Commit graph

2389 commits

Author SHA1 Message Date
Julian Bouzas
e975d5d1c4 tracer: latency: Show element's source pad name instead of element's name
The full pad name gives more information than the element's name, which is very
useful when elements have multiple source pads.
2019-04-08 12:28:07 -04:00
Nicolas Dufresne
5470f6df00 latency: Dot not override already stored events
First, the event would be leaved, but also when an element takes
several buffers before producing one, we want the reported latency to be
the aggregation, so the distance from the oldest buffer.
2019-04-08 12:28:07 -04:00
Nicolas Dufresne
bdb75a1450 tracer: latency: Add parameter to select latency type
This sets back the default to trace only pipeline latency, and add flags
to enabled element tracing. It is now possible to only trace element
latency, only trace pipeline latency, trace both or none.
2019-04-08 12:28:07 -04:00
Nicolas Dufresne
6a35153e4e tracer: latency: Add per element latency tracer
This adds per element latency tracing.
2019-04-08 12:28:07 -04:00
Nicolas Dufresne
4585d22a8d tracer: Don't pass pads inside GstEvent
This removes the passing of pad inside of a GstEvent. While this is not
a bug, it may affect the live time of the pad, hense change the pipeline
behaviour.
2019-04-08 12:28:07 -04:00
Jonny Lamb
63170d52ae identity: fixes to the eos-after and error-after properties
I copied `error-after` to make the `eos-after` property, but it turned
out there were some problems with that one, so this patch: adds
separate counters (so setting to NULL and reusing the element will
still work); clarifies the properties' min values; and reports an
error when both are set.
2018-12-18 22:55:28 +00:00
Edward Hervey
1bca404191 queue2: Add details of query in debug log 2018-12-17 09:41:45 +01:00
Jonny Lamb
460c0edbfb identity: add eos-after property
Using `num-buffers` can be unpredictable as buffer sizes are often
arbitrary (filesrc, multifilesrc, etc.). The `error-after` property on
`identity` is better but obviously reports an error afterwards. This
adds `eos-after` which does exactly the same thing but reports EOS
instead.
2018-12-11 21:14:11 +00:00
Philippe Normand
9f5e65f0a2 input-selector: Let context queries pass through
By doing so GL source elements can successfully reuse the GL context and display
of downstream elements. This change fixes an issue in playbin when using
gltestsrc where the context query made by the source element would fail and the
source element would create a second (useless) GLDisplay.
2018-11-28 11:00:21 +00:00
Jordan Petridis
84512152c1
Run gst-indent through the files
This is required before we enabled an indent test in the CI.

https://gitlab.freedesktop.org/gstreamer/gstreamer-project/issues/33
2018-11-28 05:58:53 +02:00
Niels De Graef
cf3fdbe4c0 typefind: cleanup (un)reffing of several objects.
By using these functions, we can shave off a few lines, and make the
intent of that line more clear.
2018-11-09 08:44:11 +00:00
Matthew Waters
4fc4ad87d5 query: add a new bitrate query
Allows determining from downstream what the expected bitrate of a stream
may be which is useful in queue2 for setting time based limits when
upstream does not provide timing information.

Implement bitrate query handling in queue2

https://gitlab.freedesktop.org/gstreamer/gst-plugins-base/issues/60
2018-11-07 15:04:14 +00:00
Matthew Waters
c4ccff7861 queue2: avoid ping-pong between 0% and 100% buffering messages
If upstream is pushing buffers larger than our limits, only 1 buffer
is ever in the queue at a time.  Once that single buffer has left the
queue, a 0% buffering message would be posted followed immediately by a
100% buffering message when the next buffer was inserted into the queue
a very short time later.  As per the recommendations, This would result
in the application pausing for a short while causing the appearance of
a short stutter.

The first step of a solution involves not posting a buffering message if
there is still data waiting on the sink pad for insertion into the queue.
This successfully drops the 0% messages from being posted however a
message is still posted on each transition to 100% when the new buffer
arrives resulting in a string of 100% buffering messages.  We silence
these by storing the last posted buffering percentage and only posting a
new message when it is different from or last posted message.
2018-11-07 15:04:14 +00:00
Nicolas Dufresne
e1be065293 tracers: log: Fix post query trace
The post tracer hooks have a GstQuery argument which was truncated from
the trace. As the post hook is the one that contains the useful data,
this bug was hiding the important information from that trace.
2018-11-06 15:41:01 -05:00
Matthew Waters
874ad5faca meson: generate pkg-config files for our plugins 2018-11-05 15:18:41 +00:00
Sebastian Dröge
23e6096d0b typefind: Always forward RECONFIGURE events upstream
Based on a patch by Vincent Penquerc'h <vincent.penquerch@collabora.co.uk>

Fixes https://gitlab.freedesktop.org/gstreamer/gstreamer/issues/67
2018-11-04 12:45:57 +02:00
Sebastian Dröge
3eb334a7f9 typefind: Lower debug level of some output related to the URI query
It's not a warning if an URI doesn't have an extension, and it's also
not mandatory that sources have an URI or even answer the URI query.
2018-10-26 09:21:42 +01:00
Edward Hervey
127e211004 multiqueue: Don't clamp running times for position calculation
Since we use full signed running times, we no longer need to clamp
the buffer time.

This avoids having the position of single queues not advancing for
buffers that are out of segment and never waking up non-linked
streams (resulting in an apparent "deadlock").
2018-10-22 15:28:45 +02:00
Edward Hervey
98fabd2fa2 queue2: Reset result flow when retrying
If we ever get a GST_FLOW_EOS from downstream, we might retry
pushing new data. But if pushing that data doesn't return a
GstFlowReturn (such as pushing events), we would end up returning
the previous GstFlowReturn (i.e. EOS).

