According to [1] EINTR is a possible errno for fsync() and it happens in
reality on linux (video writing via splitmuxsink with robust muxing enabled
on a cifs mounted network share), so handle it as all other EINTR
(do/while(errno == EINTR)).
Fixes:
GError.message: Error while writing to file "vidoe_001.mp4". GError.domain: 2372 GError.code: 10 from: FileSink debug: gstfilesink.c(849): gst_file_sink_render (): /GstPipeline:Pipeline/GstSplitMuxSink:SplitMuxSink/GstBin:QueueBin/GstFileSink:FileSink: Interrupted system call
Signed-off-by: Peter Seiderer <ps.report@gmx.net>
`g_object_notify()` actually takes a global lock to look up the
`GParamSpec` that corresponds to the given property name. It's not a
huge performance hit, but it's easily avoidable by using the
`_by_pspec()` variant.
This reverts a96002bb28, which is not
necessary anymore. If we release the pad after removing it then none of
the deactivation code will actually be called because the pad has no
parent anymore, and we require a parent on the pad for deactivation to
happen.
This can then, among other things, cause a streaming thread to be still
stuck in a pad probe because the pad was never flushed, and waiting
there forever because now the pad will actually never be flushed anymore.
If a pad is currently being released we don't want to forward the
FLUSHING flow return but instead consider it as NOT_LINKED. FLUSHING
would also cause upstream to be FLUSHING.
This part was missed in a3c4a3201a and
resulted in a different (and wrong) workaround in
a96002bb28.
Otherwise we're not guaranteed to read the very latest value that
another thread might've written in there when the pad was released, and
could instead work with an old value.
By passing NULL to `g_signal_new` instead of a marshaller, GLib will
actually internally optimize the signal (if the marshaller is available
in GLib itself) by also setting the valist marshaller. This makes the
signal emission a bit more performant than the regular marshalling,
which still needs to box into `GValue` and call libffi in case of a
generic marshaller.
Note that for custom marshallers, one would use
`g_signal_set_va_marshaller()` with the valist marshaller instead.
If the element before the sink needs $n buffers to produce one output
buffer, we were reffing $n events and unreffing only one.
Prevent this by using g_object_set_qdata_full() to handle the event
unreffing so we're sure no ref will be lost.
The records are static and so appear as false positives when using those
tracers with the leaks tracer as well.
The leaks tracer was already setting this flag on its record so let's
set it on the other ones as well.
In gst_download_buffer_wait_for_data(), when a seek is made with
perform_seek_to_offset() the `qlock` is released temporarily. Therefore,
the flushing condition can be set during this period and should be
checked.
This was not being checked before, causing occasional deadlocks when
GST_DOWNLOAD_BUFFER_WAIT_ADD_CHECK() was called.
GST_DOWNLOAD_BUFFER_WAIT_ADD_CHECK() assumes that the caller has already
checked that we're not flushing before, since this is done when
acquiring the lock; so if we release it temporarily somewhere, we need
to check for flush again.
Without that check, the function would keep waiting for the condition
variable to be notified before checking for flushing condition again,
and that may very well never happen. This was reproduced when during pad
deactivation when running WebKit in gdb.
sync=TRUE implementation changes the latency query of a non-live
upstream into live, though it wrongly set the upstream max latency to 0.
As non-live sources won't loose data if we wait longer, this should have
been reported as have no max latency limite (-1).
In the hotdoc inspector for example, pads are instantiated with
g_object_new, other code paths to get/set properties already make
that check.
And update doc cache
This feature was previously available only through the SIGUSR2 signal,
which meant it wasn't available on platforms that don't have UNIX
signals, such as Windows and with applications that already use
SIGUSR1 for something else.
Now we have action-signals for doing the same. These action signals
can also be used for fetching the checkpoint information
programmatically instead of printing to the debug log.
This allows programs to inspect the leaked objects directly, log them,
and so on. Unlike the existing mechanism to use SIGUSR1, this also
works on platforms that do not support UNIX signals, such as Windows
and with applications that already use SIGUSR1 for something else.
This will be useful in combination with the next commit when we add
API to get a list of active tracers so that consumers of the API can
easily distinguish tracer objects.
The code implicitly uses this value when the stack trace is not FULL.
Mostly useful for documenting the behaviour when each flag is passed
and for translating to/from strings.
