[chromium] TilingData::tilePositionY has typo of X where Y should be used
Created attachment 64753 [details] Patch
Single character typo from X to Y.
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).
btw, the recursive nature of this call seems odd to me when a simple loop would avoid it.
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.
(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?
Created attachment 64918 [details] Patch
Comment on attachment 64918 [details] Patch Clearing flags on attachment: 64918 Committed r65726: <http://trac.webkit.org/changeset/65726>
All reviewed patches have been landed. Closing bug.