RenderBlockFlow::nextFloatLogicalBottomBelow should not use ShapeOutsideFloatOffsetMode

[Bem Jones-Bey]

Once this is refactored, we can remove ShapeOutsideFloatOffsetMode.

[Bem Jones-Bey]

Created attachment 216426 [details]
Patch

[Sam Weinig]

Comment on attachment 216426 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=216426&action=review

> Source/WebCore/rendering/FloatingObjects.cpp:159
> +    FindNextFloatLogicalBottomAdapter(const RenderBlockFlow* renderer, LayoutUnit belowLogicalHeight)

You should take the RenderBlockFlow by reference.

> Source/WebCore/rendering/FloatingObjects.cpp:176
> +    const RenderBlockFlow* m_renderer;

This should be a reference.

[Bem Jones-Bey]

Created attachment 216431 [details]
Updated Patch

[Bem Jones-Bey]

(In reply to comment #2)
> (From update of attachment 216426 [details])
> View in context: https://bugs.webkit.org/attachment.cgi?id=216426&action=review
> 
> > Source/WebCore/rendering/FloatingObjects.cpp:159
> > +    FindNextFloatLogicalBottomAdapter(const RenderBlockFlow* renderer, LayoutUnit belowLogicalHeight)
> 
> You should take the RenderBlockFlow by reference.
> 
> > Source/WebCore/rendering/FloatingObjects.cpp:176
> > +    const RenderBlockFlow* m_renderer;
> 
> This should be a reference.

Thanks, I've updated that.

[WebKit Commit Bot]

Comment on attachment 216431 [details]
Updated Patch

Clearing flags on attachment: 216431

Committed r158954: <http://trac.webkit.org/changeset/158954>

[WebKit Commit Bot]

All reviewed patches have been landed.  Closing bug.

[Zan Dobersek]

Roughly 80 tests started failing on the GTK port after this landed, due to that port enabling subpixel layout. All tests pass again if the saturated layout arithmetic in LayoutUnit is enabled. The patch in bug #119273 does just that.