Don't always full invalidate on change of align-content, align-items and align-self.

Don't always full invalidate on change of align-content, align-items and align-self.

On change of align-content, align-items and align-self, if there are
real geometry changes, the renderer can handle through the normal
geometry change path. Otherwise we should not repaint.

Move the cases from
RenderStyle::diffNeedsFullLayoutAndPaintInvalidation()
to
RenderStyle::diffNeedsFullLayout

Tests cover cases of align-content, align-items and align-self changes
needing invalidation, and align-content and align-self changes not
needing invalidation.

For now any change of align-items causes reattaching of RenderObject
so we still always invalidate.

This change is needed for bug 408697 (position style change) because
changing position:relative to position:absolute also changes align-self.

BUG=375876
TEST=fast/repaint/align*.html
R=dsinclair@chromium.org, jchaffraix@chromium.org

Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=181415

[Xianzhu, 2014-09-03 20:38:29]

wangxianzhu@chromium.org changed reviewers:
+ dsinclair@chromium.org, jchaffraix@chromium.org, leviw@chromium.org

[dsinclair, 2014-09-03 20:48:12]

https://codereview.chromium.org/538613003/diff/1/LayoutTests/fast/repaint/ali...
File LayoutTests/fast/repaint/align-content-change.html (right):

https://codereview.chromium.org/538613003/diff/1/LayoutTests/fast/repaint/ali...
LayoutTests/fast/repaint/align-content-change.html:27: <p style="height:
20px">Tests repaint on align-content style change. Passes if there is no
red.</p>
s/repaint/invalidation

https://codereview.chromium.org/538613003/diff/1/LayoutTests/fast/repaint/ali...
LayoutTests/fast/repaint/align-content-change.html:27: <p style="height:
20px">Tests repaint on align-content style change. Passes if there is no
red.</p>
Can this be a ref test instead of a -expected.txt? (and the others which pass if
there is no red)

[Xianzhu, 2014-09-03 21:29:20]

https://codereview.chromium.org/538613003/diff/1/LayoutTests/fast/repaint/ali...
File LayoutTests/fast/repaint/align-content-change.html (right):

https://codereview.chromium.org/538613003/diff/1/LayoutTests/fast/repaint/ali...
LayoutTests/fast/repaint/align-content-change.html:27: <p style="height:
20px">Tests repaint on align-content style change. Passes if there is no
red.</p>
On 2014/09/03 20:48:12, dsinclair wrote:
> s/repaint/invalidation

Done.

https://codereview.chromium.org/538613003/diff/1/LayoutTests/fast/repaint/ali...
LayoutTests/fast/repaint/align-content-change.html:27: <p style="height:
20px">Tests repaint on align-content style change. Passes if there is no
red.</p>
On 2014/09/03 20:48:12, dsinclair wrote:
> Can this be a ref test instead of a -expected.txt? (and the others which pass
if
> there is no red)

I think ref test can't replace a repaint(invalidation)-test for testing correct
invalidation:
- a ref test might hide under invalidation because we may paint more than
invalidated area;
- we can't test over-invalidation with a ref test.

[dsinclair, 2014-09-04 14:10:46]

lgtm

https://codereview.chromium.org/538613003/diff/1/LayoutTests/fast/repaint/ali...
File LayoutTests/fast/repaint/align-content-change.html (right):

https://codereview.chromium.org/538613003/diff/1/LayoutTests/fast/repaint/ali...
LayoutTests/fast/repaint/align-content-change.html:27: <p style="height:
20px">Tests repaint on align-content style change. Passes if there is no
red.</p>
On 2014/09/03 21:29:20, Xianzhu wrote:
> On 2014/09/03 20:48:12, dsinclair wrote:
> > Can this be a ref test instead of a -expected.txt? (and the others which
pass
> if
> > there is no red)
> 
> I think ref test can't replace a repaint(invalidation)-test for testing
correct
> invalidation:
> - a ref test might hide under invalidation because we may paint more than
> invalidated area;
> - we can't test over-invalidation with a ref test.


Ah, k, I thought you were just testing correctness, didn't realize you were also
looking for minimal invalidation rects.

[Julien - ping for review, 2014-09-04 18:45:17]

The CQ bit was checked by jchaffraix@chromium.org

[Julien - ping for review, 2014-09-04 18:45:17]

lgtm

[commit-bot: I haz the power, 2014-09-04 18:45:35]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/wangxianzhu@chromium.org/538613003/40001

[Xianzhu, 2014-09-04 22:12:21]

The CQ bit was unchecked by wangxianzhu@chromium.org

[Xianzhu, 2014-09-04 22:12:26]

The CQ bit was checked by wangxianzhu@chromium.org

[commit-bot: I haz the power, 2014-09-04 22:12:44]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/wangxianzhu@chromium.org/538613003/40001

[Xianzhu, 2014-09-04 23:11:22]

Committed patchset #3 (id:40001) manually as 181415 (presubmit successful).