[Webkit-unassigned] [Bug 102352] Avoid a second layout of flex items in layoutAndPlaceChildren()

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Wed Nov 28 17:28:16 PST 2012


https://bugs.webkit.org/show_bug.cgi?id=102352





--- Comment #15 from Tony Chang <tony at chromium.org>  2012-11-28 17:30:31 PST ---
I debugged the mathml failures and they uncovered a real problem.  Here's an example:
http://plexode.com/u/#2N=0&-Mle2L%3A2K%0AEJheightEH5olid!EG%3Cdiv29%3E28%3Dj7K!!G!id26%22r54borderL!1px!sb4!styM86b*%3C%2Fdiv92!%20~http://pMxode.com/eval3/#ht=G4displayL-webkit-fMx6978aHblue%3B!JL!30px69a*78bHgreen%3B69b*K*&ohjNjt=document.getEMmentById(6a6).styM.J!8!620px6K&ojhNojjNcex=0


Press eval JS now to see the bug.  What happens is that we previously stretched a child and since we don't layout the second time, we keep using the stretched height, which is incorrect.  We should include this test as a layout test.

A simple fix is to ignore this optimization if the child is stretching (i.e., always relayout stretching children).

For the auto margins case, updateLogicalHeight should work, but we may be able to be smarter about it (maybe just set the margins to 0 if they style is auto?).

I'll try to put together a patch tomorrow.

-- 
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.


More information about the webkit-unassigned mailing list