Not properly resetting it would cause cases where queue2 would
stop pushing on the first GstEvent stored (even if there is more
data contained within).
2018-10-22 13:52:30 +02:00
Nicolas Dufresne
c81b2f6207 filesink: Fix wrong printf format
We add a guint64 and a guint, the result is a guint64. On 64bit
architecture, this is the same, but on 32bit architecture, it's not.

https://bugzilla.gnome.org/show_bug.cgi?id=797127
2018-09-11 21:33:51 -04:00
Sebastian Dröge
95fcced84d input-selector: Bring latency handling in sync with GstPad code 2018-08-31 12:16:43 +03:00
Sebastian Dröge
7c179e3f58 input-selector: Apply GstPad default latency handler fixes here too 2018-08-31 11:47:03 +03:00
Sebastian Dröge
6fc136ad84 concat: Improve debug output a bit by printing pad names 2018-08-29 19:26:04 +03:00
Sebastian Dröge
b324e4b802 filesink: Flush buffers before directly writing out buffers with the SYNC_AFTER flag
Otherwise we write out the SYNC_AFTER buffer immediately, and the
previously queued up buffers afterwards which then breaks the order of
data.

Also add various debug output.
2018-08-28 14:23:32 +03:00
Sebastian Dröge
035045c981 filesink: Consider the current buffer size when checking the current position 2018-08-17 17:26:17 +03:00
Sebastian Dröge
cbcad53b45 filesink: Reset the current buffer size to NULL and clear the buffer on close and FLUSH_STOP 2018-08-17 17:26:17 +03:00
Sebastian Dröge
6b4fc62b7b filesink: Implement buffering internally
We use writev() so every call ends up going to the kernel but for small
buffers we generally would prefer to do as few write calls as possible.

https://bugzilla.gnome.org/show_bug.cgi?id=794173
2018-08-16 17:01:07 +03:00
Sebastian Dröge
e975e0cae8 filesink: Remove buffer, deprecate line-buffer mode and don't use fflush()
fflush() has no effect because we use writev() directly, so fsync()
should be used instead which is actually flushing the kernel-side
buffers.

As a next step, a non-line-buffered buffering mode is to be added.

https://bugzilla.gnome.org/show_bug.cgi?id=794173
2018-08-14 13:56:08 +03:00
Sebastian Dröge
ef9619200b inputselector: Forward LATENCY query to all sinkpads
Otherwise downstream will consider the pipeline not live if the active
pad is live, even though some inactive pads might be live and might
require a non-zero latency configuration.

https://bugzilla.gnome.org/show_bug.cgi?id=796901
2018-08-01 14:23:04 +03:00
Sebastian Dröge
41f1a40704 typefind: Add new gst_type_find_helper_get_range_full() that returns flow return
And make use of it in the typefind element. It's useful to distinguish
between the different errors why typefinding can fail, and especially to
not consider GST_FLOW_FLUSHING as an actual error.

https://bugzilla.gnome.org/show_bug.cgi?id=796894
2018-08-01 14:23:04 +03:00
Nirbheek Chauhan
38ec95460f meson: Don't add static printf library to executables
They should only need to link to libgstreamer.
2018-07-25 16:02:06 +05:30
Sebastian Dröge
d196b0b904 typefind: Add _with_extension() variants for typefinding data or a buffer
And make use of that in the typefind element to also be able to make use
of the extension in push mode. It previously only did that in pull mode
and this potentially speeds up typefinding and might also prevent false
positives.

https://bugzilla.gnome.org/show_bug.cgi?id=796865
2018-07-25 01:12:11 +03:00
Edward Hervey
eeb30c2a97 concat: Properly forward the SEGMENT seqnum 2018-06-19 07:27:33 +02:00
Dimitrios Katsaros
ac7f8b6337 queue2: use GstQueueArray
When using queue2 as a queue it was using GQueue with
individually allocated queue items, so two allocs for
each item. With GstQueueArray we can avoid those.

https://bugzilla.gnome.org/show_bug.cgi?id=796483
2018-06-04 11:10:05 +01:00
Nirbheek Chauhan
4fb02fc85b meson: Update option names to omit disable_ and with- prefixes
Also yield common options to the outer project (gst-build in our case)
so that they don't have to be set manually.
2018-05-05 20:30:43 +05:30
luz.paz
4a402c1c7d Fix typos in comments and docs
Found via `codespell`

https://bugzilla.gnome.org/show_bug.cgi?id=795610
2018-05-01 11:18:03 +01:00
Xavier Claessens
b00b1d5361 Meson: Use library() to build both static and shared libs
Meson supports building both static and shared libraries in a single
library() call. It has the advantage of reusing the same .o objects and
thus avoid double compilation.

https://bugzilla.gnome.org/show_bug.cgi?id=794627
2018-04-25 00:40:30 +01:00
Tim-Philipp Müller
21decc643b multiqueue, inputselector: show pad properties in gst-inspect-1.0 2018-04-17 18:57:20 +01:00
Tim-Philipp Müller
65e0907798 inputselector, outputselector: add guards for wrong pads being set as active pads
Catch users wrongly setting foreign pads or wrong pads as
the selector's active pad, which leads to all kinds of
other issues. It's a programming error so handle it just
like we would if we had direct API.

https://bugzilla.gnome.org/show_bug.cgi?id=795309
2018-04-17 18:57:17 +01:00
Matthew Waters
5a5a548b09 plugins: Don't force 64-bit file/seek functions variants on android
Most functions are automatically chosen from the _FILE_OFFSET_BITS
define, the remaining one (fstat) is only available on API >= 21 so
check for that
2018-03-02 18:25:21 +11:00
Olivier Crête
23b32d5600 queue: Ignore thresholds if a query is queued
The queue gets filled by the tail, so a query will always be the tail
object, not the head object. Also add a _peek_tail_struct() method to the
GstQueueArray to enable looking at the tail.

