Original commit message from CVS:
2007-01-12 Andy Wingo <wingo@pobox.com>
* libs/gst/base/gstbasetransform.c (gst_base_transform_activate):
Factor start/stop into this private function instead of partially
in activate functions and partially in the change_state function.
Fixes setup before the element has changed from READY->PAUSED, as
is the case in pull-mode pipelines.
(gst_base_transform_sink_activate_push)
(gst_base_transform_src_activate_pull): Refactor to use
gst_base_transform_activate().
(gst_base_transform_change_state): Removed, not needed any more.
* libs/gst/base/gstbasesink.c (gst_base_sink_negotiate_pull):
Truncate before fixating.
Original commit message from CVS:
2007-01-12 Andy Wingo <wingo@pobox.com>
* libs/gst/base/gstbasesink.c (gst_base_sink_negotiate_pull):
Don't set_caps() if the result of fixating is ANY, as it's not
supported, and not necessary in the case of a link with no
template caps on either side. Fixes tests/check/libs/basesrc in
some pull-mode tests.
Original commit message from CVS:
2007-01-12 Andy Wingo <wingo@pobox.com>
* libs/gst/base/gstbasetransform.c (_GstBaseTransformPrivate):
(gst_base_transform_init, gst_base_transform_sink_activate_push)
(gst_base_transform_src_activate_pull):
Track the activation mode.
(gst_base_transform_setcaps): In pull mode, when activating the
src pad, after activating the sink pad, activate the sink pad's
peer, as discussed in part-negotiation.txt.
* libs/gst/base/gstbasesrc.h:
* libs/gst/base/gstbasesrc.c (gst_base_src_fixate): Add fixate
vmethod, as in basesink.
* libs/gst/base/gstbasesink.h: Reformat docs, add fixate vmethod.
* libs/gst/base/gstbasesink.c (gst_base_sink_pad_setcaps): In pull
mode, first proxy the setcaps to the peer pad.
(gst_base_sink_pad_fixate): Add a fixate function that calls the
new fixate vmethod.
(gst_base_sink_default_activate_pull): Rename from
gst_base_sink_activate_pull.
(gst_base_sink_negotiate_pull): New function, performs negotiation
in pull mode before calling ::activate_pull().
(gst_base_sink_pad_activate_pull): Actually call the activate_pull
vmethod instead of the default implementation. I have no idea how
this worked before. Negotiate before calling activate_pull.
Original commit message from CVS:
2007-01-06 Andy Wingo <wingo@pobox.com>
* libs/gst/base/gstbasesink.h: New GstBaseSinkClass vmethod,
activate_pull(), providing for a way to specialize the process of
spawning a thread to pull on the sink pad. There is a default
implementation.
* libs/gst/base/gstbasesink.c (gst_base_sink_pad_activate_pull)
(gst_base_sink_pad_activate_push, gst_base_sink_pad_activate)
(gst_base_sink_init): Renamed pad activation functions (inserting
"_pad" in their names). Refactor to use the new activate_pull
vmethod, as appropriate.
(gst_base_sink_class_init, gst_base_sink_activate_pull): Set the
default activate_pull function to start a task pulling from the
sink pad, as before.
Original commit message from CVS:
* gst/gstelement.h:
Clarify _NO_PREROLL a bit more.
* gst/gstevent.c:
Fix docs.
* gst/gstpad.c: (gst_pad_link_check_hierarchy),
(gst_pad_get_caps_unlocked), (gst_pad_save_thyself),
(handle_pad_block), (gst_pad_push_event), (gst_pad_send_event):
Patch by: Yves Lefebvre <ivanohe at abacom dot com> Fix possible deadlock
due to wrong locking order. Fixes#361769.
Remove some redundant/misplaced checks in pad_block.
* libs/gst/base/gstbasesink.c: (gst_base_sink_get_position):
For negative rates, count backwards from the duration.
