MediaStream API: Stop means stop

MediaStream API: Stop means stop
Removing notification functionality from MediaStreamTrack::stop because the world is being torn down.

BUG=232064
Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=148673

[eseidel, 2013-04-17 17:32:30]

Can we test this?

[Tommy Widenflycht, 2013-04-18 07:37:38]

On 2013/04/17 17:32:30, Eric Seidel (Google) wrote:
> Can we test this?

Don't know how to write a layout test for it but it is tested by the fuzzers
phoglund have added. That's how this bug was found.

[eseidel, 2013-04-18 07:45:27]

Do the fuzzers spit out HTML?  Could we just dump one of the fuzzer results?

[Tommy Widenflycht, 2013-04-18 07:46:09]

On 2013/04/18 07:37:38, Tommy Widenflycht wrote:
> On 2013/04/17 17:32:30, Eric Seidel (Google) wrote:
> > Can we test this?
> 
> Don't know how to write a layout test for it but it is tested by the fuzzers
> phoglund have added. That's how this bug was found.

And btw I manually verified that the crash went away after applying this patch.

[Tommy Widenflycht, 2013-04-18 07:57:17]

On 2013/04/18 07:45:27, Eric Seidel (Google) wrote:
> Do the fuzzers spit out HTML?  Could we just dump one of the fuzzer results?

Yes, but detecting this requires a combination of html and page reloads. 
But isn't this clearly something one shouldn't do?

[eseidel, 2013-04-18 08:15:44]

I mean, I'm not objecting to the change. :)  It's not obvious to me that stop()
shouldn't call didEndTrack(), but I don't know the mediastream code particularly
well.

In general, we like to test as much as possible so we don't repeat our mistakes.
 If it's impractical to write a layout test for this, I totally understand.  But
if it's reasonable, and you think it has any utility, I think we should do so. 
We have lots of fuzzer-generated layout tests which are just supposed to not
crash.  At least they help us to make sure we don't crash in the same way twice,
and they help to feed inferno's fuzzer (which takes recently added layout tests
as seeds).

[Tommy Widenflycht, 2013-04-18 14:43:35]

I did a decent effort to try to get DRT to repro this issue but without any
success.

[abarth-chromium, 2013-04-18 16:05:37]

lgtm

ok

[commit-bot: I haz the power, 2013-04-18 16:05:46]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/tommyw@chromium.org/14031012/1

[eseidel, 2013-04-18 16:06:11]

lgtm

rs=me.

[commit-bot: I haz the power, 2013-04-18 16:44:28]

Retried try job too often on linux_layout_rel for step(s) webkit_tests
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=linux_layo...

[commit-bot: I haz the power, 2013-04-18 16:45:26]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/tommyw@chromium.org/14031012/1

[commit-bot: I haz the power, 2013-04-18 17:35:45]

Retried try job too often on linux_layout_rel for step(s) webkit_tests
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=linux_layo...

[commit-bot: I haz the power, 2013-04-18 17:40:51]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/tommyw@chromium.org/14031012/1

[commit-bot: I haz the power, 2013-04-18 18:00:36]

Change committed as 148673