We need to take the state lock here to ensure that we're
not currently just before setting the state of this child
element. Otherwise it can happen that we removed the element
here and e.g. set it to NULL state, and shortly afterwards
have another thread set it to a higher state again as part of
a state change for the whole bin.

When adding an element to the bin this is not needed as we
require callers to always ensure after adding to the bin that
the new element is set to the correct state.

Sebastian Dröge authored Jan 02, 2019 and Sebastian Dröge committed Jan 03, 2019

We have to ensure that all background threads from thread pools are shut
down, or otherwise they might not have had a chance yet to drop their
last reference to the pipeline and then the assertion for a reference
count of 1 on the pipeline fails.

Sebastian Dröge authored Jan 02, 2019 and Sebastian Dröge committed Jan 03, 2019

And check for exact times as we can now do that thanks to the test clock
being deterministic.

Fixes #313

Michael Olbrich authored Sep 26, 2018 and Tim-Philipp Müller committed Dec 31, 2018

This adds two custom gdb commands:

'gst-dot' creates dot files that a very close to what
GST_DEBUG_BIN_TO_DOT_FILE() produces. Object properties and buffer content
(e.g. codec-data in caps) are not available.

'gst-print' produces high-level information about GStreamer objects. This
is currently limited to pads for GstElements and events for the pads. The
output can look like this:

(gdb) gst-print pad.object.parent
GstMatroskaDemux (matroskademux0) {
    SinkPad (sink, pull) {
    }
    SrcPad (video_0, push) {
      events:
        stream-start:
          stream-id: 0463ccb080d00b8689bf569a435c4ff84f9ff753545318ae2328ea0763fd0bec/001:1274058367
        caps: video/x-theora
          width: 1920
          height: 800
          pixel-aspect-ratio: 1/1
          framerate: 24/1
          streamheader: < 0x5555557c7d30 [GstBuffer], 0x5555557c7e40 [GstBuffer], 0x7fffe00141d0 [GstBuffer] >
        segment: time
          rate: 1
        tag: global
          container-format: Matroska
    }
    SrcPad (audio_0, push) {
      events:
        stream-start:
          stream-id: 0463ccb080d00b8689bf569a435c4ff84f9ff753545318ae2328ea0763fd0bec/002:1551204875
        caps: audio/mpeg
          mpegversion: 4
          framed: true
          stream-format: raw
          codec_data: 0x7fffe0014500 [GstBuffer]
          level: 2
          base-profile: lc
          profile: lc
          channels: 2
          rate: 44100
        segment: time
          rate: 1
        tag: global
          container-format: Matroska
        tag: stream
          audio-codec: MPEG-4 AAC audio
          language-code: en
    }
}

Looks like an earlier version of the .c file.

This follows what git and systemd tools would do.

Fixes #341

Fixes flaky appsrc unit test where depending on scheduling
the submitted list might not be writable if submitted via
an action signal from the application thread.

Fixes gst-plugins-base#522

Jonny Lamb authored Dec 14, 2018 and Olivier Crête committed Dec 18, 2018

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.

The documentation for WIN32 mktime indicates that for struct tm*
before January 1, 1970, that -1 is returned, and since mktime is timezone
dependent, the struct tm corresponding to 1:00, Jan. 1, 1970 might be failed.

See also
https://docs.microsoft.com/en-us/cpp/c-runtime-library/reference/mktime-mktime32-mktime64

Edward Hervey authored Dec 15, 2018 and Edward Hervey committed Dec 15, 2018

There was a dead assignment used outside of the bin/pipeline creation
which was confusing (and unused). Just move that variable to
where it is actually used.

(Note that that variable was not needed outside of that block since
the refactoring done in 2b33d331 )

Edward Hervey authored Dec 15, 2018 and Edward Hervey committed Dec 15, 2018

Those values are always set after before usage

Edward Hervey authored Dec 15, 2018 and Edward Hervey committed Dec 15, 2018

That assertion was accidentally removed in the refactoring done in
60de1f26

Edward Hervey authored Dec 15, 2018 and Edward Hervey committed Dec 15, 2018

readable is set just after before usage since 906bbd38

Otherwise it can easily happen that the pad is destroyed before the
thread disappears, as happened sometimes in the test_pad_probe_block_add_remove
test where joining of the thread was done *after* the pad was unreffed
and destroyed.

