Bug 130665

Summary:

[GTK] [TextureMapper] Weird brightness with some videos with acceletared compositing

Product:

WebKit

Reporter:

Xabier Rodríguez Calvar <calvaris>

Component:

WebCore Misc.

Assignee:

Xabier Rodríguez Calvar <calvaris>

Status:

RESOLVED FIXED

Severity:

Normal

CC:

abucur, agomez, cgarcia, cmarcelo, commit-queue, eric.carlson, glenn, gustavo, jer.noble, koivisto, kondapallykalyan, luiz, menard, mrobinson, noam, philipj, pnormand, sergio, simon.fraser, vjaquez

Priority:

Version:

528+ (Nightly build)

Hardware:

Unspecified

OS:

Unspecified

URL:

http://people.igalia.com/xrcalvar/talks/20131022-GstConf-WebKit/index.html#4

Attachments:

Description	Flags
Example of the problem (left: correct, right: wrong)	none
Patch	none
Patch for landing	none
Patch for landing	none
Patch for landing	none

Description Xabier Rodríguez Calvar 2014-03-24 05:47:32 PDT

Created attachment 227639 [details]
Example of the problem (left: correct, right: wrong)

Visit http://people.igalia.com/xrcalvar/talks/20131022-GstConf-WebKit/index.html#4 with AC activated and deactivated.

When it deactivated the video is ok, when it isn't the video has 'an excess of brightness'.

Comment 1 Xabier Rodríguez Calvar 2014-03-25 09:59:26 PDT

The pipeline is the same so there must be a problem when dealing with the texture.

Comment 2 Xabier Rodríguez Calvar 2014-03-26 12:20:16 PDT

Created attachment 227874 [details]
Patch

Comment 3 Martin Robinson 2014-03-26 12:26:19 PDT

Comment on attachment 227874 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=227874&action=review

> Source/WebCore/ChangeLog:8
> +        No new tests needed, current set suffices.

They don't actually! You need to tell the truth here which is is that we do not yet have the ability to test accelerated composited contents on the bots.

> Source/WebCore/platform/graphics/texmap/TextureMapper.h:50
> +        None = 0x00,

None should probably be called NoFlag and you should use 0 here instead of 0x00.

Comment 4 Xabier Rodríguez Calvar 2014-03-26 14:13:52 PDT

(In reply to comment #3)
> (From update of attachment 227874 [details])
> View in context: https://bugs.webkit.org/attachment.cgi?id=227874&action=review
> 
> > Source/WebCore/ChangeLog:8
> > +        No new tests needed, current set suffices.
> 
> They don't actually! You need to tell the truth here which is is that we do not yet have the ability to test accelerated composited contents on the bots.

I should have been more verbose. The current test would perfectly suffice if we were able to activate accelerated composition in the bots.

> > Source/WebCore/platform/graphics/texmap/TextureMapper.h:50
> > +        None = 0x00,
> 
> None should probably be called NoFlag and you should use 0 here instead of 0x00.

Roger.

Comment 5 Xabier Rodríguez Calvar 2014-03-28 04:04:17 PDT

Created attachment 228037 [details]
Patch for landing

Comment 6 WebKit Commit Bot 2014-03-28 04:05:28 PDT

Comment on attachment 228037 [details]
Patch for landing

Rejecting attachment 228037 [details] from commit-queue.

Failed to run "['/Volumes/Data/EWS/WebKit/Tools/Scripts/webkit-patch', '--status-host=webkit-queues.appspot.com', '--bot-id=webkit-cq-01', 'validate-changelog', '--check-oops', '--non-interactive', 228037, '--port=mac']" exit_code: 1 cwd: /Volumes/Data/EWS/WebKit

mrobinson@webkit.org found in /Volumes/Data/EWS/WebKit/Source/WebCore/ChangeLog does not appear to be a valid reviewer according to contributors.json.
/Volumes/Data/EWS/WebKit/Source/WebCore/ChangeLog neither lists a valid reviewer nor contains the string "Unreviewed" or "Rubber stamp" (case insensitive).

Full output: http://webkit-queues.appspot.com/results/4775424450101248

Comment 7 Xabier Rodríguez Calvar 2014-03-28 04:35:12 PDT

Created attachment 228039 [details]
Patch for landing

Comment 8 WebKit Commit Bot 2014-03-28 04:36:28 PDT

Comment on attachment 228039 [details]
Patch for landing

Rejecting attachment 228039 [details] from commit-queue.

Failed to run "['/Volumes/Data/EWS/WebKit/Tools/Scripts/webkit-patch', '--status-host=webkit-queues.appspot.com', '--bot-id=webkit-cq-02', 'validate-changelog', '--check-oops', '--non-interactive', 228039, '--port=mac']" exit_code: 1 cwd: /Volumes/Data/EWS/WebKit

mrobinson@webkit.org found in /Volumes/Data/EWS/WebKit/Source/WebCore/ChangeLog does not appear to be a valid reviewer according to contributors.json.
/Volumes/Data/EWS/WebKit/Source/WebCore/ChangeLog neither lists a valid reviewer nor contains the string "Unreviewed" or "Rubber stamp" (case insensitive).

Full output: http://webkit-queues.appspot.com/results/5825698572795904

Comment 9 Xabier Rodríguez Calvar 2014-03-28 04:38:44 PDT

Created attachment 228040 [details]
Patch for landing

Comment 10 Xabier Rodríguez Calvar 2014-03-28 05:09:27 PDT

Comment on attachment 228040 [details]
Patch for landing

Let's avoid a land patch, go home. I can this when I have time to look at the bots.

Comment 11 WebKit Commit Bot 2014-03-30 23:40:16 PDT

Comment on attachment 228040 [details]
Patch for landing

Clearing flags on attachment: 228040

Committed r166480: <http://trac.webkit.org/changeset/166480>

Comment 12 WebKit Commit Bot 2014-03-30 23:40:23 PDT

All reviewed patches have been landed.  Closing bug.