Fix a crash in WebViewImpl::copyImageAt.

Fix a crash in WebViewImpl::copyImageAt.

We need null check because hit result may be null for various reasons.

BUG=440894
Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=188373

[zino, 2015-01-07 10:20:33]

jinho.bang@samsung.com changed reviewers:
+ junov@chromium.org

[Justin Novosad, 2015-01-13 15:53:24]

I just saw this now, sorry about the long review delay.

LGTM

[Justin Novosad, 2015-01-13 15:56:25]

As we discussed, there is probably a better way to fix this, but I think that
that we should land this for now, and file a follow-up bug that addresses the
root cause (related to the FIXME).  Also, remove "[WIP]" form the title before
submitting.

[Justin Novosad, 2015-01-13 15:58:17]

junov@chromium.org changed reviewers:
+ dglazkov@chromium.org

[Justin Novosad, 2015-01-13 15:58:18]

+ dglazkov for Source/web OWNERS

[dglazkov, 2015-01-13 16:36:33]

On 2015/01/13 at 15:58:18, junov wrote:
> + dglazkov for Source/web OWNERS

Where's the null-check?

[Justin Novosad, 2015-01-13 16:39:39]

On 2015/01/13 16:36:33, dglazkov wrote:
> On 2015/01/13 at 15:58:18, junov wrote:
> > + dglazkov for Source/web OWNERS
> 
> Where's the null-check?

in isHTMLCanvasElement

[dglazkov, 2015-01-13 16:41:51]

On 2015/01/13 at 16:39:39, junov wrote:
> On 2015/01/13 16:36:33, dglazkov wrote:
> > On 2015/01/13 at 15:58:18, junov wrote:
> > > + dglazkov for Source/web OWNERS
> > 
> > Where's the null-check?
> 
> in isHTMLCanvasElement

It's already been there, right? I must be looking at this wrong.

[Justin Novosad, 2015-01-13 16:49:28]

On 2015/01/13 16:41:51, dglazkov wrote:
> On 2015/01/13 at 16:39:39, junov wrote:
> > On 2015/01/13 16:36:33, dglazkov wrote:
> > > On 2015/01/13 at 15:58:18, junov wrote:
> > > > + dglazkov for Source/web OWNERS
> > > 
> > > Where's the null-check?
> > 
> > in isHTMLCanvasElement
> 
> It's already been there, right? I must be looking at this wrong.

It is subtle. The old code calls the non-pointer flavor of isHTMLCanvasElement,
so the call site was dereferencing a null pointer.

[dglazkov, 2015-01-13 21:24:14]

On 2015/01/13 at 16:49:28, junov wrote:
> On 2015/01/13 16:41:51, dglazkov wrote:
> > On 2015/01/13 at 16:39:39, junov wrote:
> > > On 2015/01/13 16:36:33, dglazkov wrote:
> > > > On 2015/01/13 at 15:58:18, junov wrote:
> > > > > + dglazkov for Source/web OWNERS
> > > > 
> > > > Where's the null-check?
> > > 
> > > in isHTMLCanvasElement
> > 
> > It's already been there, right? I must be looking at this wrong.
> 
> It is subtle. The old code calls the non-pointer flavor of
isHTMLCanvasElement, so the call site was dereferencing a null pointer.

Oh jesus. LGTM.

[zino, 2015-01-14 09:18:02]

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

[commit-bot: I haz the power, 2015-01-14 09:19:00]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/836283002/1

[commit-bot: I haz the power, 2015-01-14 10:40:01]

Committed patchset #1 (id:1) as
https://src.chromium.org/viewvc/blink?view=rev&revision=188373