Original commit message from CVS:
2006-10-09 Zaheer Abbas Merali <zaheerabbas at merali dot org>
Patch by: Josep Torre Valles <josep@fluendo.com>
* common/m4/gst-error.m4:
Disable warning of statement not reached on Forte.
* gst/gstmessage.h:
Fix warning on Forte (value doesn't fit on enumeration).
* libs/gst/base/gstbasesink.c: (gst_base_sink_chain_unlocked):
Fix warning on Forte (value doesn't fit on enumeration).
* libs/gst/net/gstnetclientclock.c: (gst_net_client_clock_thread):
DEBUG macro says it takes minimum of 2 args and so Forte
complains about the use with just 1 arg.
* plugins/elements/gstfdsink.c:
* plugins/elements/gstfdsrc.c:
* plugins/elements/gstfilesink.c:
* plugins/elements/gstfilesrc.c:
Use correct return type for the uri handler implementations.
All these fix warnings in Forte. Fixes bug #360860.
Original commit message from CVS:
* docs/random/moving-plugins:
Make it clear that the "compiled-in descriptions" really mean
the element details.
* libs/gst/base/gstbasesink.c: (gst_base_sink_commit_state),
(gst_base_sink_wait_preroll):
Update docs.
* docs/libs/gstreamer-libs-sections.txt:
* libs/gst/base/gstbasesrc.c: (gst_base_src_wait_playing),
(gst_base_src_get_range), (gst_base_src_activate_push):
* libs/gst/base/gstbasesrc.h:
Added function to block while waiting for PLAYING, this function
is used by live sources that block on the clock.
API: gst_base_src_wait_playing()
Original commit message from CVS:
* libs/gst/base/gstbasesink.c: (gst_base_sink_commit_state),
(gst_base_sink_wait_preroll), (gst_base_sink_do_sync),
(gst_base_sink_preroll_object):
Make sure that our internal state is correct when we commit our state
asynchronously. This solves a race where a state change to PLAYING
could cause the sink to remain blocked in preroll in some situations.
Original commit message from CVS:
* docs/libs/gstreamer-libs-sections.txt:
* libs/gst/base/gstbasesink.c: (gst_base_sink_wait_preroll),
(gst_base_sink_do_sync):
* libs/gst/base/gstbasesink.h:
Expose logic to wait for preroll so that subclasses such as audiosink
can also use this method.
API: gst_base_sink_wait_preroll()
Original commit message from CVS:
* libs/gst/base/gstbasesink.c: (gst_base_sink_get_sync_times):
Inform GstSegment of the last stop position in order for the current
segment to have a proper duration if it doesn't have a specific stop
position from which a duration could be calculated.
This bug was noticeable when a non-flushing, non-update new segment was
followed by another segment (all buffers from the new segment were being
dropped).
Original commit message from CVS:
* libs/gst/base/gstbasesink.c: (gst_base_sink_preroll_object),
(gst_base_sink_loop), (gst_base_sink_get_position):
When operating in pull mode update the offset so that we
read sequentially.
Original commit message from CVS:
* libs/gst/base/gstbasesink.c: (gst_base_sink_configure_segment),
(gst_base_sink_preroll_object), (gst_base_sink_get_position):
* libs/gst/base/gstbasetransform.c:
(gst_base_transform_sink_eventfunc):
* libs/gst/base/gstcollectpads.c: (gst_collect_pads_event):
Don't randomly and silently reset a segment when the format
changes as this is a bug somewhere upstream. Fixes#330379.
Original commit message from CVS:
* libs/gst/base/gstbasesink.c: (gst_base_sink_get_position):
Don't report POSITION based on clock time if sync is
disabled in a sink.
Original commit message from CVS:
* libs/gst/base/gstbasesink.c: (gst_base_sink_configure_segment):
Add applied_rate to the debug
* libs/gst/base/gstbasesrc.c: (gst_base_src_perform_seek):
Copy applied_rate into the outgoing NEWSEGMENT event
Original commit message from CVS:
Patch by: Philippe Rouquier <philippero at libertysurf dot fr>
* libs/gst/base/gstbasesink.c: (gst_base_sink_set_flushing),
(gst_base_sink_change_state):
call ::unlock before taking the PREROLL_LOCK so we can safely
handle elements that lock in ::render.