With unit test to prevent future regression.

https://bugzilla.gnome.org/show_bug.cgi?id=762875
2018-03-01 16:31:57 -05:00
Matthew Waters
7fad93d035 tracers: latency: allow for non parented pads to send latency probes
Such a setup is used in rtspsrc for its TCP connection

https://bugzilla.gnome.org/show_bug.cgi?id=793478
2018-02-16 13:14:40 +00:00
Jan Alexander Steffens (heftig)
b17d03cb0a multiqueue: Don't start new pads until parented
Start task on new source pads added at runtime after they
have been added to the element, not during activation.

This ensures the pads can post their CREATE stream-status
messages and the application can set thread priorities.

https://bugzilla.gnome.org/show_bug.cgi?id=756867
2017-12-24 11:51:06 +01:00
Jan Alexander Steffens (heftig)
10c3a7bd55 multiqueue: Split task handling from gst_single_queue_flush
https://bugzilla.gnome.org/show_bug.cgi?id=756867
2017-12-24 11:51:06 +01:00
Sebastian Dröge
d431ef7190 downloadbuffer: Don't hold the mutex while posint the download-complete message
Something might handle it from a sync message handler and call back into
downloadbuffer, causing a deadlock.
2017-12-21 13:47:52 +02:00
Matthew Waters
a0400a0d28 docs: include tracers in the documentation
Requires exposing the tracer GType from the GstTracerFactory in order
to link the plugin with the tracer in the documentation.

https://bugzilla.gnome.org/show_bug.cgi?id=791253
2017-12-14 14:46:46 +11:00
Tim-Philipp Müller
e04c6694f1 tracers: log: no need to link to our internal printf implementation
The call to __gst_vasprintf() was removed in commit 1a3e218b8.
2017-11-26 11:14:01 +00:00
Mikhail Fludkov
6577f5e91e gsttraceutils: actually disable tracing system hooks if configured
`./configure --disable-gst-tracer-hooks` didn't do anything, hooks were
always enabled regardless of the option. It works correctly in the
Meson build though.
2017-11-24 13:40:34 +01:00
Havard Graff
df27ec3e67 gstbasetranform: replace GST_BASE_TRANSFORM with GST_BASE_TRANSFORM_CAST
To avoid a global type-lock on chain etc.
2017-11-24 13:39:39 +01:00
Edward Hervey
922e048379 queue: Only calculate level if we have valid levels
Doing calculations with GST_CLOCK_STIME_NONE would result in
completely bogus levels
2017-11-23 15:28:39 +01:00
Edward Hervey
f03443f90c concat: Handle single-pad use-cases
When EOS reaches concat, it will switch to the next candidate as its
activate pad.

The problem arises when there is only one sinkpad, the "active" pad
becomes NULL. This results in concat becoming unusable after it receives
a *single* EOS on its single sinkpad.

If we detect there is a single sinkpad and there is no current active pad:
* If we are waiting (from selected sink event/buffer), become the current
  active pad.
* If there is a seek request, send it upstream. We don't switch the
  active_sinkpad property at that point in time, since the seek could
  fail. If the seek succeeds, the following SEGMENT (or STREAM_START)
  will cause the pad_wait() to elect that pad as the new active one.
* Flush events get forwarded

https://bugzilla.gnome.org/show_bug.cgi?id=790167
2017-11-10 14:15:46 +01:00
Edward Hervey
08ad748ced concat: Make QoS forward MT-safe
In the same way it's done for other event forwarding.
2017-11-09 17:41:51 +01:00
Guillaume Desmottes
4670036691 latency tracer: add timestamp to tracer records
Include the timestamp of the recorded log as in the 'stats' tracer.
This can be useful, for example, to plot a graph showing the latency
over time.

https://bugzilla.gnome.org/show_bug.cgi?id=781315
2017-11-04 13:51:00 -04:00
Mathieu Duponchelle
f911fe4314 gstleaks.c: always log leaks listing
https://bugzilla.gnome.org/show_bug.cgi?id=789556
2017-11-02 14:53:48 +01:00
Nicolas Dufresne
41e35c334b latency-tracer: Exclude synchronization time
The goal of this tracer is to measure the processing latency between a
src and a sink. In push mode, the time was read after the chain function
have returned. As the amount of time we wait to get synched is reverse
to the amount of latency the source introduced, the result was quite
surprising.

This patch moves the latency calculation in the pre-push hook. When
there is no processing in a a pipeline (e.g. fakesrc ! fakesink), the
latency will now be 0 as it's supposed to. For pull mode, the code was
already correct. When GstBaseSink operate in pull mode, the processing
time is done durring the pull, so pull-post is the right hook. The
synchronization will happen after the pull has ended. Note that
GstBaseSink rarely operate in pull mode.

https://bugzilla.gnome.org/show_bug.cgi?id=788431
2017-10-30 15:33:15 -04:00
Reynaldo H. Verdejo Pinochet
402159fb88 Use proper GtkDoc notation for NULL/FALSE/TRUE 2017-10-03 13:54:25 -07:00
Stefan Sauer
21c3f0ff22 tee: don't create a pool if none is needed
If the aggregated size is 0 and we create a pool, the pool would provide
buffers with no memory assigned. Handle that case and skip the pool.
This was the behaviour before cf803ea9f4.

