[CSS Shapes] Normalize margin box size

[CSS Shapes] Normalize margin box size

Since the reference box may end up with negative widths/heights
due to negative margins, clamp to zero for the reference box
calculation in the MarginBox case. Since fixing this (466942)
ends up hitting the ASSERT from bug 450625, fix that by
also clamping the floatMarginBoxWidth to zero since a float
with negative width has no float area.

BUG=450625, 466942
Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=192180

[rwlbuis, 2015-03-13 22:52:14]

rob.buis@samsung.com changed reviewers:
+ bemjb@chromium.org

[rwlbuis, 2015-03-13 22:52:15]

PTAL

[rwlbuis, 2015-03-14 02:34:31]

rob.buis@samsung.com changed reviewers:
+ bjonesbe@adobe.com

[bjonesbe, 2015-03-16 16:21:43]

For the reference box size, you should clamp to 0, not abs. A negative height or
width on the reference box should give a shape with no extent.

[rwlbuis, 2015-03-16 17:21:32]

Patchset #3 (id:40001) has been deleted

[bjonesbe, 2015-03-16 22:26:34]

bjonesbe@601t.com changed reviewers:
+ bjonesbe@601t.com

[bjonesbe, 2015-03-16 22:26:35]

https://codereview.chromium.org/1008893002/diff/60001/Source/core/layout/shap...
File Source/core/layout/shapes/ShapeOutsideInfo.cpp (right):

https://codereview.chromium.org/1008893002/diff/60001/Source/core/layout/shap...
Source/core/layout/shapes/ShapeOutsideInfo.cpp:59: if
(newReferenceBoxLogicalSize.isEmpty()) {
Sorry I missed this earlier, but in this case you could just set
newReferenceBoxLogicalSize to LayoutSize(), as I don't think there's any
situation where one of the dimensions can have a 0 length and still be able to
create a shape with an extent.

https://codereview.chromium.org/1008893002/diff/60001/Source/core/layout/shap...
Source/core/layout/shapes/ShapeOutsideInfo.cpp:294: LayoutUnit
floatMarginBoxWidth =
containingBlock.logicalWidthForFloat(&floatingObject).abs();
This one also needs to clamp to 0, as a float with negative width has no float
area.

[rwlbuis, 2015-03-17 20:33:10]

On 2015/03/16 22:26:35, bjonesbe wrote:
>
https://codereview.chromium.org/1008893002/diff/60001/Source/core/layout/shap...
> File Source/core/layout/shapes/ShapeOutsideInfo.cpp (right):
> 
>
https://codereview.chromium.org/1008893002/diff/60001/Source/core/layout/shap...
> Source/core/layout/shapes/ShapeOutsideInfo.cpp:59: if
> (newReferenceBoxLogicalSize.isEmpty()) {
> Sorry I missed this earlier, but in this case you could just set
> newReferenceBoxLogicalSize to LayoutSize(), as I don't think there's any
> situation where one of the dimensions can have a 0 length and still be able to
> create a shape with an extent.

Done.

>
https://codereview.chromium.org/1008893002/diff/60001/Source/core/layout/shap...
> Source/core/layout/shapes/ShapeOutsideInfo.cpp:294: LayoutUnit
> floatMarginBoxWidth =
> containingBlock.logicalWidthForFloat(&floatingObject).abs();
> This one also needs to clamp to 0, as a float with negative width has no float
> area.

Done.

[Bem Jones-Bey (adobe), 2015-03-17 20:35:55]

lgtm

[rwlbuis, 2015-03-17 20:47:06]

rob.buis@samsung.com changed reviewers:
+ leviw@chromium.org

[rwlbuis, 2015-03-17 20:47:07]

@leviw PTAL

[leviw_travelin_and_unemployed, 2015-03-19 15:31:40]

The CQ bit was checked by leviw@chromium.org

[leviw_travelin_and_unemployed, 2015-03-19 15:31:41]

LGTM.

[commit-bot: I haz the power, 2015-03-19 15:32:03]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1008893002/100001

[rwlbuis, 2015-03-19 15:52:59]

On 2015/03/19 15:31:41, leviw wrote:
> LGTM.

Thanks! Two bugs squashed for the price of one.

[commit-bot: I haz the power, 2015-03-19 16:59:41]

Committed patchset #5 (id:100001) as
https://src.chromium.org/viewvc/blink?view=rev&revision=192180