WebKit Bugzilla
New
Browse
Log In
×
Sign in with GitHub
or
Remember my login
Create Account
·
Forgot Password
Forgotten password account recovery
NEW
141924
Use accelerated drawing to generate snapshots.
https://bugs.webkit.org/show_bug.cgi?id=141924
Summary
Use accelerated drawing to generate snapshots.
Jeremy Jones
Reported
2015-02-23 14:04:52 PST
Use accelerated drawing to generate snapshots.
Attachments
Patch
(3.94 KB, patch)
2015-02-23 14:15 PST
,
Jeremy Jones
beidson
: review-
Details
Formatted Diff
Diff
View All
Add attachment
proposed patch, testcase, etc.
Jeremy Jones
Comment 1
2015-02-23 14:15:23 PST
Created
attachment 247147
[details]
Patch
Jeremy Jones
Comment 2
2015-02-23 17:42:14 PST
rdar://problem/18072471
Tim Horton
Comment 3
2015-02-23 17:53:00 PST
Comment on
attachment 247147
[details]
Patch View in context:
https://bugs.webkit.org/attachment.cgi?id=247147&action=review
> Source/WebKit2/WebProcess/WebPage/WebPage.cpp:1678 > +
No trailing whitespace.
> Source/WebKit2/WebProcess/WebPage/WebPage.cpp:1691 > + FloatPoint destination(0, 0);
No need for the 0,0. Or the temporary; this can just be an inline FloatPoint() on the next line if you want.
> Source/WebKit2/WebProcess/WebPage/WebPage.cpp:1692 > + snapshotGraphicsContext->drawImageBuffer(imageBuffer.get(), ColorSpaceDeviceRGB, destination);
We should have a plan to go further and return accelerated image buffers (in the form of IOSurfaces probably) to clients, so that we can avoid this readback. But, I bet you've already learned that painting and then doing the readback is faster in a lot of important cases, and that's why you've posted this patch? Maybe we can make a parallel API (that hands back surfaces and paints acceleratedly) and have people start switching to it? Sadly for very simple cases this will reduce performance (because the win from accelerating painting will not be enough to offset the loss of the readback), so we should probably do that sooner rather than later.
Jon Lee
Comment 4
2015-02-24 13:12:09 PST
Is there a way we could restrict this snapshotting path to only when a video is presented fullscreen?
Brady Eidson
Comment 5
2017-04-24 19:05:27 PDT
Comment on
attachment 247147
[details]
Patch This patch has been pending review since 2015 with no recent activity. It seems unlikely that it would even still apply to trunk in its current form. Clearing from the review queue. Feel free to update and resubmit if the patch is still relevant.
Note
You need to
log in
before you can comment on or make changes to this bug.
Top of Page
Format For Printing
XML
Clone This Bug