Bug 44195

Summary: [chromium] TilingData::tilePositionY has typo of X where Y should be used
Product: WebKit Reporter: Vincent Scheib <scheib>
Component: Layout and RenderingAssignee: Nobody <webkit-unassigned>
Status: RESOLVED FIXED    
Severity: Normal CC: amarinichev, commit-queue, levin
Priority: P2    
Version: 528+ (Nightly build)   
Hardware: Other   
OS: All   
Attachments:
Description Flags
Patch
none
Patch none

Vincent Scheib
Reported 2010-08-18 12:18:37 PDT
[chromium] TilingData::tilePositionY has typo of X where Y should be used
Attachments
Patch (1.14 KB, patch)
2010-08-18 12:19 PDT, Vincent Scheib 		no flags
DetailsFormatted DiffDiff
Patch (8.53 KB, patch)
2010-08-19 17:31 PDT, Vincent Scheib 		no flags
DetailsFormatted DiffDiff
Show Obsolete (1) View All Add attachment proposed patch, testcase, etc.
Vincent Scheib
Comment 1 2010-08-18 12:19:30 PDT
Created attachment 64753 [details] Patch
Vincent Scheib
Comment 2 2010-08-18 12:23:46 PDT
Single character typo from X to Y.
David Levin
Comment 3 2010-08-18 12:25:57 PDT
Comment on attachment 64753 [details] Patch Please either add a layout test to catch this OR Mention which layout test covers it (in the ChangeLog) OR Mention why it is impossible to add a layout test for this -- hopefully not -- (in the ChangeLog).
David Levin
Comment 4 2010-08-18 12:27:58 PDT
btw, the recursive nature of this call seems odd to me when a simple loop would avoid it.
Vincent Scheib
Comment 5 2010-08-18 14:36:10 PDT
Layout tests are not yet possible for this code for a few reasons: - TilingData is currently only used behind a compile time flag as we develop GPU accelerated support for Canvas 2D. - TilingData is driven by a system GPU's maximum supported texture size, which varies on different hardware. We will need to plumb testing code that enables a layout test to supply a false value for GLES2Texture to mock as the maximium texture size at test time only.
David Levin
Comment 6 2010-08-19 08:34:40 PDT
(In reply to comment #5) > Layout tests are not yet possible for this code for a few reasons: > - TilingData is currently only used behind a compile time flag as we develop GPU accelerated support for Canvas 2D. Yes but a layout test could be written that would hit this code when the flag is flipped. > - TilingData is driven by a system GPU's maximum supported texture size, which varies on different hardware. We will need to plumb testing code that enables a layout test to supply a false value for GLES2Texture to mock as the maximium texture size at test time only. Is there a current typically upper bound? How hard would it be to hit that (or some large multiple of that) for the layout test? Lastly, why is this method recursive when a loop could easily do the same thing?
Vincent Scheib
Comment 7 2010-08-19 17:31:10 PDT
Created attachment 64918 [details] Patch
WebKit Commit Bot
Comment 8 2010-08-19 21:52:26 PDT
Comment on attachment 64918 [details] Patch Clearing flags on attachment: 64918 Committed r65726: <http://trac.webkit.org/changeset/65726>
WebKit Commit Bot
Comment 9 2010-08-19 21:52:31 PDT
All reviewed patches have been landed. Closing bug.
Note You need to log in before you can comment on or make changes to this bug.