[Webkit-unassigned] [Bug 50119] New: RenderObject::minimumPreferredLogicalWidth() sometimes reports wrong value

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=50119

           Summary: RenderObject::minimumPreferredLogicalWidth() sometimes
                    reports wrong value
           Product: WebKit
           Version: 528+ (Nightly build)
          Platform: PC
        OS/Version: Mac OS X 10.5
            Status: NEW
          Severity: Normal
          Priority: P2
         Component: Layout and Rendering
        AssignedTo: webkit-unassigned at lists.webkit.org
        ReportedBy: aa at chromium.org
                CC: hyatt at apple.com, sam at webkit.org, jamesr at chromium.org


I noticed a change in the behavior of RenderObject::minimumPreferredLogicalWidth() for the root renderer recently, and tracked it to this change:

http://trac.webkit.org/changeset/70546/trunk/WebCore/rendering/RenderBox.cpp

Also see http://code.google.com/p/chromium/issues/detail?id=61370 for the Chromium bug where this was noticed.

It looks to me like a style can sometimes be applied to a body element before it is appended to the tree. In this case, the dirty bit for minimumPreferredLogicalWidth will not be propagated up the tree.

I have a patch that appears to fix the problem, but I would like feedback on:

a) Whether this is a good way to fix it, and
b) How to create a test case

Patch to follow.

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