Bug 33501 - [chromium] FontLinux::selectionRectForComplexText can assert when TextRun gets normalized
Summary: [chromium] FontLinux::selectionRectForComplexText can assert when TextRun get...
Status: RESOLVED WONTFIX
Alias: None
Product: WebKit
Classification: Unclassified
Component: Platform (show other bugs)
Version: 528+ (Nightly build)
Hardware: Other Linux
: P2 Normal
Assignee: Nobody
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2010-01-11 16:22 PST by John Gregg
Modified: 2013-04-09 16:09 PDT (History)
3 users (show)

See Also:


Attachments
Patch (3.97 KB, patch)
2010-01-11 18:01 PST, John Gregg
no flags Details | Formatted Diff | Diff
Patch (3.98 KB, patch)
2010-01-11 18:07 PST, John Gregg
no flags Details | Formatted Diff | Diff

Note You need to log in before you can comment on or make changes to this bug.
Description John Gregg 2010-01-11 16:22:32 PST
This bug causes LayoutTests/fast/text/find-kana.html to assert in the following way:

When a find is *successful* on a string with U+0300 character, chromium updates the selection and
Font::selectionRectForComplexText(run, point, height, from, to)
   is called with run={length: 3, ...}, from=0, to=3,
but the TextRunWalker walker(run) gets normalized so that
   walker={length: 2}, from=0, to=3,
and we essentially hit an out-of-bounds problem when working with the walker.

I think the correct fix is to not normalize in this case; I will submit a patch for this.
Comment 1 John Gregg 2010-01-11 18:01:14 PST
Created attachment 46326 [details]
Patch
Comment 2 John Gregg 2010-01-11 18:07:40 PST
Created attachment 46327 [details]
Patch
Comment 3 John Gregg 2010-01-11 18:13:30 PST
CC+ hbono
Comment 4 Jungshik Shin 2010-03-05 10:53:51 PST
The normalization was added by Yusuke with my review (in r47922). (bug 28742). The patch looks ok, but I'm not sure.
Comment 5 Stephen Chenney 2013-04-09 16:09:38 PDT
LayoutTest failures for Chromium are being marked WontFix. The Bug is still accessible and referenced from TestExpectations.