Refactor CanvasRenderingContext2D::scrollPathIntoViewInternal().

Refactor CanvasRenderingContext2D::scrollPathIntoViewInternal().

This fixes a crash when a RenderBoxModelObject is null.
Also, we can simplify the logic to take border and padding into account by using absoluteContextBox().

BUG=391185
TEST=https://cluster-fuzz.appspot.com/testcase?key=6636208132194304

Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=177625

[zino, 2014-07-04 04:26:41]

Please take a look.
Thank you.

[yosin_UTC9, 2014-07-04 05:45:02]

rsLGTM

I'm not familiar with Canvas 2D code, but the changes seems to  be OK.

[Justin Novosad, 2014-07-04 14:46:04]

Could you add a layout test that triggers the crash that you are fixing?

[zino, 2014-07-05 05:08:49]

On 2014/07/04 14:46:04, junov wrote:
> Could you add a layout test that triggers the crash that you are fixing?

I added a layout test but I'm not sure if the test is correct.
The test codes seem to be strange. (It is based on minimized testcase to be
attached in crbug/391185)
I don't know why the strange codes make renderer to be null.

Nevertheless, I think we always need to check whether the renderer is null.

Thank you.

[Justin Novosad, 2014-07-07 13:42:14]

On 2014/07/05 05:08:49, zino wrote:
> On 2014/07/04 14:46:04, junov wrote:
> > Could you add a layout test that triggers the crash that you are fixing?
> 
> I added a layout test but I'm not sure if the test is correct.
> The test codes seem to be strange. (It is based on minimized testcase to be
> attached in crbug/391185)
> I don't know why the strange codes make renderer to be null.
> 
> Nevertheless, I think we always need to check whether the renderer is null.
> 
> Thank you.

Oh, this is from clusterfuzz? Sorry I did not realize that, I had not looked
at the bug. In that case, you don't have to add a test. But you should
mention in the change description that this is a fix for a culterfuzz crash.
One way of doing that is to add the line
TEST=https://cluster-fuzz.appspot.com/testcase?key=6636208132194304
at the end of the change description

Just remove the test, edit the description, then you can commit. No need
to wait for another review.

lgtm

[zino, 2014-07-08 00:21:51]

The CQ bit was checked by jinho.bang@samsung.com

[commit-bot: I haz the power, 2014-07-08 00:22:14]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/jinho.bang@samsung.com/367113007/1

[commit-bot: I haz the power, 2014-07-08 01:24:49]

FYI, CQ is re-trying this CL (attempt #1).
The failing builders are:
  win_blink_rel on tryserver.blink
(http://build.chromium.org/p/tryserver.blink/builders/win_blink_rel/builds/15785)

[commit-bot: I haz the power, 2014-07-08 01:54:21]

Change committed as 177625