WebKit Bugzilla
New
Browse
Search+
Log In
×
Sign in with GitHub
or
Remember my login
Create Account
·
Forgot Password
Forgotten password account recovery
RESOLVED WONTFIX
51186
[chromium] Add asserts to test for contiguous-pixel Skia bitmaps.
https://bugs.webkit.org/show_bug.cgi?id=51186
Summary
[chromium] Add asserts to test for contiguous-pixel Skia bitmaps.
W. James MacLean
Reported
2010-12-16 07:26:44 PST
[chromium] Add asserts to test for contiguous-pixel Skia bitmaps.
Attachments
Patch
(4.63 KB, patch)
2010-12-16 07:29 PST
,
W. James MacLean
no flags
Details
Formatted Diff
Diff
Patch
(4.75 KB, patch)
2010-12-23 07:56 PST
,
W. James MacLean
no flags
Details
Formatted Diff
Diff
Patch
(4.74 KB, patch)
2010-12-23 08:26 PST
,
W. James MacLean
no flags
Details
Formatted Diff
Diff
Show Obsolete
(2)
View All
Add attachment
proposed patch, testcase, etc.
W. James MacLean
Comment 1
2010-12-16 07:29:37 PST
Created
attachment 76768
[details]
Patch
W. James MacLean
Comment 2
2010-12-16 07:33:12 PST
Since Skia bitmaps may have rowBytes() >= bytesPerPixel() * width(), and since some places in the code assume the equality above is enforced, add asserts() to detect when there are extra bytes at the end of each line where padding=0 is assumed.
W. James MacLean
Comment 3
2010-12-16 07:38:21 PST
James R, can you please review this? It's a very short (and hopefully straightforward) CL.
Kenneth Russell
Comment 4
2010-12-22 11:58:48 PST
Comment on
attachment 76768
[details]
Patch Looks OK.
WebKit Commit Bot
Comment 5
2010-12-22 14:03:39 PST
Comment on
attachment 76768
[details]
Patch Clearing flags on attachment: 76768 Committed
r74503
: <
http://trac.webkit.org/changeset/74503
>
WebKit Commit Bot
Comment 6
2010-12-22 14:03:46 PST
All reviewed patches have been landed. Closing bug.
W. James MacLean
Comment 7
2010-12-23 07:56:10 PST
Created
attachment 77333
[details]
Patch
W. James MacLean
Comment 8
2010-12-23 07:58:22 PST
Comment on
attachment 77333
[details]
Patch The patch failed on mac compile (I thought I had tested this, but apparently not). I have revised to correct, and this patch has been re-tested to make sure it compiles on Mac.
W. James MacLean
Comment 9
2010-12-23 07:59:47 PST
Patch reverted due to not compiling on mac.
Kenneth Russell
Comment 10
2010-12-23 08:01:53 PST
Comment on
attachment 77333
[details]
Patch View in context:
https://bugs.webkit.org/attachment.cgi?id=77333&action=review
The "if" with an ASSERT as its sole arm is sloppy.
> WebCore/platform/graphics/chromium/ImageLayerChromium.cpp:125 > + ASSERT(skiaBitmap->rowBytes() == SkBitmap::ComputeRowBytes(skiaConfig, skiaBitmap->width()));
ASSERT(!pixels || (skiaBitmap->rowBytes() == ...))
W. James MacLean
Comment 11
2010-12-23 08:26:21 PST
Created
attachment 77335
[details]
Patch
W. James MacLean
Comment 12
2010-12-23 08:27:20 PST
(In reply to
comment #10
)
> > ASSERT(!pixels || (skiaBitmap->rowBytes() == ...))
This is much better, thanks.
Kenneth Russell
Comment 13
2010-12-23 08:28:27 PST
Comment on
attachment 77335
[details]
Patch OK.
WebKit Commit Bot
Comment 14
2010-12-23 08:56:48 PST
Comment on
attachment 77335
[details]
Patch Clearing flags on attachment: 77335 Committed
r74561
: <
http://trac.webkit.org/changeset/74561
>
WebKit Commit Bot
Comment 15
2010-12-23 08:56:54 PST
All reviewed patches have been landed. Closing bug.
Kenneth Russell
Comment 16
2010-12-23 15:23:11 PST
These assertions seem to have broken the Chromium UI tests. See
https://bugs.webkit.org/show_bug.cgi?id=51565
, but in particular:
http://build.chromium.org/p/chromium/builders/Vista%20Tests%20(dbg)(2)/builds/3052
the failure of UITestCanLaunchWithOSMesa, and:
http://build.chromium.org/p/chromium/builders/Vista%20Tests%20(dbg)(2)/builds/3052/steps/ui_tests/logs/stdio
http://build.chromium.org/p/chromium/builders/Vista%20Tests%20(dbg)(2)/builds/3052/steps/Process%20Dumps/logs/stdio
Reopening. Please test this patch with this test on Windows before resubmitting.
Kenneth Russell
Comment 17
2010-12-23 15:23:40 PST
See above comment.
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