Clobbering these tables will temporarily clobber the template used
as a basis for other threads to start decoding from. If the other
decoding thread updates from the template right at that moment,
subsequent threads will get invalid (or, usually, none at all) mmco
tables. This leads to invalid reference lists and subsequent decode
failures.
Therefore, instead, decode the mmco tables only for the first slice in
a field or frame. For other slices, decode the bits and ensure they
are identical to the mmco tables in the first slice, but don't ever
clobber the context state. This prevents other threads from using a
clobbered/invalid template as starting point for decoding, and thus
fixes decoding in these cases.
This fixes occasional (~1%) failures of h264-conformance-mr1_bt_a with
frame-multithreading enabled.
(cherry picked from commit bad446e251405dc250c3cbee199072e083a1e4b9)
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
Attempting to re-parse the headers at demuxer level is a
pandora box the way its done currently.
This allows full reconfiguration of vorbis streams
Fixes Ticket2117
Fixes Ticket2121
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
(cherry picked from commit c994bb2fb7727f5c20ef71cc6727fb219fd49d60)
Previously we re synced from where we where which cam lead
to loosing pages.
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
(cherry picked from commit c5cf58d4b9b04cee1487a3095b83300791c21f15)
By default libav sets it to 3/4 while x264 sets it to 9/10.
CC: libav-stable@libav.org
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
(cherry picked from commit 47812070a267cbdf74164e154d03d99bf8ced100)
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
rc_buffer_size is not set before.
Solve the initial the rate control underflow issue reported in
bug 222.
CC: libav-stable@libav.org
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
(cherry picked from commit bff3607547fdbb6e32b3830a351e6a33280c1e0d)
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
Accept the buffer size in bytes and check for overflow before passing
the value in bits to init_get_bits.
(cherry picked from commit e28ac6e5e27e64a206e399e958481c1e6f992189)
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
Fix an undefined behaviour and make the function return a proper
error in case of overflow.
CC: libav-stable@libav.org
(cherry picked from commit d9cf5f516974c64e01846ca685301014b38cf224)
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
Prevent the crash on fuzzed files as reported in bug 63.
(cherry picked from commit c1d1ef4ecd9c4f1ca01c8149c7e57c14968ca588)
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
Handle pred_flag parameter not given to get_mvdata_interlaced()
Signed-off-by: Vladimir Pantelic <vladoman@gmail.com>
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
(cherry picked from commit 7b8c5b263bc680eff5710bee5994de39d47fc15e)
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
Regression since: ea6da80
Fixes Ticket2123
I cannot reproduce any regressions by flipping the wrong condition
to how it should have been.
Thanks-to: ubitux
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
(cherry picked from commit 57bdd67646cfffa2921a8b28bb5f88cfe5c0989e)
Sometimes the muxer modifies the packet, like for instance lavf/mp3enc
changing pkt->destruct in order to keep a copy. These changes must be
kept, even though the muxer behaviour is questionable. Regression since
0072116.
Fixes#2124.
(cherry picked from commit 119d70db5099f9513d954283245efd7f699ad321)
The filter has been added recently and does not yet pass tests on all platforms.
It should not be included in the tests in the release until it works.
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
We dont include mathematics.h as the headers get configured by config.h
for the target not the host.
Reviewed-by: Clément Bœsch <ubitux@gmail.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