Fixes#340174.
Original commit message from CVS:
* docs/design/part-overview.txt:
Make upsteam/downstream concepts more clear.
Give an example of serialized/non-serialized events.
* docs/design/part-events.txt:
* docs/design/part-streams.txt:
Mention applied_rate.
* docs/design/part-trickmodes.txt:
Mention applied rate, flesh out some more use cases.
* gst/gstevent.c: (gst_event_new_new_segment),
(gst_event_parse_new_segment), (gst_event_new_new_segment_full),
(gst_event_parse_new_segment_full), (gst_event_new_tag),
(gst_event_parse_tag), (gst_event_new_buffer_size),
(gst_event_parse_buffer_size), (gst_event_new_qos),
(gst_event_parse_qos), (gst_event_parse_seek),
(gst_event_new_navigation):
* gst/gstevent.h:
Add applied_rate field to NEWSEGMENT event.
API: gst_event_new_new_segment_full()
API: gst_event_parse_new_segment_full()
* gst/gstsegment.c: (gst_segment_init), (gst_segment_set_seek),
(gst_segment_set_newsegment), (gst_segment_set_newsegment_full),
(gst_segment_to_stream_time), (gst_segment_to_running_time):
* gst/gstsegment.h:
Add applied_rate to GstSegment structure.
Make calculation of stream_time and running_time more correct
wrt rate/applied_rate.
Add some more docs.
API: GstSegment::applied_rate field
API: gst_segment_set_newsegment_full();
* libs/gst/base/gstbasesink.c: (gst_base_sink_configure_segment),
(gst_base_sink_get_sync_times), (gst_base_sink_get_position):
* libs/gst/base/gstbasetransform.c:
(gst_base_transform_sink_eventfunc),
(gst_base_transform_handle_buffer):
Parse and use applied_rate in the GstSegment field.
* tests/check/gst/gstevent.c: (GST_START_TEST):
Add check for applied_rate field.
* tests/check/gst/gstsegment.c: (GST_START_TEST),
(gstsegments_suite):
Add more checks for various GstSegment operations.
Original commit message from CVS:
* libs/gst/base/gstbasesink.c: (gst_base_sink_get_sync_times),
(gst_base_sink_do_sync), (gst_base_sink_chain_unlocked),
(gst_base_sink_get_position), (gst_base_sink_change_state):
Store the sync time of the buffer end position separatly in a
new variable eos_rtime so we can properly sync the EOS event.
Fixes#340697.
Fix the docs for gst_base_sink_set_qos_enabled().
Don't set segment start to invalid value when we receive a
non TIME newsegment.
get closer to handling position reporting for negative rates
correctly.
Original commit message from CVS:
* libs/gst/base/gstbasesink.c: (gst_base_sink_event):
* libs/gst/base/gstcollectpads.c: (gst_collect_pads_base_init),
(gst_collect_pads_is_collected), (gst_collect_pads_event):
more detailed debug and formatting cleanup
Original commit message from CVS:
* libs/gst/base/gstbasesink.c: (gst_base_sink_is_too_late),
(gst_base_sink_do_render_stats), (gst_base_sink_render_object),
(gst_base_sink_get_position), (gst_base_sink_change_state):
When frame dropping is enabled, we should not ignore frames
without a duration.
Update some documentation.
Original commit message from CVS:
* gst/gstsegment.c:
Added some more docs.
* libs/gst/base/gstbasesink.c: (gst_base_sink_perform_qos),
(gst_base_sink_reset_qos):
Calculate more accurate rate values.