Fixes #339

Jonny Lamb authored Dec 11, 2018 and Nicolas Dufresne committed Dec 11, 2018

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.

Convenient helper setting a caps feature on all the structures of a
caps.

gst: fixed the install command for gdb python macros on macos - `install -D` is not supported by BSD install

Dardo D Kleiner authored Dec 01, 2018 and Sebastian Dröge committed Dec 05, 2018

Existing test for iterating/removing buffer meta data was insufficient
to detect linked list corruption when removing multiple items, and could
also suffer from such corruption in attempting to count remaining items.
Modified the one test and added several others to exercise multiple
scenarios.

Validates fix for issue #332.

Dardo D Kleiner authored Dec 01, 2018 and Sebastian Dröge committed Dec 05, 2018

Fix corruption of meta list head when removing metas at the beginning
during iteration. Linked list handling in gst_buffer_foreach_meta
failed to track the previous entry and update the correct next pointer
when removing items from beyond the head of the list, resulting in
arbitrary list pointer corruption.

Closes #332

From cd1dee0 to 59cb678

This was no longer optional, leading to deadcode. This regression was
found trying to fix the unwind variant in cerbero.

They seemed incompatible with other colors.

Simple blue doesn't work on Linux console, which also happens to be a
gnome-terminal theme. Use bright-blue instead.

KimTaeSoo authored Nov 26, 2018 and Sebastian Dröge committed Nov 28, 2018

Before the previous commit, buffer pulling count and chain function call
counts are not equal due to EOS. After the modification, these counts
are equal so unit test is passing.

!33
#294

KimTaeSoo authored Nov 15, 2018 and Sebastian Dröge committed Nov 28, 2018

baseparse internally uses a 64kb buffer for pulling data from upstream.
If a 64kb pull is failing with a short read, it would previously pull
again the requested size.

Doing so is not only inefficient but also seems to cause problems with
some elements (rawvideoparse) where the second pull would fail with EOS.

Short reads are only allowed in GStreamer at EOS.

Closes #294

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.

This is required before we enabled an indent test in the CI.

gstreamer-project#33

Zeeshan Ali authored Nov 24, 2018 and Nirbheek Chauhan committed Nov 27, 2018

Not only this will make colored output work on old terminals and console
as well, terminals can theme the actual colors this way to make it fit
with their different themes this way.

This reverts commit 2faf93c0.

THis broke half our unit tests, oops:

https://ci.gstreamer.net/job/GStreamer-master/11203/testReport/

Sebastian Dröge authored Nov 13, 2018 and Mathieu Duponchelle committed Nov 26, 2018

Without this bindings get confused about the meaning of references, and
we really own these references if they are not already owned by
something else.

This change was originally part of 2cf16838 (gst-inspect: Colored
output) but got lost during the recent rebase.

Let's make the output a bit pretty to read. The colored output can be
disabled with `--no-colors` option or by setting `GST_INSPECT_NO_COLORS'
env (to any value).

The chosen colors are based on the popular Solarized theme, which is
targeted for both dark and light backgrounds.

Note:

* We only support true colors. If the terminal doesn't signal support for
  that via 'COLORTERM' env, we disable colored output.

* We don't add colors to --print-plugin-auto-install-info output, as
  that's meant for machines, not humans. Not only machines don't care
  about beauty, the existing ones will likely not expect colors and choke
  on it and we'll get angry mob at our doors.

[1] https://ethanschoonover.com/solarized

When printing info about a specific plugin, there is no need to prefix
some of the details with plugin's name. It's not only redundant but also
inconsistent and makes the task of adding consistent coloring to the
output (which we'll do in a follow patch), harder.

This emulates the default behaviour of git help pages, and also fixes
a bug on macOS where `less -F` doesn't display anything at all when
the output is shorter than one terminal screen.

Also moved the DEFAULT_PAGER define to after the includes, because
it's an unprefixed define.

Fixes gstreamer/gstreamer#330