Commits · 7d2a6826912b629f17f3b0c5268e334fa123c4c9 · Linshizhi / ffmpeg.wasm-core

31 May, 2015 33 commits
- huffyuv: Check memory allocation · 7d2a6826
  Vittorio Giovara authored May 31, 2015
  
  7d2a6826
- eatgv: Check memory allocation · 7fccc96d
  Vittorio Giovara authored May 31, 2015
  
  7fccc96d
- dct: Check memory allocation · 69277069
  Vittorio Giovara authored May 31, 2015
  
  69277069
- asv: Check memory allocation · e524f373
  Vittorio Giovara authored May 31, 2015
  
  e524f373
- aacpsy: Check memory allocation · 074a1b37
  Vittorio Giovara authored May 31, 2015
  
  074a1b37
- mpegvideo: Move MotionEstContext and function declarations to a separate header · 149fa0b7
  Vittorio Giovara authored May 29, 2015
  
  149fa0b7
- mpegvideo: mpeg12: Move function declarations · b2b76691
  Vittorio Giovara authored May 29, 2015
  
  b2b76691
- mpegvideo: Move tables to a separate file · 378a0008
  Vittorio Giovara authored May 29, 2015
  
  378a0008
- mpegvideo: msmpeg4: Move function declarations · 31a117a0
  Vittorio Giovara authored May 29, 2015
  
  31a117a0
- mpegvideo: wmv2: Move function declarations · 2f15846a
  Vittorio Giovara authored May 29, 2015
  
  2f15846a
- dxva2_hevc: re-write reference frame handling · 902a55f7
  Hendrik Leppkes authored May 28, 2015
```
The old logic required an explicit clearing of the lists first and was
prone to overflow the DXVA2 struct in some circumstances.
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
```
  902a55f7
- dxva2_hevc: fix 32x32 scaling lists · 35818b8a
  Hendrik Leppkes authored May 28, 2015
```
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
```
  35818b8a
- mov: always check avio_read return value · 5c720657
  Andreas Cadhalpun authored May 26, 2015
```
If avio_read fails, the buffer can contain uninitialized data.

This fixes 'Conditional jump or move depends on uninitialised value(s)'
valgrind warnings, and addresses a few memleaks.
Signed-off-by: Andreas Cadhalpun <Andreas.Cadhalpun@googlemail.com>
Signed-off-by: Vittorio Giovara <vittorio.giovara@gmail.com>
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
```
  5c720657
- avio: Add avio_read wrapper to simplify error checking · bff0349d
  Vittorio Giovara authored May 26, 2015
```
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
```
  bff0349d
- mov: reject zero bytes_per_frame with non-zero samples_per_frame · cffb9ea8
  Andreas Cadhalpun authored May 26, 2015
```
In this case the mov demuxer can return a large number of empty packets.
Signed-off-by: Andreas Cadhalpun <Andreas.Cadhalpun@googlemail.com>
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
```
  cffb9ea8
- mov: abort on EOF in ff_mov_read_chan · 8e9c39e8
  Andreas Cadhalpun authored May 26, 2015
```
Otherwise the loop can take a lot of time if num_descr is very large.
Signed-off-by: Andreas Cadhalpun <Andreas.Cadhalpun@googlemail.com>
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
```
  8e9c39e8
- mov: Correctly allocate ctts_data · a8a90906
  Luca Barbato authored May 31, 2015
```
It can be reallocated.

CC: libav-stable@libav.org
```
  a8a90906
- mov: Fix two memleaks · f261a55d
  Andreas Cadhalpun authored May 26, 2015
```
Signed-off-by: Andreas Cadhalpun <Andreas.Cadhalpun@googlemail.com>
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>

CC: libav-stable@libav.org
```
  f261a55d
- vp9: Parse subsampling and report missing feature · 3f38d4b8
  Vittorio Giovara authored May 27, 2015
  