Original commit message from CVS:
* gst/gst_private.h:
add a new #ifdef to use __declspec(dllimport) only for
other modules and not for gstreamer core
* gst/gstbasesink.c: (gst_base_sink_perform_qos):
use gst_guint64_to_gdouble for conversion
* win32/common/libgstreamer.def:
add new exported functions
* win32/vs6/gst_inspect.dsp:
* win32/vs6/gst_launch.dsp:
* win32/vs6/libgstbase.dsp:
* win32/vs6/libgstcontroller.dsp:
* win32/vs6/libgstcoreelements.dsp:
* win32/vs6/libgstdataprotocol.dsp:
* win32/vs6/libgstnet.dsp:
update project files
Original commit message from CVS:
* libs/gst/base/gstbasesink.c: (gst_base_sink_get_sync_times):
Be a bit nicer to badly behaving upstream elements that expect
us to deal with non TIME segments and timestamps (such as fakesrc
in the testsuite).
Original commit message from CVS:
* gst/gstbus.c:
Small documentation clarification about the signal watch.
* libs/gst/base/gstbasesink.c: (gst_base_sink_get_sync_times),
(gst_base_sink_wait_clock), (gst_base_sink_do_sync),
(gst_base_sink_perform_qos), (gst_base_sink_reset_qos),
(gst_base_sink_do_render_stats), (gst_base_sink_render_object),
(gst_base_sink_get_position_last),
(gst_base_sink_get_position_paused), (gst_base_sink_change_state):
Convert and store timestamps in stream time and running time, the
raw timestamps are not usefull, also document this better.
Use different window sizes for good and bad QoS observations so
we react to badness a little quicker.
Keep track of the amount of rendered and dropped buffers.
Send QoS timestamps in running time.
* libs/gst/base/gstbasetransform.c:
(gst_base_transform_sink_eventfunc),
(gst_base_transform_handle_buffer):
Compare QoS timestamps against running time.
Original commit message from CVS:
* docs/libs/gstreamer-libs-sections.txt:
* gst/gsttaglist.c:
* libs/gst/base/gstbasesink.c:
* libs/gst/base/gstbasesink.h:
* libs/gst/base/gstbasesrc.c:
* libs/gst/base/gstbasesrc.h:
Documentation updates. Make BaseSink and BaseSrc docs contain the
class structure so that people can actually see the prototypes for
virtual functions they're supposed to be overriding.
Original commit message from CVS:
* libs/gst/base/gstbasesink.c: (gst_base_sink_init),
(gst_base_sink_finalize), (gst_base_sink_set_qos_enabled),
(gst_base_sink_is_qos_enabled), (gst_base_sink_do_sync),
(gst_base_sink_record_qos_observation),
(gst_base_sink_perform_qos), (gst_base_sink_reset_qos),
(gst_base_sink_is_too_late), (gst_base_sink_render_object),
(gst_base_sink_change_state):
More QoS measurements as described in the design doc.
Get rid of ringbuffer with observations, running average is
more simple and equally good.
Calculates valid proportion now.
Added beginning of flood measurement.
Original commit message from CVS:
* libs/gst/base/gstbasesink.c: (gst_base_sink_class_init),
(gst_base_sink_init), (gst_base_sink_finalize),
(gst_base_sink_set_qos_enabled), (gst_base_sink_is_qos_enabled),
(gst_base_sink_set_property), (gst_base_sink_get_property),
(gst_base_sink_commit_state), (gst_base_sink_get_sync_times),
(gst_base_sink_wait_clock), (gst_base_sink_do_sync),
(gst_base_sink_add_qos_observation), (gst_base_sink_send_qos),
(gst_base_sink_perform_qos), (gst_base_sink_reset_qos),
(gst_base_sink_is_too_late), (gst_base_sink_render_object),
(gst_base_sink_preroll_object), (gst_base_sink_event),
(gst_base_sink_chain_unlocked), (gst_base_sink_get_position_last),
(gst_base_sink_get_position_paused), (gst_base_sink_get_position),
(gst_base_sink_query), (gst_base_sink_change_state):
Decouple max-lateness and the fact that QoS messages are generated
with a new property (qos).
Add vmethod so subclasses can be notified of ASYNC playing
state changes.