There was a race where we could still get the pad event function
called when its private member were already unset, leading to
a segfault in the event handler:
```
0 gst_multi_queue_src_event (pad=<optimized out>, parent=<optimized out>, event=0x7f3ff0007600) at ../subprojects/gstreamer/plugins/elements/gstmultiqueue.c:2534
2534 ret = gst_pad_push_event (sq->sinkpad, event);
[Current thread is 1 (Thread 0x7f406c0258c0 (LWP 21925))]
(gdb) bt
0 0x00007f4062ec1399 in gst_multi_queue_src_event (pad=<optimized out>, parent=<optimized out>, event=0x7f3ff0007600 [GstEvent]) at ../subprojects/gstreamer/plugins/elements/gstmultiqueue.c:2534
1 0x00007f406b40f46d in gst_validate_pad_monitor_src_event_check (handler=0x7f4062ec1360 <gst_multi_queue_src_event>, event=0x7f3ff0007600 [GstEvent], parent=0x7f3fcc01f090 [GstMultiQueue|multiqueue167], pad_monitor=0x7f3fe809e7c0 [GstValidatePadMonitor|validatepadmonitor2213]) at ../subprojects/gst-devtools/validate/gst/validate/gst-validate-pad-monitor.c:2101
2 0x00007f406b40f46d in gst_validate_pad_monitor_src_event_func (pad=<optimized out>, parent=0x7f3fcc01f090 [GstMultiQueue|multiqueue167], event=0x7f3ff0007600 [GstEvent]) at ../subprojects/gst-devtools/validate/gst/validate/gst-validate-pad-monitor.c:2374
3 0x00007f406b904387 in gst_pad_send_event_unchecked (pad=pad@entry=0x7f3fdc027650 [GstPad|src_0], event=event@entry=0x7f3ff0007600 [GstEvent], type=<optimized out>, type@entry=GST_PAD_PROBE_TYPE_EVENT_UPSTREAM) at ../subprojects/gstreamer/gst/gstpad.c:5772
4 0x00007f406b90481b in gst_pad_push_event_unchecked (pad=pad@entry=0x7f4058182fc0 [GstPad|sink], event=event@entry=0x7f3ff0007600 [GstEvent], type=type@entry=GST_PAD_PROBE_TYPE_EVENT_UPSTREAM) at ../subprojects/gstreamer/gst/gstpad.c:5417
5 0x00007f406b90f016 in gst_pad_push_event (pad=0x7f4058182fc0 [GstPad|sink], event=event@entry=0x7f3ff0007600 [GstEvent]) at ../subprojects/gstreamer/gst/gstpad.c:5554
6 0x00007f406a1c99ba in gst_video_decoder_src_event_default (decoder=0x7f3fe81c6060 [GstTheoraDec|theoradec46], event=<optimized out>) at ../subprojects/gst-plugins-base/gst-libs/gst/video/gstvideodecoder.c:1532
7 0x00007f406b40f46d in gst_validate_pad_monitor_src_event_check (handler=0x7f406a1ca270 <gst_video_decoder_src_event>, event=0x7f3ff0007600 [GstEvent], parent=0x7f3fe81c6060 [GstTheoraDec|theoradec46], pad_monitor=0x7f4028163aa0 [GstValidatePadMonitor|validatepadmonitor2216]) at ../subprojects/gst-devtools/validate/gst/validate/gst-validate-pad-monitor.c:2101
8 0x00007f406b40f46d in gst_validate_pad_monitor_src_event_func (pad=<optimized out>, parent=0x7f3fe81c6060 [GstTheoraDec|theoradec46], event=0x7f3ff0007600 [GstEvent]) at ../subprojects/gst-devtools/validate/gst/validate/gst-validate-pad-monitor.c:2374
```
This make the GstSingleQueue a MiniObject, mainly so it is properly
refcounted.
This also make use of the GstMultiQueuePad class for srcpads which
is totally valid as srcpads and sinkpads share the same SingleQueue
object.
The `query` argument of gst_pad_query is "transfer none".
Query objects are "borrowed" by the pad query handlers and those
should never unref them.
This was leading to double freed queries in a very racy way with nested
GESTimelines.
Otherwise when seeking backwards we would keep the last_stop at the last
position we saw until playback passed the seek position again, and if
switching to the next pad happens in the meantime we would set the wrong
offset in the outgoing segment.
This URI is valid:
data:,;base64
(It encodes the literal string ";base64")
But would lead to a crash because the code assumed the semicolon would
be placed before the colon.
Quoting RFC 2396:
For resiliency, programs interpreting URI should treat upper case
letters as equivalent to lower case in scheme names (e.g., allow
"HTTP" as well as "http").
This seems to happen when another client is accessing the file at the
same time, and retrying after a short amount of time solves it.
Sometimes partial data is written at that point already but we have no
idea how much it is, or if what was written is correct (it sometimes
isn't) so we always first seek back to the current position and repeat
the whole failed write.
It happens at least on Linux and macOS on SMB/CIFS and NFS file systems.
Between write attempts that failed with EACCES we wait 10ms, and after
enough consecutive tries that failed with EACCES we simply time out.
In theory a valid EACCES for files to which we simply have no access
should've happened already during the call to open(), except for NFS
(see open(2)).
This can be enabled with the new max-transient-error-timeout property, and
a new o-sync boolean property was added to open the file in O_SYNC mode
as without that it's not guaranteed that we get EACCES for the actual
writev() call that failed but might only get it at a later time.
Fixes https://gitlab.freedesktop.org/gstreamer/gstreamer/issues/305