  3f38d4b8
- x86: check for AV_CPU_FLAG_AVXSLOW where useful · d68c0538
  James Almer authored May 26, 2015
```
Signed-off-by: James Almer <jamrial@gmail.com>
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
```
  d68c0538
- x86: Add helper macros to check for slow cpuflags · cae39851
  James Almer authored May 26, 2015
```
Signed-off-by: James Almer <jamrial@gmail.com>
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
```
  cae39851
- x86: add AV_CPU_FLAG_AVXSLOW flag · f7cafb5d
  James Almer authored May 26, 2015
```
Signed-off-by: James Almer <jamrial@gmail.com>
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
```
  f7cafb5d
- ppc: vsx: Implement diff_pixels and get_pixels · d0bf20a4
  Luca Barbato authored May 12, 2015
```
Use a macro to abstract the endianness.
```
  d0bf20a4
- ppc: vsx: Implement float_dsp · eecd29b3
  Luca Barbato authored May 11, 2015
  
  eecd29b3
- ppc: cpu: Add support for VSX and POWER8 extensions · 7d07ee5a
  Luca Barbato authored May 10, 2015
  
  7d07ee5a
- ppc: Restrict some Altivec implementations to Big Endian · da60b99a
  Luca Barbato authored May 10, 2015
```
In Little Endian the vec_ld/vec_st operations work as
expected only for byte-vectors.
```
  da60b99a
- ppc: configure: Support ISA 2.06 and later · f22ebd25
  Luca Barbato authored May 10, 2015
```
POWER 7 and POWER 8 support VSX and ldbrx.

POWER 8 supports additional extended VSX instructions introduced
with ISA 2.07.
```
  f22ebd25
- ppc: Clarify and extend the cpuid check · 3058872c
  Luca Barbato authored May 09, 2015
```
Add POWER entries.
```
  3058872c
- ppc: linux: Check altivec using the auxv · baa94563
  Luca Barbato authored Apr 03, 2015
```
Should prevent trying to use altivec when it is disabled by the kernel.
```
  baa94563
- ppc: pixblockdsp: Use the abriged vector types · 7014b659
  Luca Barbato authored May 11, 2015
  
  7014b659
- ppc: avutil: Use the abriged vector types · 72cebae0
  Luca Barbato authored May 10, 2015
  
  72cebae0
- ppc: avutil: Drop a potentially dangerous workaround · 254eb5b6
  Luca Barbato authored May 10, 2015
```
The compiler is free to optimize such expressions in any sort of way.
```
  254eb5b6
- network: prevent SIGPIPE on OSX · 881b80b3
  wm4 authored May 28, 2015
```
OSX does not know MSG_NOSIGNAL. BSD (which OSX is based on) has got
the socket option SO_NOSIGPIPE (even if modern BSDs also support
MSG_NOSIGNAL).
Signed-off-by: Martin Storsjö <martin@martin.st>
```
  881b80b3
30 May, 2015 2 commits
- h264: Make sure reinit failures mark the context as not initialized · 0181ae9a
  Luca Barbato authored May 25, 2015
```
Bug-Id: CVE-2015-3417
CC: libav-stable@libav.org
```
  0181ae9a
- msrle: Use FFABS to determine the frame size in msrle_decode_pal4 · 5ecabd3c
  Luca Barbato authored May 25, 2015
```
As done in msrle_decode_8_16_24_32.

Bug-Id: CVE-2015-3395
CC: libav-stable@libav.org
```
  5ecabd3c
29 May, 2015 5 commits
- rtmpdh: Generate the whole private exponent using av_get_random_seed() with nettle/gmp · b2f0f37d
  Martin Storsjö authored May 29, 2015
```
Don't use a PRNG for generating it; that defies the intended use
within the cryptograhic handshake.
Signed-off-by: Martin Storsjö <martin@martin.st>
```
  b2f0f37d
- rtmpdh: Create sufficiently long private keys for gcrypt/nettle · e9e86d9e
  Martin Storsjö authored May 28, 2015
```
There was a misunderstanding betewen bits and bytes for the parameter
value for generating random big numbers.
Signed-off-by: Martin Storsjö <martin@martin.st>
```
  e9e86d9e
- rtmpdh: Remove an unnecessary check in the gcrypt/nettle dh_compute_key · 8016a1bd
  Martin Storsjö authored May 28, 2015
```
Signed-off-by: Martin Storsjö <martin@martin.st>
```
  8016a1bd
- rtmpdh: Add fate test for the DH handshake routine · 063f7467
  Martin Storsjö authored May 28, 2015
```
This helps if these functions need to be implemented using another
crypto API.
Signed-off-by: Martin Storsjö <martin@martin.st>
```
  063f7467
- rtmpdh: Pass the actual buffer size of the output secret key · 0508faaa
  Martin Storsjö authored May 28, 2015
```
Signed-off-by: Martin Storsjö <martin@martin.st>
```
  0508faaa