[Webkit-unassigned] [Bug 191627] [css-grid] Incorrect track sizing when using relative sized items in 'auto' column tracks

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=191627

--- Comment #15 from zsun at igalia.com ---
(In reply to zsun from comment #14)
> Yes, the value might not have been initialized/set. While running the test
> case here
> (grid-self-baseline-not-applied-if-sizing-cyclic-dependency-001.html),
> without the valueOr(), this test crashes and throws "STDERR: ASSERTION
> FAILED: initialized()".
> 
> Would the width be nullopt? I'm just wondering why in
> RenderGrid::placeItemsOnGrid(), it's initialized as LayoutUnit() rather than
> LayoutUnit(-1) like the height?

(In reply to Javier Fernandez from comment #13)
> Comment on attachment 421640 [details]
> Patch
> 
> View in context:
> https://bugs.webkit.org/attachment.cgi?id=421640&action=review
> 
> >> Source/WebCore/rendering/RenderBox.cpp:2046
> >> +        return overridingContainingBlockContentLogicalWidth().valueOr(LayoutUnit());
> > 
> > Hmm if hasOverridingContainingBlockContentLogicalWidth() is true then we know that it has a value, then why do we need the valueOr() ?
> 
> No, the HasOverrideXXX function only tells whether a value has been set
> (which is always true for grid items), bit it might be nullopt (to specify
> indefinite size)

Would it make sense to initiate it as 0 for width just as RenderGrid::placeItemsOnGrid() does?

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.webkit.org/pipermail/webkit-unassigned/attachments/20210308/eae33c4c/attachment.htm>