Commit graph

129 commits

Author SHA1 Message Date
Jan Schmidt
41d2b6f19e splitmux: Handle a hang draining out at EOS
Make sure that all data is drained out when the reference pad
goes EOS. Fixes a problem where data that arrives on other
pads after the reference pad finishes can stall forever and
never pass EOS.

https://bugzilla.gnome.org/show_bug.cgi?id=763711
2016-04-01 00:48:05 +11:00
Xavier Claessens
fb835c100a splitmuxsink: Fix occasional deadlock when ending file with subtitle
Deadlock occurs when splitting files if one stream received no buffer during
the first GOP of the next file. That can happen in that scenario for example:
 1) The first GOP of video is collected, it has a duration of 10s.
    max_in_running_time is set to 10s.
 2) Other streams catchup and we receive the first subtitle buffer at ts=0 and
    has a duration of 1min.
 3) We receive the 2nd subtitle buffer with a ts=1min. in_running_time is set to
    1min. That buffer is blocked in handle_mq_input() because
    max_in_running_time is still 10s.
 4) Since all in_running_time are now > 10s, max_out_running_time is now set to
    10s. That first GOP gets recorded into the file. The muxer pop buffers out
    of the mq, when it tries to pop a 2nd subtitle buffer it blocks because the
    GstDataQueue is empty.
 5) A 2nd GOP of video is collected and has a duration of 10s as well.
    max_in_running_time is now 20s. Since subtitle's in_running_time is already
    1min, that GOP is already complete.
 6) But let's say we overran the max file size, we thus set state to
    SPLITMUX_STATE_ENDING_FILE now. As soon as a buffer with ts > 10s (end of
    previous GOP) arrives in handle_mq_output(), EOS event is sent downstream
    instead. But since the subtitle queue is empty, that's never going to
    happen. Pipeline is now deadlocked.

To fix this situation we have to:
 - Send a dummy event through the queue to wakeup output thread.
 - Update out_running_time to at least max_out_running_time so it sends EOS.
 - Respect time order, so we set out_running_tim=max_in_running_time because
   that's bigger than previous buffer and smaller than next.

https://bugzilla.gnome.org/show_bug.cgi?id=763711
2016-04-01 00:48:05 +11:00
Thiago Santos
d738fa0787 splitmuxsink: only try to create internal sink if it doesn't exist
This allows splitmuxsink to be reused after being put to NULL.

Test included

https://bugzilla.gnome.org/show_bug.cgi?id=762893
2016-03-24 20:10:25 -03:00
Vineeth TM
1071309870 good: use new gst_element_class_add_static_pad_template()
https://bugzilla.gnome.org/show_bug.cgi?id=763076
2016-03-24 14:32:20 +02:00
Jan Schmidt
675a4088e5 splitmuxsink: Only update running time when it increases.
Don't increment running time from every buffer. The correct
logic to only increment when running time advances is a
little further down, so delete this left-over line.
2015-12-19 03:49:28 +11:00
George Kiagiadakis
4a78048cc5 splitmuxsink: do not destroy the multiqueue & muxer when going to NULL
Instead, delay it until all request pads have been released. This is
because the release_pad() vfunc requires the multiqueue and muxer to
be there in order to release their request pads as well. If those
elements are destroyed earlier, release_pad() does not work, no
pads are released and some resources are leaked.

https://bugzilla.gnome.org/show_bug.cgi?id=753622
2015-10-28 22:39:44 +11:00
George Kiagiadakis
5e4caca709 splitmuxsink: post messages when fragments are being opened and closed
This can be useful for applications that need to track the created fragments
(to log them in a recording database, for example)

https://bugzilla.gnome.org/show_bug.cgi?id=750108
2015-10-03 00:52:23 +10:00
Ramiro Polla
f0a47d0b60 splitmuxsink: allow non-video streams to serve as reference
In the absence of a video stream, the first stream will be used as
reference.

https://bugzilla.gnome.org/show_bug.cgi?id=753617
2015-10-03 00:44:58 +10:00
George Kiagiadakis
20754db26f splitmuxsink: initialize mux_start_time properly
mux_start_time refers to the running_time of the buffer
that goes first in the output file. Normally this time is
0, so this variable is initialized to 0 during the state
change to PAUSED.

However, when dealing with dynamic pipelines and starting
a recording while the pipeline has already run for a while,
the running_time of the first buffer is > 0 and this causes
a problem with detecting the end of the first file(s) when
splitting by duration, because the code will later compare
the threshold_time with (last buffer running_time - mux_start_time)
and will get it wrong until mux_start_time advances enough
to make this difference < threshold_time, creating empty files
in the meantime.

https://bugzilla.gnome.org/show_bug.cgi?id=753624
2015-10-03 00:44:01 +10:00
Thiago Santos
72212198c7 splitmuxsink: remove extra \n at debug message 2015-08-03 13:46:16 -03:00
Thiago Santos
a930a1364a splitmuxsink: prevent deadlock when states change too fast
If the GOP is completed, pads have to start gathering for the
next one but it is possible that the the state might go to
COLLECTING_GOP_START and back to WAITING_GOP_COMPLETE before the
thread has a chance to wake up and proceed, leaving it trapped in
the check_completed_gop loop and deadlocking the other threads
waiting for it to advance.

