Do not create a suspended thread so ResumeThread becomes unnecessary.
Busy loop over thread handle if by misfortune the thread ends before
_beginthreadex has returned.

Do not use SetThreadPriority on Windows Store.

Set the cancelled thread bit outside of the APC procedure.
APC is still used to wake up the cancelled thread when it is in an alertable state
(SleepEx or WaitForMultipleObjectsEx).
Since QueueUserAPC is forbidden on Windows Store we check the cancelled state every
50 ms.

Provide extradata also if the video is not palettized. All samples in tickets #4017, #5952 and #9546 plays nicely.
Signed-off-by: Rafaël Carré <funman@videolan.org>

Unbreak avformat-only builds.
Signed-off-by: Rafaël Carré <funman@videolan.org>

Fix decoding of MSRLE samples with avformat demuxer

Disable the API function returning version string as it uses old
automake functionality that might be buggy/not supported anymore.
Don't bother fixing properly as we don't use that function and libspeex is now
an abandoned projet (move to Opus!)

This patch copies the Qt interface implementation wrt file dropping, namely:
- first try it as a subtitle file if it is relevant
- second, try it as an item to be added to the playlist if not a subtitle

The previous code does not make sense as there is no unmute action.

Closes #9414

This reverts commit 7e2edc0d.

http://webm.googlecode.com/files/libvpx-v1.1.0.tar.bz2
 is a release tarball
http://webm.googlecode.com/files/libvpx-v1.2.0.tar.bz2
 is a daily updated git snapshot

libvpx has no release tarballs (or releases) anymore apparently.

AV_NOPTS_VALUE already contains a cast to int64_t

Ref #9022

Fixes #6476
Fixes #9442

As the library symbols are suffixed with the exact build version, and
it is not clear yet whether the API will be broken or simply extended,
watch for the exact version number (which is 1)

This patch changes the way seeking is done.

Previously it was a dumb stream size based seeking, and was
very approximative with multi-streams or non fixed bitrate
stream.

There was some code in oggseek to bisect search for frames,
which i previously linked to opus seeking, but it was
not reusable by all codecs as the ogg spec says granule to
absolute time is one way only.

New code does bisect search using absolute time only, and
then tries to sync to a specific keyframe (if codec has any),
backward or forward by checking packets.

Bisect and backward searchs are disabled for non FASTSEEK inputs.
In that case, it just behaves like any other player and syncs
to the next keyframe.

DirectShow/OggDS has also been fixed.

It is not clear which connections and which hardware are 10 bits able,
so let's stay on the safe side.

Fixes #9361

adding RecentsMRL::toPlaylist()
refactoring functions to eliminate repeating code
refactoring public DialogProvider::saveAPlaylist() to private DialogProvider::saveAPlaylist(playlist_t *p_playlist, playlist_item_t *p_node)
and making new DialogProvider::savePlayingToPlaylist() and DialogProvider::saveRecentsToPlaylist()

this commit includes fixes related to Jean-Baptiste Kempf code review.
 - fixing whitespace problem
 - RecentsMRL::toPlaylist null handling
Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

It fixes bug #9311 and resolves xml special chars on several other xml based playlist format (b4s, podcast, shoutcast, qtl).

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

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

Fix to actually conform to the range specified in doc.

First introduced in 9d20efbbSigned-off-by: Jean-Baptiste Kempf <jb@videolan.org>

To test this change, I used a MiniMaxwell which allows me to introduce
jitter, delay, and packet re-ordering in the video stream. I ran for
over 12 hours with no effect on the video and no increase in buffer
size. I'm reattaching the final patch here. I've made 2 basic changes
to the buffering strategy:

1) I now check to see if I and done buffering on both the
ES_OUT_SET_PCR and ES_OUT_SET_GROUP_PCR calls. This causes the buffered
amount to be checked more frequently and prevents over buffering which
I have seen in some instances
2) When ignoring the jitter, I do not flush the buffer and rebuffer,
since I am not increasing the buffer size anyway. Instead I just reset
the PCR and continue on.
Signed-off-by: Rémi Denis-Courmont <remi@remlab.net>