430352 - [FIX]"ASSERTION: width less than zero: 'aContainingBlockWidth >= 0'"

Status: RESOLVED FIXED

(Core :: Layout, defect)

Description

[Jesse Ruderman]

Attachment: testcase

###!!! ASSERTION: width less than zero: 'aContainingBlockWidth >= 0', file /Users/jruderman/trunk/mozilla/layout/base/nsLayoutUtils.cpp, line 1857

Comment 1

[Boris Zbarsky [:bzbarsky]]

We're calling CalculateContainingBlockSizeForAbsolutes with an aReflowState which has:

  (gdb) p aReflowState.mStyleBorder->GetBorder()
  $12 = (const nsMargin &) @0xb144d558: {top = 60, right = 60,
                                         bottom = 60, left = 60}

but also has:

  (gdb) p aReflowState.mComputedBorderPadding
  $9 = {top = 0, right = 0, bottom = 0, left = 0}

aFrameSize is 0x0 in this case.  So CalculateContainingBlockSizeForAbsolutes is computing a negative cbSize.width (and same for height), which ends up propagating down to the child reflow state.  I assume the moz-appearance is what clobbers the computed border/padding here?

Is there a reason we're not using the mComputedBorderPadding - mComputedPadding here to find the border?  And should the moz-appearance really lead to a 0 mComputedBorderPadding in this case?

Comment 2

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

You're right, we should be using mComputedBorderPadding - mComputedPadding there.

Comment 3

[Boris Zbarsky [:bzbarsky]]

Attachment: Fix

Comment 4

[Shawn Wilsher :sdwilsh]

Pushed to mozilla-central:
http://hg.mozilla.org/mozilla-central/index.cgi/rev/bee72059de15