Commit graph

21 commits

Author SHA1 Message Date
Mathieu Duponchelle
cddb0e951f splitmuxsrc: don't queue data on unlinked pads
Once a pad has returned NOT_LINKED, the part reader shouldn't let its
corresponding data queue run full and eventually (after 20 seconds)
stall playback.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/3145>
2022-10-10 18:11:12 +00:00
Mathieu Duponchelle
f8d8d67b8b splitmuxsrc: don't consider unlinked pads when deactivating part
If splitmuxsrc exposes multiple pads, but only one is linked, part pads
will never see an EOS event. This shouldn't prevent the part from being
eventually deactivated.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/3099>
2022-10-01 02:33:08 +00:00
Thibault Saunier
6a4425e46a meson: Call pkgconfig.generate in the loop where we declare plugins dependencies
Removing some copy pasted code

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2970>
2022-09-01 21:17:35 +00:00
Jan Schmidt
4a6c2e6720 splitmuxsrc: Stop pad task before cleanup
When stopping the element, make sure the pad task
is stopped before destroying the part readers.

Closes a race where the pad task might access
a freed pointer.

Also add a guard against this sort of thing
by holding a ref to the reader in the pad loop.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2901>
2022-08-17 09:42:50 +10:00
Jan Schmidt
ab459f0528 splitmuxsink: Fix memory leak
Fix a leak of the buffer info struct when reaching
EOS without data on the reference input.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2751>
2022-07-12 11:22:33 +00:00
Seungha Yang
b233df3537 splitmuxsink: Don't crash on EOS without buffer
Fix a case where upstream pushed EOS without buffers.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2174>
2022-07-05 11:33:35 +00:00
Jan Schmidt
7322a6d004 splitmuxsrc: Re-queue sticky events after probing.
When processing the first event after probing the
file and being activated, requeue sticky events
as there's no requirement that demuxers send tag
and other events again after a seek - that's
why they're sticky.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2432>
2022-05-17 11:55:40 +00:00
Jan Alexander Steffens (heftig)
718d31fe63 splitmuxsink: When flushing, exit handle_mq_input quickly
If we just break the loop, we might run into the `gop != NULL` assert
that follows it. Rather, exit immediately with flushing flow.

Also use this flushing mechanism when we release a pad. This avoids
having an extra flag.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1030>
2022-05-17 09:24:10 +00:00
Jan Alexander Steffens (heftig)
fd27ee1537 splitmuxsink: Avoid deadlock on release, harder
Unlock after broadcasting and wait for the pad to be free before
relocking the muxer, giving the input probe a chance to react to our
broadcast.

Improves the fix from
https://gitlab.freedesktop.org/gstreamer/gst-plugins-good/-/merge_requests/838.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1030>
2022-05-17 09:24:10 +00:00
Xavier Claessens
b99ecc78ca Replace gst-i18n-*.h with gi18n-lib.h
GLib guarantees libintl is always present, using proxy-libintl as
last resort. There is no need to mock gettex API any more.

This fix static build on Windows because G_INTL_STATIC_COMPILATION must
be defined before including libintl.h, and glib does it for us as part
as including glib.h.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2028>
2022-04-19 18:01:06 +00:00
Sebastian Dröge
241a26631d splitmuxsink: Warn when calculating the next fragment time in timecode mode fails
But only if timecode mode is enabled as it will fail all the time
otherwise.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1557>
2022-01-25 12:08:20 +00:00
Célestin Marot
f509578de5 multifilesrc: fix caps leak
since `gst_caps_replace()` and `gst_pad_set_caps()` both ref the caps and neither of them takes the ownership of the caps -> it must be unreffed in `gst_multi_file_src_set_property()`

to test the leak (on Unix): `echo coucou > /tmp/file.txt && GST_TRACERS=leaks GST_DEBUG="GST_TRACER:7" gst-launch-1.0 multifilesrc location=/tmp/file.txt caps='txt' ! fakesink`

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1436>
2021-12-10 18:34:57 +01:00
Sebastian Dröge
dac82a8932 multifilesink: Make minimum distance between keyframes in next-file=key-frame mode configurable
Previously this was hardcoded to 10s, which is not necessarily the
desired behaviour.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1292>
2021-11-02 20:21:10 +00:00
Sebastian Dröge
2d4491e195 splitmuxsink: Don't assert on the input side if no GOP is available when shutting down
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1119>
2021-10-12 19:34:16 +00:00
Sebastian Dröge
4316be6571 splitmuxsink: Keep track of the pending input GOPs in a queue
This cleans up input GOP handling and makes it possible to handle more
than 2 pending GOPs, which could happen before if keyframes are arriving
with too short of a distance between them.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1005>
2021-10-05 12:35:19 +00:00
Sebastian Dröge
ae8ceb801c splitmuxsink: Handle frame reordering due to B frames better
Instead of assuming that the PTS of a keyframe is the lowest PTS of a
GOP, wait until the DTS has passed this PTS and take the minimum PTS up
to that point. That way the minimum PTS of a GOP can be determined, at
least for closed GOP streams. Open GOP streams still can't be handled
properly.

By knowing the minimum PTS of each GOP, keyframes can be requested at
the correct time relative to the GOP (and thus fragment) start and
fragment overflow calculations can calculate the correct durations of
the GOPs.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1005>
2021-10-05 12:35:19 +00:00
Sebastian Dröge
f83ed50c20 splitmuxsink: Reset timecodes when resetting the sink
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1005>
2021-10-05 12:35:19 +00:00
Sebastian Dröge
ed6823ad67 splitmuxsink: Use GST_CLOCK_STIME_NONE instead of 0 to initialize max out running time
Otherwise streams with only DTS would misbehave while it is negative.

For outputting everything immediately at EOS, use G_MAXINT64 instead
which is bigger-or-equal to any other running time.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1005>
2021-10-05 12:35:19 +00:00
Sebastian Dröge
eed843eea0 splitmuxsink: Only update max in/out running time if it's actually bigger
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1005>
2021-10-05 12:35:19 +00:00
Sebastian Dröge
68683d15c4 splitmuxsink: Only count keyframes for the reference context, consistently
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1005>
2021-10-05 12:35:19 +00:00
Thibault Saunier
5ff769d731 Move files from gst-plugins-good into the "subprojects/gst-plugins-good/" subdir 2021-09-24 16:13:50 -03:00