GitLab

Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

effectively reverting 351cd5a8

Close #16905
Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

The time tooltip no longer puts the window in front on Windows, when
compiling VLC with Qt5

Close #12084 #8532
Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

Some files do not have their first cluster at PTS=0, meaning that we
might spend an eternity looking for such PTS (even though it is not at
all present in the file) when starting playback (seeking to 0).

This patch fixes the issue by always assigning a seekpoint for PTS=0 pointing
at the first cluster position. Instead of using the cluster-timecode, a
hardcoded value of 0 is used (so that every file will have a seekpoint at that
location).
Signed-off-by: Thomas Guillem <thomas@gllm.fr>

Signed-off-by: Hugo Beauzée-Luyssen <hugo@beauzee.fr>

Signed-off-by: Hugo Beauzée-Luyssen <hugo@beauzee.fr>

vlc_thread_id is already defined in src/linux/thread.c

Each codecs has its own callback.

This was not happening because blocks are allocated with a pre-padding of 32
bits. But this behavior could change in function of the packetizer or demuxer.

Because we may have to realloc blocks inside this callback (corner case).

Therefore, this is not a quirk.

CSD and video size checks are now done in StartMediaCodec.

Call ParseVideoExtra from Open(). This was useless to call it from
StartMediaCodec since p_dec.fmt_in.p_extra couldn't change during the lifetime
of the module.

Additionnaly, ensure we pass "0" instead of an empty string for the root
object

As not doing so prevents MS Media Sharing servers from understanding the
request.

since defined values have changed

Restarting the audio output is not enough when a device is unplugged. Indeed,
the Start function will use the old (and invalid) device.
To fix this issue, re-select the default device in case of error.

Some devices expose more than one endpoint such as speakers and headphone
output. Disconnecting headphones does not remove the endpoint but instead marks
it as being in unplugged state.

Respond to the appropriate activation events in order to enable/disable
endpoints correctly when they are added/removed from set of "active" endpoints.

Fix #15966

Original patch by Sean McGovern <gseanmcg@gmail.com>

Original patch by Sean McGovern <gseanmcg@gmail.com>

The IF_EXTRACT_FMT macro is to be used in a if-else-tree, the usage of
IF_EXTRACT prior to the following if-statement did however cause the
single if-else-tree to be split up into two (causing the fallback-else
at the end to be unconditionally hit, even if another branch had already
handled the data).

If "TITLE=" metadata was found, we would still end up inside the
fallback since the previous implementation of IF_EXTRACT_FMT would start
a new if-else-tree.
Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

And set the image base to >4GB

This should improve the application security
Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

The functionality remains the same, though the code should be easier to
follow and maintain.
Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>