Add a test for this scenario.

https://bugzilla.gnome.org/show_bug.cgi?id=730758
2017-10-02 13:58:04 -04:00
Mathieu Duponchelle
053f1523ad Multiqueue: don't allow dropping SEGMENT_DONE events
https://bugzilla.gnome.org/show_bug.cgi?id=780795
2017-09-18 14:12:06 +02:00
Tim-Philipp Müller
c2917b1ae2 plugins: use new gst_buffer_list_calculate_size() 2017-09-17 12:36:03 +01:00
Nicolas Dufresne
2cc5c539a3 identity: Add a drop-allocation property
When enabled, this property will make the allocation query fail. This is
the same as one could have done using a tee before the tee started
implementing the allocation query.

https://bugzilla.gnome.org/show_bug.cgi?id=730758
2017-09-06 10:03:20 -04:00
Nicolas Dufresne
1cd0dd3503 tee: Allocate one more buffer when multi-plexing
This extra buffer ensure that the downstream threads are not starved
when multiplexing a stream.

https://bugzilla.gnome.org/show_bug.cgi?id=730758
2017-09-05 16:00:09 -04:00
Nicolas Dufresne
cf803ea9f4 tee: Implement allocation query aggregation
This will aggregate allocation params, pool and will keep all
meta that has no parameters.

https://bugzilla.gnome.org/show_bug.cgi?id=730758
2017-09-05 16:00:09 -04:00
Nicolas Dufresne
945322f5a6 tee: Deprecate alloc-pad property
It has no effect, not implemented, and would lead to bad rendering.

https://bugzilla.gnome.org/show_bug.cgi?id=730758
2017-09-05 15:59:46 -04:00
Sebastian Dröge
82ed369991 identity: Return FLUSHING instead of EOS and don't start waiting for anything if currently flushing
Otherwise we might try unscheduling a clock id (that does not exist
yet), then the streaming thread waits for id and the state change never
continues because the streaming thread is blocked.

Also shutting down and flushing and similar should return FLUSHING, not
EOS. The stream is not over, we're just not accepting any buffers
anymore.
2017-08-16 22:50:49 +03:00
Sebastian Dröge
36372a9f4c queue: Allow re-usability after EOS
After EOS, it is possible for a pad to be resetted by sending
either a STREAM_START or SEGMENT event

Mimic the same behaviour when receiving STREAM_START/SEGMENT events
in queue if we are EOS'd

https://bugzilla.gnome.org/show_bug.cgi?id=786056
2017-08-11 11:12:09 +03:00
Edward Hervey
6566eff20d queue2: Allow re-usability after EOS
After EOS, it is possible for a pad to be resetted by sending
either a STREAM_START or SEGMENT event

Mimic the same behaviour when receiving STREAM_START/SEGMENT events
in queue2 if we are EOS'd

https://bugzilla.gnome.org/show_bug.cgi?id=786056
2017-08-09 16:18:40 +02:00
Edward Hervey
e2f8dc0aef plugins: *queue* elements: Handle STREAM_START in EOS situation
When queue-like elements are in "EOS" situation (received GST_FLOW_EOS
from downstream or EOS was pushed), they drain buffers/events that
wouldn't be processed anyway and let through events that might
modify the EOS situation.

Previously only GST_EVENT_EOS and GST_EVENT_SEGMENT events were let
through, but we also need to allow GST_EVENT_STREAM_START to go
through since it resets the EOS state of pads since 1.6

https://bugzilla.gnome.org/show_bug.cgi?id=786034
2017-08-09 10:54:17 +02:00
Edward Hervey
4312119d16 queue2: Handle buffering levels on NOT_LINKED
When downstream returns NOT_LINKED, we return the buffering level
as being 100%.

Since the queue is no longer being consumed/used downstream, we
want applications to essentially "ignore" this queue for buffering
purposes.

If other streams are still being used, those stream buffering levels
will be used. If none are used, upstream will post an error message
on the bus indicating no streams are used.

https://bugzilla.gnome.org/show_bug.cgi?id=785799
2017-08-07 15:20:55 +02:00
Tim-Philipp Müller
6575389908 downloadbuffer: remove unused struct member
This was used in queue2 when handling in coming serialized
queries, but downloadbuffer just refuses serialized queries.
2017-07-26 14:20:44 +01:00
Edward Hervey
78ee0dfad1 multiqueue: Fix access to NULL pointer
sq can be NULL.

Also fix commit message (it's the queue we are iterating over that we
are logging, not the one passed as argument).

CID #1415569
2017-07-20 14:18:51 +02:00
Seungha Yang
1b0059e0c5 multiqueue: Calculate interleave only within each streaming thread
... and use the biggest interleave value among streaming threads.
This is to optimize multiqueue size adaptation on adaptive streaming
use case with "use-interleave" property.

https://bugzilla.gnome.org/show_bug.cgi?id=784448
2017-07-19 17:19:14 +02:00
Nicolas Dufresne
4a108c7858 meson: Use join_paths and .set_quoted where possible 2017-07-11 14:57:26 -04:00
Tim-Philipp Müller
d93ebba58b meson: add option to disable tracer hooks 2017-07-05 13:27:48 +01:00
Thibault Saunier
bd73551cc0 leaks: Handle subclasses in filters even for unhandled/lazy loaded types
Using typename in the set of unhandled types instead of the quark so
that we also handle subclasses as with other filters.
2017-05-20 15:42:16 +02:00
Nicolas Dufresne
d143af2c77 Remove plugin specific static build option
Static and dynamic plugins now have the same interface. The standard
--enable-static/--enable-shared toggle are sufficient.
2017-05-16 13:29:38 -04:00
Thiago Santos
058bdcfe6b queue: avoid return flushing if we have a not-linked
Return the correct flow return instead of returning always flushing.
This would cause queue to convert not-linked to flushing and making
upstream elements stop.