Collect timestamp start and stop to report better current
position in EOS/PLAYING/PAUSED/READY/NULL.
Refactor QoS/frame dropping and other measurements.
API: GstBaseSrc::qos
* libs/gst/base/gstbasesink.h:
Added Private struct.
API: gst_base_sink_set_qos_enabled
API: gst_base_sink_is_qos_enabled
Original commit message from CVS:
* libs/gst/base/gstbasesink.c: (gst_base_sink_configure_segment),
(gst_base_sink_get_sync_times), (gst_base_sink_wait_clock),
(gst_base_sink_do_sync), (gst_base_sink_do_qos):
Better debug info when we receive a segment event.
Reorganize a bit so we can pass the get_times() results around.
Use the segment format when calculating the running time.
Don't do QoS is sync is disabled or we have no clock or the
element does not want us to sync to the clock.
Don't drop buffers if QoS is disabled for now.
Original commit message from CVS:
* libs/gst/base/gstbasesink.c: (gst_base_sink_get_sync_times):
Can't use gst_segment_to_running_time() when the segment
is not in GST_TIME_FORMAT (like with filesink, for example).
Stops flac encoding pipelines from spewing critical warnings
at EOS (#331248).
Original commit message from CVS:
* libs/gst/base/gstbasesink.c: (gst_base_sink_do_sync),
(gst_base_sink_do_qos):
Separate QoS calculation.
Only drop buffers when lateness is bigger than the
duration of the buffer.
Original commit message from CVS:
* libs/gst/base/gstbasesink.c: (gst_base_sink_init),
(gst_base_sink_do_sync):
Small cleanups.
Use QOS debug category.
Original commit message from CVS:
* gst/gstclock.c: (gst_clock_entry_new),
(gst_clock_id_compare_func), (gst_clock_id_wait),
(gst_clock_id_wait_async), (gst_clock_id_unschedule),
(gst_clock_init), (gst_clock_get_internal_time),
(gst_clock_set_master), (do_linear_regression),
(gst_clock_add_observation), (gst_clock_set_property):
* gst/gstclock.h:
Review docs.
Small cleanups.
Fix a possible segfault when the window-size is made smaller.
Calculate jitter before performing the clock wait. Ideally
the clock implementation should calculate jitter but we need
API breakage for that.
* gst/gstsystemclock.c: (gst_system_clock_init):
Docs review.
* libs/gst/base/gstbasesink.c: (gst_base_sink_do_sync):
Remove leftover else
* tests/check/gst/gstsystemclock.c: (GST_START_TEST),
(gst_systemclock_suite):
Added check to test GST_CLOCK_DIFF.
Original commit message from CVS:
* docs/libs/gstreamer-libs-sections.txt:
Added basesink new methods.
* gst/gstevent.c:
* gst/gstevent.h:
Docs updates. Flesh out the QoS docs.
* libs/gst/base/gstadapter.c:
Small doc clarification about ownership and flushing.
* libs/gst/base/gstbasesink.c: (gst_base_sink_set_sync),
(gst_base_sink_get_sync), (gst_base_sink_set_max_lateness),
(gst_base_sink_get_max_lateness), (gst_base_sink_set_property),
(gst_base_sink_get_property), (gst_base_sink_do_sync):
* libs/gst/base/gstbasesink.h:
Added new methods to allow subclass to control max-lateness
and sync.
Generate very basic QoS events based on last sync observation.
Updated docs, fix typo, added some QoS blurb.
* libs/gst/base/gstbasesrc.c:
Remove obsolete _get_state() calls from docs.
Original commit message from CVS:
* libs/gst/base/gstbasesink.c: (gst_base_sink_class_init),
(gst_base_sink_init), (gst_base_sink_set_property),
(gst_base_sink_get_property), (gst_base_sink_do_sync):
* libs/gst/base/gstbasesink.h:
Make max-lateness a property.
