as to add a ref on the object (...

Original commit message from CVS:
2005-05-26  Andy Wingo  <wingo@pobox.com>

* gst/gstminiobject.c (gst_value_mini_object_collect): Use
gst_value_set_mini_object so as to add a ref on the object (which
will be removed when the value is unset).

* gst/elements/gstfakesink.c (gst_fakesink_class_init): Fix signal
arg type in ::handoff.

* gst/gstelement.c (gst_element_change_state): Also deactivate
pads in READY->NULL, just in case the element didn't make it to
PAUSED. Wingo tested, Wim approved.

19 years agogst/gstpad.c: A flushing pad cannot be used to alloc_buffer from.
Wim Taymans [Thu, 26 May 2005 10:50:12 +0000 (10:50 +0000)]
gst/gstpad.c: A flushing pad cannot be used to alloc_buffer from.

Original commit message from CVS:
* gst/gstpad.c: (gst_pad_set_active), (gst_pad_link_prepare),
(gst_pad_link), (gst_pad_accept_caps), (gst_pad_alloc_buffer),
(gst_pad_query), (gst_pad_send_event), (gst_pad_start_task):
A flushing pad cannot be used to alloc_buffer from.

19 years agogst/gstbus.*: Implement a real GSource and use g_main_context_wakeup() to signal...
Wim Taymans [Thu, 26 May 2005 10:48:53 +0000 (10:48 +0000)]
gst/gstbus.*: Implement a real GSource and use g_main_context_wakeup() to signal new messages instead of the socketpair.

Original commit message from CVS:
* gst/gstbus.c: (gst_bus_init), (gst_bus_dispose), (gst_bus_post),
(gst_bus_pop), (gst_bus_source_prepare), (gst_bus_source_check),
(gst_bus_source_dispatch), (gst_bus_source_finalize),
(gst_bus_create_watch), (gst_bus_add_watch_full):
* gst/gstbus.h:
Implement a real GSource and use g_main_context_wakeup() to
signal new messages instead of the socketpair.

19 years agogst/: Fix state changes for non sinks. We now change sinks, then elements with unconn...
Wim Taymans [Wed, 25 May 2005 19:33:39 +0000 (19:33 +0000)]
gst/: Fix state changes for non sinks. We now change sinks, then elements with unconnected srcpads, then the rest.

Original commit message from CVS:
* gst/gstbin.c: (bin_element_is_sink), (has_ancestor),
(bin_element_is_semi_sink), (append_child), (gst_bin_change_state):
* gst/gstpad.c: (gst_pad_set_active), (gst_pad_link_prepare),
(gst_pad_link), (gst_pad_accept_caps), (gst_pad_query),
(gst_pad_send_event), (gst_pad_start_task):
* gst/gstqueue.c: (gst_queue_init), (gst_queue_locked_flush),
(gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop),
(gst_queue_sink_activate), (gst_queue_src_activate),
(gst_queue_change_state):
* gst/gstqueue.h:
Fix state changes for non sinks. We now change sinks, then elements
with unconnected srcpads, then the rest.
More efficient queue unlocking in flush and state changes.
Set the pad activate mode even if it does not have an activate
function.

19 years agohappify buildbot
Ronald S. Bultje [Wed, 25 May 2005 16:09:34 +0000 (16:09 +0000)]
happify buildbot

Original commit message from CVS:
happify buildbot

19 years agogst/base/gstbasesrc.c: Don't go in pull mode for non-seekable sources.
Ronald S. Bultje [Wed, 25 May 2005 15:57:57 +0000 (15:57 +0000)]
gst/base/gstbasesrc.c: Don't go in pull mode for non-seekable sources.

Original commit message from CVS:
* gst/base/gstbasesrc.c: (gst_basesrc_activate):
Don't go in pull mode for non-seekable sources.
* gst/elements/gsttypefindelement.c: (gst_type_find_element_init),
(gst_type_find_element_dispose), (gst_type_find_handle_src_query),
(free_entry), (stop_typefinding),
(gst_type_find_element_handle_event), (find_peek),
(gst_type_find_element_chain), (do_pull_typefind),
(gst_type_find_element_change_state):
Allow typefinding (w/o seeking) in push-mode, simplified version
of what was in 0.8.
* gst/gstutils.c: (gst_buffer_join):
* gst/gstutils.h:
gst_buffer_join() from 0.8.

19 years agogst/gstpad.c: Disable attempt at mode switching until it is figured out.
Wim Taymans [Wed, 25 May 2005 13:59:18 +0000 (13:59 +0000)]
gst/gstpad.c: Disable attempt at mode switching until it is figured out.

Original commit message from CVS:
* gst/gstpad.c: (gst_pad_set_active), (gst_pad_link_prepare),
(gst_pad_link), (gst_pad_accept_caps), (gst_pad_query),
(gst_pad_send_event), (gst_pad_start_task):
Disable attempt at mode switching until it is figured out.

19 years agogst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function.
Wim Taymans [Wed, 25 May 2005 11:50:11 +0000 (11:50 +0000)]
gst/: Implement gst_pad_pause/start/stop_task(), take STREAM lock in task function.

Original commit message from CVS:
* gst/base/gstadapter.c: (gst_adapter_peek), (gst_adapter_flush):
* gst/base/gstbasesink.c: (gst_basesink_preroll_queue_push),
(gst_basesink_finish_preroll), (gst_basesink_chain),
(gst_basesink_loop), (gst_basesink_activate),
(gst_basesink_change_state):
* gst/base/gstbasesrc.c: (gst_basesrc_do_seek),
(gst_basesrc_get_range), (gst_basesrc_loop),
(gst_basesrc_activate):
* gst/elements/gsttee.c: (gst_tee_sink_activate):
* gst/gstpad.c: (gst_pad_dispose), (gst_real_pad_class_init),
(gst_real_pad_init), (gst_real_pad_set_property),
(gst_real_pad_get_property), (gst_pad_set_active),
(gst_pad_is_active), (gst_pad_get_query_types), (gst_pad_unlink),
(gst_pad_link_prepare), (gst_pad_link), (gst_pad_get_real_parent),
(gst_real_pad_get_caps_unlocked), (gst_pad_peer_get_caps),
(gst_pad_accept_caps), (gst_pad_get_peer), (gst_pad_realize),
(gst_pad_event_default_dispatch), (gst_pad_event_default),
(gst_pad_dispatcher), (gst_pad_query), (gst_real_pad_dispose),
(gst_pad_save_thyself), (handle_pad_block), (gst_pad_chain),
(gst_pad_push), (gst_pad_get_range), (gst_pad_pull_range),
(gst_pad_send_event), (gst_pad_start_task), (gst_pad_pause_task),
(gst_pad_stop_task):
* gst/gstpad.h:
* gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain),
(gst_queue_loop), (gst_queue_src_activate):
* gst/gsttask.c: (gst_task_init), (gst_task_set_lock),
(gst_task_get_state):
* gst/gsttask.h:
* gst/schedulers/threadscheduler.c:
(gst_thread_scheduler_task_start), (gst_thread_scheduler_func):
Implement gst_pad_pause/start/stop_task(), take STREAM lock
in task function.
Remove ACTIVE pad flag, use FLUSHING everywhere
Added _pad_chain(), _pad_get_range() to call chain/getrange
functions.
Add locks around IS_FLUSHING when reading.
Take STREAM lock in chain(), get_range() functions so plugins
don't need to take it anymore.