Based on the previous patch for queue2.

https://bugzilla.gnome.org/show_bug.cgi?id=776999
2017-03-14 22:20:04 -07:00
Thiago Santos
045700e80e queue2: avoid return flushing if we have a not-linked
Return the correct flow return instead of returning always flushing.
This would cause queue2 to convert not-linked to flushing and making
upstream elements stop.

https://bugzilla.gnome.org/show_bug.cgi?id=776999
2017-03-11 11:19:33 -08:00
Wim Taymans
586b34436e buffer: handle gst_buffer_map failures 2017-03-10 10:13:05 +01:00
Wim Taymans
64a1d4ad0b downloadbuffer: unlock mutex in error case 2017-03-10 10:12:49 +01:00
Jan Schmidt
9b0a6ca7ab multiqueue: Make min-interleave-time a configurable property
Remove a FIXME about making the minimum interleave
buffering a configurable property
2017-03-09 12:09:57 +11:00
Seungha Yang
89082272f8 inputselector: Always proxy position/duration query
active-pad switch causes reconfigure event with lock taken,
and upstream element might query the current position or duration
before returning the reconfigure event.
Meanwhile, gst_input_selector_get_linked_pad() is used to get srcpad
inside of default query handle, and it takes also lock.
Since inputselector is still locked by active-pad switch, and so the query
cannot be handled further.

https://bugzilla.gnome.org/show_bug.cgi?id=775445
2017-03-07 12:48:47 +02:00
Thibault Saunier
a87b4551a6 Port gtk-doc comments to their equivalent markdown syntax
Modernizing our documentation and preparing a possible move to hotdoc.
This commits also adds missing @title metadatas to all SECTIONs
2017-01-27 16:36:38 -03:00
Jan Schmidt
0528645f7c identity: Add ts-offset property.
Add a property to delay or advance sync time
when sync=true, with the same behaviour as
the ts-offset property in basesink
2017-01-16 09:40:25 +11:00
Jan Schmidt
748687fc30 queue: Don't generate GST_FLOW_ERROR without logging
At least log a message to the debug log when generating
a GST_FLOW_ERROR, to make it possible to find where it came from.
2017-01-03 02:21:10 +11:00
Stefan Sauer
a5246f91af gstrusage: explicitly register to hooks
We were attaching to any probe point to take rusage samples. The new refcount
hooks are called way too frequently though to make this still feasible.
2016-12-21 22:19:51 +01:00
Nirbheek Chauhan
97f39cae63 meson: Derive defines from header/function names
This is what Autoconf already does for us, so just do this. Avoids
people making typos while adding header or function checks. Because we
use a config.h.meson, such typos won't even be noticed.

Also, starting from Meson 0.36.0, the XCode 8 workaround that we use for
clock_gettime is no longer needed.
2016-12-21 23:51:36 +05:30
Stefan Sauer
11cc260448 latency: the latency is not an aggregated value
The logged latencies are individual meassurements.
2016-12-20 21:07:14 +01:00
Thibault Saunier
29f0a79880 leaks: Allow user to set the flags to use to retrieve stack traces
https://bugzilla.gnome.org/show_bug.cgi?id=775541
2016-12-20 15:29:10 -03:00
Thibault Saunier
30133909ce leaks: Allow tracing Gst(Mini)Object reffing operations
It makes it much simpler to later debug refcount issues.

https://bugzilla.gnome.org/show_bug.cgi?id=775541
2016-12-20 15:29:10 -03:00
Thibault Saunier
32b17a8f3d leaks: Allow passing a GstStructure to configure the tracer
But keep understanding the simple synthax with a comma separated
list of filters

https://bugzilla.gnome.org/show_bug.cgi?id=775541
2016-12-20 15:29:10 -03:00
Jan Schmidt
e64765c0d8 typefind: Switch to normal mode before have-type
Before emitting have-type, switch to NORMAL
mode, as part of the have-type processing sends
the caps event downstream, which might trigger
actions like downstream autoplugging or
flushing seeks - and the latter are only
passed upstream if we've set typefind to NORMAL
mode.
2016-12-21 00:40:10 +11:00
Sebastian Dröge
47f9223750 multiqueue: Protect against spurious wakeups of the condition variable 2016-12-18 13:59:12 +02:00
Stefan Sauer
1bfd04a5fe tracers/stats: log optional fields instead of GST_CLOCK_TIME_NONE
Simplify the traces and avoid trace analyzer to know that ((1<<64) - 1) means
we had no value.
2016-12-16 15:55:21 +01:00
Sebastian Dröge
3fab00b46a queue/queue2: Protect against spurious condition variable wakeups
Make sure that we only wake up when we have to flush, or when this
specific query was handled.

https://bugzilla.gnome.org/show_bug.cgi?id=776039
2016-12-13 21:03:54 +02:00
Sebastian Dröge
33c239828b queue/queue2: Ensure that the streaming thread is unlocked after deactivating the srcpad
It might happen that the srcpad task function is never called at all, in
which case unlocking everything from there will never happen.

Make sure to unlock everything another time after the task function is
definitely stopped.

https://bugzilla.gnome.org/show_bug.cgi?id=776039
2016-12-13 21:03:54 +02:00
Thibault Saunier
33616d47be info: Add a 'flags' parametter to gst_debug_get_stack_trace
This is an API break but that API has not been released yet.

We are passing a flag rather than a simple boolean as we can imagine
to implement more features in the future for example to retrieve a
stack trace for all the threads, etc..

Retrieving source file and line numbers is pretty
expensive while getting a stack trace, this new argument
allows the user to decide to retrieve a backtrace
without those infos instead which is much faster.