Original commit message from CVS:
* libs/gst/base/gstbasesink.c: (gst_base_sink_wait_clock),
(gst_base_sink_do_sync), (gst_base_sink_render_object):
Don't ever draw a frame that is >10ms late.
Original commit message from CVS:
2006-02-07 Andy Wingo <wingo@pobox.com>
* libs/gst/base/gstcollectpads.c (gst_collect_pads_chain): Only
update last_stop if we're in TIME format and the timestamp is
valid.
* libs/gst/base/gstcollectpads.c (gst_collect_pads_event)
* libs/gst/base/gstbasetransform.c (gst_base_transform_eventfunc):
* libs/gst/base/gstbasesink.c (gst_base_sink_configure_segment):
If we get a new newsegment with a different format, adapt
accordingly.
* gst/gstclock.c (gst_clock_set_calibration): Accept a numerator
of 0. Not a problem, really.
Original commit message from CVS:
2006-02-07 Andy Wingo <wingo@pobox.com>
* libs/gst/base/gstbasesink.c (gst_base_sink_chain_unlocked): Only
warn if sync=true.
Original commit message from CVS:
* libs/gst/base/gstbasesink.c: (gst_base_sink_do_sync),
(gst_base_sink_render_object),
(gst_base_sink_queue_object_unlocked):
Only keep track of prerollable items when we are
prerolling.
Before rendering after preroll, always check if we
have queued items.
Added some more debugging.
Original commit message from CVS:
* docs/design/part-element-sink.txt:
Updated document.
* libs/gst/base/gstbasesink.c: (gst_base_sink_init),
(gst_base_sink_finalize), (gst_base_sink_preroll_queue_flush),
(gst_base_sink_configure_segment), (gst_base_sink_commit_state),
(gst_base_sink_get_sync_times), (gst_base_sink_wait_clock),
(gst_base_sink_do_sync), (gst_base_sink_render_object),
(gst_base_sink_preroll_object),
(gst_base_sink_queue_object_unlocked),
(gst_base_sink_queue_object), (gst_base_sink_event),
(gst_base_sink_chain_unlocked), (gst_base_sink_chain),
(gst_base_sink_loop), (gst_base_sink_activate_pull),
(gst_base_sink_get_position), (gst_base_sink_change_state):
* libs/gst/base/gstbasesink.h:
Totally refactored matching the design doc.
Use two segments, one to clip incomming buffers and another to
perform sync.
Handle queueing correctly, bypass the queue when playing.
Make EOS cancelable.
Handle errors correctly when operating in pull based mode.
* tests/check/elements/fakesink.c: (GST_START_TEST),
(fakesink_suite):
Added new check for sinks.
Original commit message from CVS:
* libs/gst/base/gstbasesink.c: (gst_base_sink_init),
(gst_base_sink_preroll_queue_empty), (gst_base_sink_commit_state),
(gst_base_sink_handle_object), (gst_base_sink_event),
(gst_base_sink_is_prerolled), (gst_base_sink_wait),
(gst_base_sink_do_sync), (gst_base_sink_handle_event),
(gst_base_sink_handle_buffer), (gst_base_sink_set_flushing),
(gst_base_sink_deactivate), (gst_base_sink_activate),
(gst_base_sink_activate_pull), (gst_base_sink_get_position),
(gst_base_sink_query), (gst_base_sink_change_state):
Basesink cleanups, remove some old code.
Handle the case where a subclass can preroll in the render
method (mostly audiosinks).
Handle more events.
Remove some locks around variables that are now protected
with the PREROLL_LOCK (clock_id, flushing, ..).
Optimize position query some more, do correct locking.
Remove old code to push queue in state change, this is not
needed anymore since preroll blocks on all prerollable items
now.
Almost implemented as described in design doc.
Original commit message from CVS:
2006-01-25 Julien MOUTTE <julien@moutte.net>
* libs/gst/base/gstbasesink.c: (gst_base_sink_get_position):
We only query position using the clock in the playing state.
Query peer in the other cases.
* win32/common/config.h: Updates.
