RESOLVED FIXED 108283
REGRESSION (r139282): Caret repainting is broken for text-align: center'd <input> 
https://bugs.webkit.org/show_bug.cgi?id=108283
Summary REGRESSION (r139282): Caret repainting is broken for text-align: center'd <in...
Tim Horton
Reported 2013-01-29 18:54:13 PST
See attached testcase; the caret is not fully erased when blinking. Reproduces in Safari on Mountain Lion. I have a hunch that this is due to a combination of r139282 (confirmed by bisection) and subpixel layout (entirely speculative). <rdar://problem/13062683>
Attachments
test (265 bytes, text/html)
2013-01-29 18:54 PST, Tim Horton 		no flags
Details
off caret (7.14 KB, image/png)
2013-01-29 18:54 PST, Tim Horton 		no flags
Details
on caret (6.83 KB, image/png)
2013-01-29 18:54 PST, Tim Horton 		no flags
Details
Patch (3.97 KB, patch)
2013-01-30 00:24 PST, Tien-Ren Chen 		no flags
DetailsFormatted DiffDiff
Patch (4.98 KB, patch)
2013-01-30 02:32 PST, Tien-Ren Chen 		no flags
DetailsFormatted DiffDiff
Patch (4.90 KB, patch)
2013-01-30 02:38 PST, Tien-Ren Chen 		no flags
DetailsFormatted DiffDiff
Show Obsolete (2) View All Add attachment proposed patch, testcase, etc.
Tim Horton
Comment 1 2013-01-29 18:54:28 PST
Created attachment 185375 [details] test
Tim Horton
Comment 2 2013-01-29 18:54:40 PST
Created attachment 185376 [details] off caret
Tim Horton
Comment 3 2013-01-29 18:54:52 PST
Created attachment 185377 [details] on caret
Alexandre Elias
Comment 4 2013-01-29 20:32:39 PST
Also reproducing on Mac Chromium 25.0.1364.45, looks like cross-platform issue.
Tien-Ren Chen
Comment 5 2013-01-29 23:54:50 PST
Confirmed on chromium-linux 25.0.1364.45 too. Looks like we still need the 1-pixel padding workaround. I'm uploading a patch now.
Tien-Ren Chen
Comment 6 2013-01-30 00:24:53 PST
Created attachment 185414 [details] Patch
Tien-Ren Chen
Comment 7 2013-01-30 00:26:32 PST
Don't have access to a Mac right now. Waiting for Mac EWS bot result to update TestExpectations.
Tien-Ren Chen
Comment 8 2013-01-30 02:17:31 PST
Comment on attachment 185414 [details] Patch I'm surprised Mac EWS passed all the tests. Do they generate pixel results properly for repaint regions?
Tim Horton
Comment 9 2013-01-30 02:18:03 PST
(In reply to comment #8) > (From update of attachment 185414 [details]) > I'm surprised Mac EWS passed all the tests. Do they generate pixel results properly for repaint regions? Mac EWS doesn't run pixel tests.
Tien-Ren Chen
Comment 10 2013-01-30 02:19:13 PST
Got it. I'll run it on my Macbook then.
Tien-Ren Chen
Comment 11 2013-01-30 02:32:26 PST
Created attachment 185442 [details] Patch
Tim Horton
Comment 12 2013-01-30 02:35:14 PST
Comment on attachment 185442 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=185442&action=review > Source/WebCore/editing/FrameSelection.cpp:1330 > + // FIXME: Need to over-paint 1 pixel to workaround some rounding problem. s/some/a/ or s/problem/problems/. > Source/WebCore/editing/FrameSelection.cpp:1332 > + // Can't reproduce http://webkit.org/b/19086 but probably the same root cause. This line doesn't belong here.
Tien-Ren Chen
Comment 13 2013-01-30 02:38:40 PST
Created attachment 185445 [details] Patch
WebKit Review Bot
Comment 14 2013-01-30 03:22:03 PST
Comment on attachment 185445 [details] Patch Clearing flags on attachment: 185445 Committed r141243: <http://trac.webkit.org/changeset/141243>
WebKit Review Bot
Comment 15 2013-01-30 03:22:07 PST
All reviewed patches have been landed. Closing bug.
Emil A Eklund
Comment 16 2013-01-30 09:18:03 PST
Thanks for adding back that workaround, I'll try to figure out why we get the repaint rect wrong for the caret when I get a chance. We had similar problems in chromium when we first turned on subpixel layout.
Note You need to log in before you can comment on or make changes to this bug.