For example running $ GST_LEAKS_TRACER_STACK_TRACE=1 GST_DEBUG=GST_TRACER:7 \
GST_TRACERS=leaks time gst-launch-1.0 videotestsrc num-buffers=1 ! fakesink:

* With simple stack traces:

    0.04s user 0.02s system 99% cpu 0.060 total

* With full stack traces:

    0.66s user 0.23s system 96% cpu 0.926 total

https://bugzilla.gnome.org/show_bug.cgi?id=775423
2016-12-12 15:06:20 -03:00
Edward Hervey
6b73bf38d1 filesrc: Set GError in another error case
When changing the location while open, properly set the GError regarding
the failure.
2016-12-12 16:19:13 +01:00
Seungha Yang
ce11ddeb0c multiqueue: Fix overflow on get_buffering_level()
guint64 denominator factor for gst_util_uint64_scale_int() can cause overflow

https://bugzilla.gnome.org/show_bug.cgi?id=775921
2016-12-11 13:30:52 +02:00
Thibault Saunier
16bd0e7bac meson: Fix build 2016-12-09 19:31:51 -03:00
Thibault Saunier
bd5d43e8f9 meson: Support building with Gst debug disabled 2016-12-09 17:57:10 -03:00
Sebastian Dröge
d284be3505 typefind: Use gst_query_has_scheduling_mode_with_flags() convenience function 2016-12-09 18:02:41 +02:00
Stefan Sauer
a25453c2c6 tracers/log: log more detail
Log the objects like we would in GST_TRACE_OBJECT. Add the hook function into
the fucntion field.
2016-12-08 22:03:19 +01:00
Stefan Sauer
8d819a51b8 tracer/log: fix hook prototype
s/GstElement/GstPad/
2016-12-08 22:02:17 +01:00
Stefan Sauer
cfb6c7d4f5 tracer/latency: clear qdata
When reading the qdata, clear it to avoid it being read and unreffed again.
Fixes #774332
2016-12-07 21:53:49 +01:00
Tim-Philipp Müller
cf7cfb0a0e fakesink, identity: print metas attached to buffer in silent=false mode 2016-12-05 20:29:25 +00:00
Sebastian Dröge
e02405b284 elements: Handle GstIterator RESYNC return value correctly in gst_iterator_foreach() 2016-12-05 11:01:45 +02:00
Stefan Sauer
788c0d220a tracers/log: log messages in message category 2016-12-04 12:15:54 +01:00
Nirbheek Chauhan
ad592f86b6 meson: Add Autotools changes that weren't mirrored
commits:
a7d282d272
6fdb4df0f8
1aceebd67f
2016-11-28 19:28:27 +05:30
Tim-Philipp Müller
1aceebd67f elements: add dataurisrc to build
Moved from -bad.
2016-11-28 11:19:12 +00:00
Tim-Philipp Müller
76c47b7e9c dataurisrc: fix string leak in property getter 2016-11-28 11:19:07 +00:00
Tim-Philipp Müller
a837ff6581 Move dataurisrc element from -bad
https://bugzilla.gnome.org/show_bug.cgi?id=774527
2016-11-28 11:18:39 +00:00
Olivier Crête
5216322d39 fdsink: Block in preroll_wait on unlock
The correct behaviour of anything stuck in the ->render() function
between ->unlock() and ->unlock_stop() is to call
gst_base_sink_wait_preroll() and only return an error if this returns an
error, otherwise, it must continue where it left off!

https://bugzilla.gnome.org/show_bug.cgi?id=773912
2016-11-23 11:58:43 -05:00
Jan Schmidt
fa3eac2d28 multiqueue: Make sure not-linked streams get woken up
When running in sync-by-running-time mode, pad groups
that have exactly 1 pad and it's not-linked might never
wake up after computing a high time, as the per-pad-group
high time was only recomputed when a pad in the group
advances.

Wake those up using the global multiqueue high-time across
all other groups instead.

https://bugzilla.gnome.org/show_bug.cgi?id=774322
2016-11-17 00:46:31 +11:00
Jan Schmidt
a6ca8dfb42 queues: Don't return negative position queries.
When subtracting queued data sizes from upstream queries
in queue, queue2, downloadbuffer and typefind, clamp the
result to not go negative, in case upstream returned
a nonsense value that's too small (as could happen if
upstream is estimating, or just broken)
2016-11-16 00:12:22 +11:00
Sebastian Dröge
abe60a92c4 funnel: Make sure to only lock the stream lock once
We also only unlock it once, and otherwise have it locked forever from
this thread, causing deadlocks on shutdown later.
2016-11-09 11:37:09 +02:00
Thibault Saunier
85179a674f tracers: leaks: Use the new gst_debug_get_stack_trace
And remove the local implementation of it.

https://bugzilla.gnome.org/show_bug.cgi?id=772555
2016-11-04 14:22:05 -03:00
Thibault Saunier
a8d4857555 gst: Use libunwind/libdw to generate backtraces if avalaible
Making the gst_debug_print_trace function more generally useful.

API:
  + gst_debug_get_trace

https://bugzilla.gnome.org/show_bug.cgi?id=772555
2016-11-04 14:22:05 -03:00
Jan Schmidt
812d4976f0 fakesink: Add property to not drop out-of-segment buffers
Implement handling in basesink to not unconditionally discard
out-of-segment buffers and expose it as a new property on fakesink
(not unconditionally in all basesink based sinks).

The property defaults to FALSE.

https://bugzilla.gnome.org/show_bug.cgi?id=765734
2016-11-02 11:21:37 +11:00
Sebastian Dröge
c56e1d18c1 funnel: Always push all sticky events whenever we forward a serialized event
Otherwise downstream will have an inconsistent set of sticky events at this
point, e.g. when a TAG event is pushed and downstream wants to relate it to
the stream by looking at the current STREAM_START event.

https://bugzilla.gnome.org/show_bug.cgi?id=768526
2016-11-01 21:01:28 +02:00
Sebastian Dröge
25bf63d6f7 tee: Set GST_PAD_FLAG_PROXY_CAPS before forwarding sticky events
https://bugzilla.gnome.org/show_bug.cgi?id=752213
2016-11-01 20:53:43 +02:00
Sebastian Dröge
cfdb97a135 Revert "queue: Fix race when calculating cur_level.time"
This reverts commit d03bd54780.

