408292 - "ASSERTION: lost track of line breaks somehow" with -moz-column, text-transform, rtl

Status: RESOLVED FIXED

(Core :: Layout: Text and Fonts, defect)

Description

[Jesse Ruderman]

Attachment: testcase

Loading the testcase triggers:

###!!! ASSERTION: lost track of line breaks somehow: 'nextLineBreak == aTextRun->mLineBreaks.Length()', file /Users/jruderman/trunk/mozilla/layout/generic/nsTextRunTransformations.cpp, line 462

Comment 1

[Simon Montagu :smontagu]

This is probably because we don't do bidi resolution quite right across column boundaries. I think all the stuff that was blocking fixing that when we talked about it a few months ago has meanwhile been fixed.

Comment 2

[Simon Montagu :smontagu]

Attachment: Include all continuation blocks in the array of text frames

Comment 3

[Robert O'Callahan (:roc) (email my personal email if necessary)]

Comment on attachment 293535 [details] [diff] [review]
Include all continuation blocks in the array of text frames

wow, that looks easy!

Comment 4

[Simon Montagu :smontagu]

Robert, do you have a testcase which requires the offset fixups at http://mxr.mozilla.org/seamonkey/source/layout/generic/nsTextFrameThebes.cpp#6020 ? It may be possible to convert them to assertions already.

Comment 5

[Robert O'Callahan (:roc) (email my personal email if necessary)]

Not at my fingertips, but the CVS log should show you the bugs with the testcases that led to those fixups.

Comment 6

[Simon Montagu :smontagu]

Comment on attachment 293535 [details] [diff] [review]
Include all continuation blocks in the array of text frames

Too easy, apparently: this regresses a couple of the new crashtests. Go Jesse.

Comment 7

[Simon Montagu :smontagu]

Attachment: Patch for checkin

It wasn't really a regression, I was just missing a trivial null-check. In other words, our test suites are catching faults above and beyond the bugs they are actually testing for :)

Transferring r+sr=roc.

Comment 8

[Simon Montagu :smontagu]

Checked in with the testcase as a crashtest.

Comment 9

[(no longer active)]

Mass-assigning the new rtl keyword to RTL-related (see bug 349193).