RESOLVED FIXED 20482
REGRESSION (r34342-r34367): Crash at WebCore::canHaveChildrenForEditing
https://bugs.webkit.org/show_bug.cgi?id=20482
Summary REGRESSION (r34342-r34367): Crash at WebCore::canHaveChildrenForEditing
Robert Blaut
Reported 2008-08-22 01:24:41 PDT
Reproducible crash on http://my.opera.com/ODIN/blog/lovely-break-css-working-group after using ctrl + cmd + d (Lookup in Dictionary) 1) Go to http://my.opera.com/ODIN/blog/lovely-break-css-working-group 2) Place cursor on "lovely" word 3) Press ctrl + cmd + d 4) Webkit r35844 for Mac OS X crashes Note: if it doesn't crash for the first time, use shortcut one more time.
Attachments
crash log (24.25 KB, text/plain)
2008-08-22 01:25 PDT, Robert Blaut
no flags
reduction (349 bytes, text/html)
2008-09-20 17:07 PDT, Gavin Sherlock
no flags
Robert Blaut
Comment 1 2008-08-22 01:25:56 PDT
Created attachment 22932 [details] crash log
Tom Ierna
Comment 2 2008-09-20 08:37:08 PDT
Gavin Sherlock
Comment 3 2008-09-20 09:02:13 PDT
Took me about a dozen tries to reproduce it, but I see the crash also. Am unable to reproduce in Safari 3.1.2 after trying for a while, so presume this is a regression.
Gavin Sherlock
Comment 4 2008-09-20 09:23:05 PDT
Easier steps to reproduce: 1) Go to http://my.opera.com/ODIN/blog/lovely-break-css-working-group 2) Place cursor on "lovely" word 3) Press ctrl + cmd + d 4) If it hasn't crashed, place cursor on "lunch" 5) Press ctrl + cmd + d
Gavin Sherlock
Comment 5 2008-09-20 09:23:43 PDT
Regressed between r34342 and r34367 ( don't have privileges to change the title).
Gavin Sherlock
Comment 6 2008-09-20 09:28:08 PDT
I suspect: http://trac.webkit.org/changeset/34344 though am not sure.
Gavin Sherlock
Comment 7 2008-09-20 17:07:43 PDT
Created attachment 23617 [details] reduction 1) Place cursor on "One" word 2) Press ctrl + cmd + d 3) Place cursor on "Two" 4) Press ctrl + cmd + d Crash! I don't seem to be able to reduce any further.
David Kilzer (:ddkilzer)
Comment 8 2008-11-29 05:28:18 PST
I don't see this crash anymore with WebKit nightly build r38826 on Mac OS X 10.4.11 (8S165) with Safari 3.2. Can anyone else still reproduce it?
Gavin Sherlock
Comment 9 2008-11-29 07:59:41 PST
r38826 also doesn't crash for me on 10.5.5 with Safari 3.2. It would be nice to know what changeset fixed this, and to at least have a manual test to prevent it regressing.
David Kilzer (:ddkilzer)
Comment 10 2008-11-29 11:18:20 PST
Running bisect-builds to determine when this was fixed.
David Kilzer (:ddkilzer)
Comment 11 2008-11-29 11:21:52 PST
(In reply to comment #10) > Running bisect-builds to determine when this was fixed. Fails: r38297 Works: r38377
David Kilzer (:ddkilzer)
Comment 12 2008-11-29 11:28:27 PST
(In reply to comment #11) > (In reply to comment #10) > > Running bisect-builds to determine when this was fixed. > Fails: r38297 Works: r38377 I suspect r38304 fixed this. http://trac.webkit.org/changeset/38304 See also Bug 22160.
Note You need to log in before you can comment on or make changes to this bug.