It breaks the unit test, although it ensures that only correct values
are used for calculations. Needs to be fixed up.

https://bugzilla.gnome.org/show_bug.cgi?id=773096
2016-10-21 15:23:14 +03:00
Stian Selnes
d03bd54780 queue: Fix race when calculating cur_level.time
On the first buffer, it's possible that sink_segment is set but
src_segment has not been set yet. If this is the case, we should not
calculate cur_level.time since sink_segment.position may be large and
src_segment.position default is 0, with the resulting diff being larger
than max-size-time, causing the queue to start leaking (if
leaky=downstream).

One potential consequence of this is that the segment event may be
stored on the srcpad before the caps event is pushed downstream, causing
a g_warning ("Sticky event misordering, got 'segment' before 'caps'").

https://bugzilla.gnome.org/show_bug.cgi?id=773096
2016-10-20 14:14:20 +03:00
Thibault Saunier
2c6dd70217 stats: Fix warning 2016-10-14 17:30:01 +02:00
Vincent Penquerc'h
18d5d118ab tracers: fix structure leak
https://bugzilla.gnome.org/show_bug.cgi?id=772851
2016-10-13 13:06:40 +01:00
Edward Hervey
ace418d226 queue2: Implement gst_event_full_func handling
Same as we do for queue
2016-10-08 13:56:18 +02:00
Stefan Sauer
ede686c39b tracer/rusage: fix format string args
The format string contains a process id, but we did not provice one. This
caused us to log garbage since all args got shifted.
2016-10-08 13:22:23 +02:00
Tim-Philipp Müller
9791669ac3 meson: tracers: signal availability of libunwind and backtrace() to code
Not setting cdata here on purpose because of .. complications.
2016-09-27 18:24:02 +01:00
Thibault Saunier
c4aed66eff meson: Build tracers 2016-09-26 18:27:00 -03:00
Carlos Rafael Giani
7413064f06 queue2: Update buffering if its enabled and low/high watermarks are changed
https://bugzilla.gnome.org/show_bug.cgi?id=771210
2016-09-17 09:54:29 +01:00
Carlos Rafael Giani
5988095f90 multiqueue: Add higher-resolution low/high-watermark properties
low/high-watermark are of type double, and given in range 0.0-1.0. This
makes it possible to set low/high watermarks with greater resolution,
which is useful with large multiqueue max sizes and watermarks like 0.5%.

Also adding a test to check the fill and watermark level behavior.

https://bugzilla.gnome.org/show_bug.cgi?id=770628
2016-08-31 12:56:19 +03:00
Carlos Rafael Giani
67874ea86d multiqueue: Distinguish between buffering percentage and buffering level
To make the code clearer, and to facilitate future improvements, introduce
a distinction between the buffering level and the buffering percentage.

Buffering level: the queue's current fill level. The low/high watermarks
are in this range.

Buffering percentage: percentage relative to the low/high watermarks
(0% = low watermark, 100% = high watermark).

To that end, get_percentage() is renamed to get_buffering_level(). Also,
low/high_percent are renamed to low/high_watermark to avoid confusion.
mq->buffering_percent values are now normalized in the 0..100 range for
buffering messages inside update_buffering(), and not just before sending
the buffering message. Finally the buffering level range is parameterized
by adding a new constant called MAX_BUFFERING_LEVEL.

https://bugzilla.gnome.org/show_bug.cgi?id=770628
2016-08-31 12:56:19 +03:00
Carlos Rafael Giani
e83412b4fd multiqueue: Rename percent/percent_changed to buffering_percent(_changed)
This is a prerequisite for subsequent commits, and makes queue2 and
multiqueue code a little more consistent.

https://bugzilla.gnome.org/show_bug.cgi?id=770628
2016-08-31 12:56:19 +03:00
Edward Hervey
3117525cb6 multiqueue: Fix high_time wakeup logic
When calculating the high_time, cache the group value in each singlequeue.

This fixes the issue by which wake_up_next_non_linked() would use the global
high-time to decide whether to wake-up a waiting thread, instead of the group
one, resulting in those threads constantly spinning.

Tidy up a bit the waiting logic while we're at it.

With this patch, we go from 212% playing a 8 audio / 8 video file down to less
than 10% (most of it being the video decoding).

https://bugzilla.gnome.org/show_bug.cgi?id=770225
2016-08-30 12:28:40 +03:00
Thibault Saunier
4714ef2f8e Make use of the new GST_ELEMENT_FLOW_ERROR API all around.
https://bugzilla.gnome.org/show_bug.cgi?id=770158
2016-08-27 09:33:20 -03:00
Carlos Rafael Giani
db66cb51b3 queue2: Add higher-resolution low/high-watermark properties
low/high-watermark are of type double, and given in range 0.0-1.0. This
makes it possible to set low/high watermarks with greater resolution,
which is useful with large queue2 max sizes and watermarks like 0.5%.

Also adding a test to check the fill and watermark level behavior.

https://bugzilla.gnome.org/show_bug.cgi?id=769449
2016-08-25 11:54:33 +03:00
Carlos Rafael Giani
e0f1a9e618 queue2: Distinguish between buffering percentage and buffering level
To make the code clearer, and to facilitate future improvements, introduce
a distinction between the buffering level and the buffering percentage.

Buffering level: the queue's current fill level. The low/high watermarks
are in this range.

Buffering percentage: percentage relative to the low/high watermarks
(0% = low watermark, 100% = high watermark).

