Select Git revision
Search by author
- Any Author
- authors
-
Administrator root
-
Alaric Senat
asenat
-
Alexandre Janniaux alexandre-janniaux
-
Bart - The Review Bot
Bart
-
Benjamin Arnaud
bunjee
-
Christophe Massiot cmassiot
-
Claudio Cambra
claucambra
-
David dfuhrmann
-
Denis Charmet
typx
-
Fatih Uzunoğlu fuzun
-
Felix Paul Kühne
fkuehne
-
François Cartegnie
fcartegnie
-
Gabriel Lafond-Thenaille gabriel_lt
-
**** project_435_bot
-
**** project_435_bot_867220ea9cbb40f38051abad33491375
-
Ilkka Ollakka
ileoo
-
Jean-Baptiste Kempf
jbk
-
Jean-Paul Saman
jpsaman
-
Johannes Kauffmann
JohannesKauffmann
-
Konstantin Pavlov
thresh
- Feb 18, 2025
-
-
Steve Lhomme authored
-
Steve Lhomme authored
It requires loading nvEncodeAPI[64].dll from the system which is not possible in UWP.
-
OpenBSD and NetBSD do not have qsort_r().
-
This allows checking that the dependency could be found, in particular to set the enabled: flags for plugins, without first checking whether the condition to do the lookup are met.
-
The split from 91eaf42c was missing the static qualifier for the functions, leading to warnings.
-
Steve Lhomme authored
*pi_preroll is a vlc_tick_t that can never be bigger than VLC_TICK_MAX.
-
-
- Feb 17, 2025
-
-
This actually disable the tests when a module the test depends on is being disabled, and display a warning listing the modules that were missing for the test.
-
The module manifest is indexed by the name and allows checking whether a given module has been declared or not. In particular, it's used by the tests to check whether the module was enabled or not. For this feature to work, it should always be stored and not just when the module is actually enabled.
-
Without commit "input: es_out: fix non-matching main_clock bus", the test is failing with the following assertion: core debug: The input can't pace, selecting the input (PCR) as the clock source Assertion failed: (vlc_mutex_held(&main_clock->lock)), function vlc_clock_main_CreateInputMaster, file clock.c, line 998. FAIL test_input_es_out (exit status: 134) Co-authored-by:Sergio Ammirata <sergio@ammirata.net> Reported that some US feeds with mpegts were triggering the assertion on vlc_clock_main_CreateInputMaster() call.
-
Fix the following assertion in some cases multiple programs are used by the es_out, leading to a mismatch between the main_clock bus being locked and the main_clock bus creating the input vlc_clock. Assertion failed: (vlc_mutex_held(&main_clock->lock)), function vlc_clock_main_CreateInputMaster, file clock.c, line 998. FAIL test_input_es_out (exit status: 134) Co-authored-by:
-
- Feb 16, 2025
-
-
-
The gc might not release the SurfaceTexture from its destructor.
-
-
-
-
This fixes closing the window using the CSD close button crashing the application. It should not be necessary to delete the window because it has a parent window and the parent window already gets deleted.
-
-
-
Currently, if Qt's event loop is not fast enough to consume the high resolution events, the events may backlog.
-
-
-
-
-
-
-
-
The current value yields spurious failure on CI server under heavy load fix #29029
-
Check usable header "netinet/tcp.h" : NO Check usable header "sys/mount.h" : NO The headers in question need sys/types.h included as well.
-
- Feb 15, 2025
-
-
We don't want to build on darwin without X11 and android for instance, but it would be ok to build on darwin with X11.
-
Library X11 found: NO and as expected X11_SKINS was not defined. ../modules/gui/skins2/src/os_factory.cpp:46:2: error: "No OSFactory implementation !" ^ 1 error generated. There is already an X11 check and it works as expected so use that for the dependency checking.
-
N.B. The tarball structure as changed, the top-level directory is now amf-headers-v1.4.36 (rather than AMF). Since all of robux's patches have been upstreamed we can now drop them.
-
Thomas Guillem authored
Work-arround oss-fuzz that can't accept scripts like the following: ``` export VLC_TARGET=$target exec ./vlc-demux-dec-libfuzzer "\$@" ``` Therefore, append the target to the binary name. cf. https://github.com/google/oss-fuzz/issues/1462#issuecomment-393362400
-
Avoid using cdata, which will trigger a full rebuild on enabling/disabling.
-
The contrib include path was used to specify the include path to the libavcodec headers, but of course it doesn't work when libavcodec is not compiled from the contribs. Originally added with the plugin definition in 07d15477.
-
It only has C files, so there's no need for cpp flags.
-
Avoid using cdata, which will trigger a full rebuild on enabling/disabling.
-
The contrib include path was used to specify the include path to the libavcodec headers, but of course it doesn't work when libavcodec is not compiled from the contribs. Originally added with the plugin definition in 07d15477.
-
This reverts commit 4cd9bdcb. Modifying cdata requires regenerating config.h which is included by every single source file in the project. There's no point in reconfiguring and recompiling the whole project to enable or disable a library for a module just for the sake of matching what autotool buildsystem did.
-
OpenBSD was the latest OS to fix if.h being built standlone and we fixed this 7 yrs ago.
-
