Chromium Code Reviews
chromiumcodereview-hr@appspot.gserviceaccount.com (chromiumcodereview-hr) | Please choose your nickname with Settings | Help | Chromium Project | Gerrit Changes | Sign out
(114)

Issue 25758003: Exynos decoder: workaround for wrong MFC output format. (Closed)

Created:
7 years, 2 months ago by hshi1
Modified:
7 years, 2 months ago
CC:
chromium-reviews, jam, apatrick_chromium, joi+watch-content_chromium.org, feature-media-reviews_chromium.org, darin-cc_chromium.org, piman+watch_chromium.org
Visibility:
Public.

Description

Exynos decoder: workaround for wrong MFC output format. EVDA currently uses NV12MT_16X16 (tiled) format for MFC output, while EVEA uses NV12M (non-tiled) format. There appears to be a MFC driver or firmware bug when VP8 decode and H264 encode run concurrently. Workaround is to force EVDA to use non-tiled format (NV12M) so that MFC output format is the same for both EVEA and EVDA. BUG=302870 TEST=verify corruption is fixed when tab casting VP8 playback, AND hardware decoder is used. R=scherkus@chromium.org Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=226912

Patch Set 1 #

Patch Set 2 : Add TODO and reference bug. #

Patch Set 3 : Use NV12M instead of NV12MT. #

Unified diffs Side-by-side diffs Delta from patch set Stats (+4 lines, -2 lines) Patch
M content/common/gpu/media/exynos_video_decode_accelerator.cc View 1 2 2 chunks +4 lines, -2 lines 0 comments Download

Messages

Total messages: 13 (0 generated)
hshi1
PTAL. This is a temporary workaround for the VP8 corruption problem when H264 encode runs ...
7 years, 2 months ago (2013-10-02 20:23:50 UTC) #1
scherkus (not reviewing)
lgtm sheu/posciak: any comments?
7 years, 2 months ago (2013-10-02 23:53:47 UTC) #2
hshi1
sheu/posciak - per offline discussions this CL should be ok to you. I will try ...
7 years, 2 months ago (2013-10-03 01:24:40 UTC) #3
commit-bot: I haz the power
CQ is trying da patch. Follow status at https://chromium-status.appspot.com/cq/hshi@chromium.org/25758003/3001
7 years, 2 months ago (2013-10-03 02:59:58 UTC) #4
commit-bot: I haz the power
Step "update" is always a major failure. Look at the try server FAQ for more ...
7 years, 2 months ago (2013-10-03 04:38:50 UTC) #5
commit-bot: I haz the power
CQ is trying da patch. Follow status at https://chromium-status.appspot.com/cq/hshi@chromium.org/25758003/3001
7 years, 2 months ago (2013-10-03 04:49:08 UTC) #6
commit-bot: I haz the power
CQ is trying da patch. Follow status at https://chromium-status.appspot.com/cq/hshi@chromium.org/25758003/3001
7 years, 2 months ago (2013-10-03 05:17:06 UTC) #7
Pawel Osciak
On 2013/10/03 05:17:06, I haz the power (commit-bot) wrote: > CQ is trying da patch. ...
7 years, 2 months ago (2013-10-03 07:07:15 UTC) #8
hshi1
On 2013/10/03 07:07:15, posciak wrote: > On 2013/10/03 05:17:06, I haz the power (commit-bot) wrote: ...
7 years, 2 months ago (2013-10-03 16:32:50 UTC) #9
hshi1
posciak/sheu - Uploaded PS#3. I believe it is still a valid workaround to force MFC ...
7 years, 2 months ago (2013-10-03 16:59:37 UTC) #10
hshi1
I do not have a good way to measure per-frame decode time (that is actually ...
7 years, 2 months ago (2013-10-03 17:55:28 UTC) #11
commit-bot: I haz the power
CQ is trying da patch. Follow status at https://chromium-status.appspot.com/cq/hshi@chromium.org/25758003/56001
7 years, 2 months ago (2013-10-03 21:23:20 UTC) #12
hshi1
7 years, 2 months ago (2013-10-04 00:41:32 UTC) #13
Message was sent while issue was closed.
Committed patchset #3 manually as r226912 (presubmit successful).

Powered by Google App Engine
This is Rietveld 408576698