To that end, get_buffering_percent() is renamed to get_buffering_level(),
and the code at the end that transforms to the buffering percentage is
factored out into a new convert_to_buffering_percent() function. Also,
the buffering level range is parameterized by adding a new constant called
MAX_BUFFERING_LEVEL.

https://bugzilla.gnome.org/show_bug.cgi?id=769449
2016-08-25 11:54:33 +03:00
Nirbheek Chauhan
b2f9808722 Add support for Meson as alternative/parallel build system
https://github.com/mesonbuild/meson

With contributions from:

Tim-Philipp Müller <tim@centricular.com>
Mathieu Duponchelle <mathieu.duponchelle@opencreed.com>
Jussi Pakkanen <jpakkane@gmail.com> (original port)

Highlights of the features provided are:
* Faster builds on Linux (~40-50% faster)
* The ability to build with MSVC on Windows
* Generate Visual Studio project files
* Generate XCode project files
* Much faster builds on Windows (on-par with Linux)
* Seriously fast configure and building on embedded

... and many more. For more details see:

http://blog.nirbheek.in/2016/05/gstreamer-and-meson-new-hope.html
http://blog.nirbheek.in/2016/07/building-and-developing-gstreamer-using.html

Building with Meson should work on both Linux and Windows, but may
need a few more tweaks on other operating systems.
2016-08-19 21:26:14 +01:00
Edward Hervey
5154dcfbdc queue2: Post buffering messages earlier in ringbuffer mode
In ringbuffer mode we need to make sure we post buffering messages *before*
blocking to wait for data to be drained.

Without this, we would end up in situations like this:
* pipeline is pre-rolling
* Downstream demuxer/decoder has pushed data to all sinks, and demuxer thread
  is blocking downstream (i.e. not pulling from upstream/queue2).
* Therefore pipeline has pre-rolled ...
* ... but queue2 hasn't filled up yet, therefore the application waits for
  the buffering 100% messages before setting the pipeline to PLAYING
* But queue2 can't post that message, since the 100% message will be posted
  *after* there is room available for that last buffer.

https://bugzilla.gnome.org/show_bug.cgi?id=769802
2016-08-12 16:26:40 +02:00
Josep Torra
5f021759b3 multiqueue: removed redundant call to g_thread_self
Remove an unneeded call to g_thread_self and minor coding style fix.
2016-08-08 16:42:06 +02:00
Jan Schmidt
b50a12b8ac inputselector: Handle stream-group-done
Handle the new stream-group-done message to unblock pads which
are waiting for the running time to advance on that group.

https://bugzilla.gnome.org/show_bug.cgi?id=768995
2016-07-25 20:22:20 +10:00
Jan Schmidt
4981e09bc7 inputselector: Wake other pads when selected goes EOS
Other pads that are waiting for the stream on the selected
pad to advance before they finish waiting themselves
should be given the chance to do so when the selected pad
goes EOS. Fixes problems where input streams can end up
waiting forever if the active stream goes EOS earlier than
their own end time.
2016-07-25 19:34:52 +10:00
Guillaume Desmottes
714ea37282 tracers: leaks: update type filter later for unknown types
This allow us to filter using an object type which is implemented
by a plugin like, say, GstGtkGLSink.

https://bugzilla.gnome.org/show_bug.cgi?id=768989
2016-07-20 14:14:48 +01:00
Guillaume Desmottes
59885c2a92 leaks: check return values of libunwind calls 2016-07-11 12:44:19 +01:00
Edward Hervey
1ff80fbf52 queue2: Fix average input rate calculation on small input range
When dealing with small-ish input data coming into queue2, such as
adaptivedemux fragments, we would never take into account the last
<200ms of data coming in.

The problem is that usually on TCP connection the download rate
gradually increases (i.e. the rate is lower at the beginning of a
download than it is later on). Combined with small download time (less
than a second) we would end up with a computed average input rate
which was sometimes up to 30-50% off from the *actual* average input
rate for that fragment.

In order to fix this, force the average input rate calculation when
we receive an EOS so that we take into account that final window
of data.

https://bugzilla.gnome.org/show_bug.cgi?id=768649
2016-07-11 10:08:44 +02:00
Guillaume Desmottes
9ba53ff1fc leaks: warn if object is destroyed while the tracer is disposing
This should not happen and generally means some thread is still running.

https://bugzilla.gnome.org/show_bug.cgi?id=768578
2016-07-08 17:35:30 +01:00
Guillaume Desmottes
1ed4140d00 leaks tracer: use G_OS_UNIX to check for signal support
Checking for signal.h is not good enough as it's present in Windows.
Those signals are UNIX specific anyway.

https://bugzilla.gnome.org/show_bug.cgi?id=767857
2016-07-08 12:38:46 +03:00
Guillaume Desmottes
17c37efa83 leaks tracer: add creation stack trace support
This allow us to provide the trace of leaked objects making it easier
to debug.

https://bugzilla.gnome.org/show_bug.cgi?id=767862
2016-07-08 12:07:53 +03:00
Guillaume Desmottes
3bb5c1e73a leaks tracer: add checkpoint support using SIGUSR2
https://bugzilla.gnome.org/show_bug.cgi?id=767857
2016-07-08 11:23:43 +03:00
Guillaume Desmottes
53d2e8c977 leaks tracer: log alive objects when receiving SIGUSR1
We don't want to automatically catch signals so use an env variable to
enable this feature.

https://bugzilla.gnome.org/show_bug.cgi?id=767857
2016-07-08 11:23:38 +03:00
Sebastian Dröge
528fbfe7c1 funnel: Only forward sticky events on GAP events if needed
That is, if the active pad changed and if forwarding of sticky events is
requested at all. We otherwise forward events too often.
2016-07-07 13:32:39 +03:00