zqz/platform-external-webrtc
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Added EncodedImage::GetBufferPaddingBytes.

Browse Source 
The FFmpeg video decoder requires up to 8 additional bytes to be allocated for its encoded image buffer input, due to optimized byte readers over-reading on some platforms.
We plan to use FFmpeg for a soon-to-land H.264 enc/dec.

This CL adds support for padding encoded image buffers based on codec type, and makes sure calls to VCMEncodedFrame::VerifyAndAllocate use the padding.

All padding constants are 0 but making H.264 pad with 8 bytes will be a one-line change.

Also, added -framework CoreFoundation to webrtc_h264_video_toolbox which was missing.

BUG=chromium:468365
BUG=https://bugs.chromium.org/p/webrtc/issues/detail?id=5424
NOTRY=True

Review URL: https://codereview.webrtc.org/1602523004

Cr-Commit-Position: refs/heads/master@{#11337}
This commit is contained in:
hbos
2016-01-21 05:43:11 -08:00
committed by Commit bot
parent 429c345b02
commit d664836efa
5 changed files with 35 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
webrtc/modules/video_coding/frame_buffer.cc
View File

@ -105,7 +105,8 @@ VCMFrameBufferEnum VCMFrameBuffer::InsertPacket(
uint32_t requiredSizeBytes =
Length() + packet.sizeBytes +
(packet.insertStartCode ? kH264StartCodeLengthBytes : 0);
(packet.insertStartCode ? kH264StartCodeLengthBytes : 0) +
EncodedImage::GetBufferPaddingBytes(packet.codec);
if (requiredSizeBytes >= _size) {
const uint8_t* prevBuffer = _buffer;
const uint32_t increments =