Make currentColor keyword in linear-gradient work

Make currentColor keyword in linear-gradient work

In the case where the CSSGradientValue is not cacheable (i.e. a color
stop uses currentColor), pass the RenderObject so the computed color
stops gets resolved against its RenderStyle.

Add a reftest for this.

BUG=436585
Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=186438

[rwlbuis, 2014-12-02 19:47:47]

rob.buis@samsung.com changed reviewers:
+ esprehn@chromium.org, timloh@chromium.org

[rwlbuis, 2014-12-02 19:47:48]

PTAL

[Timothy Loh, 2014-12-02 21:56:21]

On 2014/12/02 19:47:48, rwlbuis wrote:
> PTAL

Is it possible to delay resolving currentColor in gradients until later?

[rwlbuis, 2014-12-02 22:53:14]

On 2014/12/02 21:56:21, Timothy Loh wrote:
> On 2014/12/02 19:47:48, rwlbuis wrote:
> > PTAL
> 
> Is it possible to delay resolving currentColor in gradients until later?

It does seem possible, I am testing a new patch.

[rwlbuis, 2014-12-03 00:23:02]

On 2014/12/02 22:53:14, rwlbuis wrote:
> On 2014/12/02 21:56:21, Timothy Loh wrote:
> > On 2014/12/02 19:47:48, rwlbuis wrote:
> > > PTAL
> > 
> > Is it possible to delay resolving currentColor in gradients until later?
> 
> It does seem possible, I am testing a new patch.

Patch set 4 is all green! I updated the description too, should be ready for
review.

[Timothy Loh, 2014-12-03 01:09:59]

On 2014/12/03 00:23:02, rwlbuis wrote:
> On 2014/12/02 22:53:14, rwlbuis wrote:
> > On 2014/12/02 21:56:21, Timothy Loh wrote:
> > > On 2014/12/02 19:47:48, rwlbuis wrote:
> > > > PTAL
> > > 
> > > Is it possible to delay resolving currentColor in gradients until later?
> > 
> > It does seem possible, I am testing a new patch.
> 
> Patch set 4 is all green! I updated the description too, should be ready for
> review.

lgtm

Just to double check that I'm understanding all the code, we call into image(..)
when we want to paint, right? So we leave the position as a CSSValue all the way
until then?

[Timothy Loh, 2014-12-03 01:10:07]

https://codereview.chromium.org/767343003/diff/60001/Source/core/css/CSSGradi...
File Source/core/css/CSSGradientValue.h (left):

https://codereview.chromium.org/767343003/diff/60001/Source/core/css/CSSGradi...
Source/core/css/CSSGradientValue.h:61: Color m_resolvedColor;
I guess we can remove this now?

[rwlbuis, 2014-12-03 15:56:39]

On 2014/12/03 01:10:07, Timothy Loh wrote:
>
https://codereview.chromium.org/767343003/diff/60001/Source/core/css/CSSGradi...
> File Source/core/css/CSSGradientValue.h (left):
> 
>
https://codereview.chromium.org/767343003/diff/60001/Source/core/css/CSSGradi...
> Source/core/css/CSSGradientValue.h:61: Color m_resolvedColor;
> I guess we can remove this now?

It is not trivial, but maybe it can be removed. I'll investigate some more and
if it is possible
make a follow-up patch.

[rwlbuis, 2014-12-03 15:56:44]

The CQ bit was checked by rob.buis@samsung.com

[commit-bot: I haz the power, 2014-12-03 15:57:03]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/767343003/60001

[commit-bot: I haz the power, 2014-12-03 17:10:33]

Committed patchset #4 (id:60001) as
https://src.chromium.org/viewvc/blink?view=rev&revision=186438