To solve it, this patch also checks that tha input running time
hasn't changed to prevent this scenario.
2015-08-03 13:46:16 -03:00
Jan Schmidt
a0182dd943 splitmuxsink: Support mpegtsmux as a muxer.
As a fallback, look for a pad template sink_%d on
the muxer when requesting pads, to support mpegtsmux

https://bugzilla.gnome.org/show_bug.cgi?id=752999
2015-07-29 23:03:30 +10:00
George Kiagiadakis
bbfa46363c splitmuxsink: add a "format-location" signal that allows better control over filenames
In certain applications, splitting into files named after a base
location template and an incremental sequence number is not enough.

This signal gives more fine-grained control to the application to
decide how to name the files.

https://bugzilla.gnome.org/show_bug.cgi?id=750106
2015-07-14 18:45:49 +02:00
Jan Schmidt
b26bbae695 Revert "splitmuxsink: Mask async-start/done while switching files."
This reverts commit d61e5393f1.

Causes failures muxing larger GOP sizes for some reason. Reverting
while I figure it out
2015-06-23 17:33:03 +10:00
Jan Schmidt
d61e5393f1 splitmuxsink: Mask async-start/done while switching files.
Sometimes, extra async-start/done from the internal sink
while the element is still starting up can cause splitmuxsink
to stall in PAUSED state when it has been set to PLAYING
by the app. Drop the child's async-start/done messages while
switching, so they don't cause state changes at the
splitmuxsink level.

https://bugzilla.gnome.org/show_bug.cgi?id=750747
2015-06-23 12:03:14 +10:00
Jan Schmidt
3f39d06338 splitmuxsink: Take released-but-not-yet-output bytes into account
When deciding whether it's time to switch to a new file, take into
account data that's been released for pushing, but hasn't yet
been pushed - because downstream is slow or the threads haven't been
scheduled.

Fixes a race in the unit test and probably in practice - sometimes
failing to switch when it should for an extra GOP or two.

Also fix a problem in splitmuxsrc where playback sometimes
stalls at startup if types are found too quickly.

https://bugzilla.gnome.org/show_bug.cgi?id=750747
2015-06-12 01:57:36 +10:00
Jan Schmidt
d78502deb1 splitmuxsink: Don't accumulate more than 2 GOPs
Don't allow large amounts of data to queue up - we only need
the GOP we're writing, and the GOP we're accumulating.
2015-06-08 18:58:43 +10:00
George Kiagiadakis
97c03449a4 splitmuxsink: do not access property variable without the object lock, use the local stack copy instead 2015-04-15 13:30:19 +02:00
George Kiagiadakis
1954726328 splitmuxsink: add probe on the multiqueue's sink pad instead of the ghost pad
because _release_pad tries to release it from ctx->sinkpad, which is
multiqueue's sink pad, and currently fails because the probe is not
installed there
2015-04-15 13:30:19 +02:00
Vincent Penquerc'h
a862db33b6 splitmuxsink: fix mutex leak 2015-04-09 13:01:23 +01:00
Jan Schmidt
b0ce43cde3 splitmuxsink: Protect property variables with the object lock.
Use the object lock instead of the splitmux lock to protect
internal property variables, so they're not locked when
switching to a new file.

https://bugzilla.gnome.org/show_bug.cgi?id=744420
2015-03-07 00:55:47 +11:00
Tim-Philipp Müller
f5b511b42b multifile: attempt to fix docs build issue on build bot 2015-02-26 19:48:33 +00:00
Luis de Bethencourt
4af5a2b760 splitmuxsin: remove dead code
Every instance of goto beach has buf_info equal NULL. Don't check
for a condition that never happens.

CID #1268399
2015-02-16 13:59:17 +00:00
Tim-Philipp Müller
3f5b690e78 splitmuxsink: flag as sink from the start 2015-02-13 20:40:48 +00:00
Tim-Philipp Müller
90badeebad splitmuxsink: fix example pipeline properly
x264enc might not have a max-key-int property, but it
has a key-int-max property...
2015-02-10 16:00:07 +00:00
Luis de Bethencourt
12aa2428e0 splitmux: update example pipeline
Element x264enc doesn't have a max-key-int property
2015-02-10 14:56:23 +00:00
Luis de Bethencourt
0373fd8f65 splitmux: fix memory leak
If execution goes to the beach in line 981, buf_info goes out of scope without
the memory being free'd. Handle this case.

CID #1268403
2015-02-10 13:33:09 +00:00
Jan Schmidt
8ceb58122e splitmux: Add unit test for file splitting
Add a unit test for file splitting, and fix the leaks in the
splitmuxsink it found
2015-02-07 03:58:30 +11:00
Jan Schmidt
5e2214d309 splitmux: Implement new elements for splitting files at mux level.
Implement 2 new elements - splitmuxsink and splitmuxsrc.

splitmuxsink is a bin which wraps a muxer and takes 1 video stream,
plus audio/subtitle streams, and starts a new file
whenever necessary to avoid overrunning a threshold of either bytes
or time. New files are started at a keyframe, and corresponding audio
and subtitle streams are split at packet boundaries to match
video GOP timestamps.

splitmuxsrc is a corresponding source element which handles
the splitmux:// URL and plays back all component files,
reconstructing the original elementary streams as it goes.
2015-02-06 04:26:59 +11:00