Update telemetry tests to be agnostic to ScrollTopLeftInterop mode

Update telemetry tests to be agnostic to ScrollTopLeftInterop mode

See https://dev.opera.com/articles/fixing-the-scrolltop-bug/
for details of the issue.  It should be possible to run telemetry
with --enable-experimental-web-platform-features.

BUG=157855
CQ_EXTRA_TRYBOTS=tryserver.chromium.perf:linux_perf_bisect;tryserver.chromium.perf:mac_10_10_perf_bisect;tryserver.chromium.perf:win_perf_bisect;tryserver.chromium.perf:android_nexus5_perf_bisect

Committed: https://crrev.com/a9b90544c07f7e54781c91856d0019e1f3a1567a
Cr-Commit-Position: refs/heads/master@{#356947}

[Rick Byers, 2015-10-29 02:23:37]

Description was changed from

==========
Update telemetry tests to be agnostic to ScrollTopLeftInterop mode

BUG=157855
==========

to

==========
Update telemetry tests to be agnostic to ScrollTopLeftInterop mode

See https://dev.opera.com/articles/fixing-the-scrolltop-bug/
for details of the issue.  It should be possible to run telemetry
with --enable-experimental-web-platform-features.

BUG=157855
==========

[Rick Byers, 2015-10-29 02:25:48]

rbyers@chromium.org changed reviewers:
+ skyostil@chromium.org

[Rick Byers, 2015-10-29 02:25:48]

Sami, PTAL
I had a developer reach out to me and mention that telemetry unit tests fail
when running with --enable-experimental-web-platform-features.  

I submitted these try jobs https://codereview.chromium.org/1422353003 with
ScrollTopLeftInterop set to stable and verified that this patch was sufficient
to get the telemetry unit tests to pass.  Do you think I should submit a perf
tryjob or anything to validate this further before landing?

Thanks,
   Rick

[Sami, 2015-10-29 10:12:43]

lgtm, thanks for the fix.

I guess document.scrollingElement has been shipped for a while now so this won't
cause problems when run on older reference builds (typically whatever is the
current stable channel)?

You might want to add extra cq trybots to make sure this doesn't break any of
the perf bots:

CQ_EXTRA_TRYBOTS=tryserver.chromium.perf:linux_perf_bisect;tryserver.chromium.perf:mac_10_10_perf_bisect;tryserver.chromium.perf:win_perf_bisect;tryserver.chromium.perf:android_nexus5_perf_bisect

[Rick Byers, 2015-10-29 14:27:43]

On 2015/10/29 10:12:43, Sami wrote:
> lgtm, thanks for the fix.
> 
> I guess document.scrollingElement has been shipped for a while now so this
won't
> cause problems when run on older reference builds (typically whatever is the
> current stable channel)?

Yes it shipped in M44.  How can I check to see what the current reference build
is?
 
> You might want to add extra cq trybots to make sure this doesn't break any of
> the perf bots:
> 
>
CQ_EXTRA_TRYBOTS=tryserver.chromium.perf:linux_perf_bisect;tryserver.chromium.perf:mac_10_10_perf_bisect;tryserver.chromium.perf:win_perf_bisect;tryserver.chromium.perf:android_nexus5_perf_bisect

Perfect, will do.  Thanks!

[Sami, 2015-10-29 19:04:13]

On 2015/10/29 14:27:43, Rick Byers wrote:
> On 2015/10/29 10:12:43, Sami wrote:
> > lgtm, thanks for the fix.
> > 
> > I guess document.scrollingElement has been shipped for a while now so this
> won't
> > cause problems when run on older reference builds (typically whatever is the
> > current stable channel)?
> 
> Yes it shipped in M44.  How can I check to see what the current reference
build
> is?

I don't exactly know where it's configured, but a good rule of thumb so far
seems to be the ~current stable release.

[Rick Byers, 2015-10-29 20:01:14]

Description was changed from

==========
Update telemetry tests to be agnostic to ScrollTopLeftInterop mode

See https://dev.opera.com/articles/fixing-the-scrolltop-bug/
for details of the issue.  It should be possible to run telemetry
with --enable-experimental-web-platform-features.

BUG=157855
==========

to

==========
Update telemetry tests to be agnostic to ScrollTopLeftInterop mode

See https://dev.opera.com/articles/fixing-the-scrolltop-bug/
for details of the issue.  It should be possible to run telemetry
with --enable-experimental-web-platform-features.

BUG=157855
CQ_EXTRA_TRYBOTS=tryserver.chromium.perf:linux_perf_bisect;tryserver.chromium.perf:mac_10_10_perf_bisect;tryserver.chromium.perf:win_perf_bisect;tryserver.chromium.perf:android_nexus5_perf_bisect
==========

[Rick Byers, 2015-10-29 20:01:47]

On 2015/10/29 19:04:13, Sami wrote:
> On 2015/10/29 14:27:43, Rick Byers wrote:
> > On 2015/10/29 10:12:43, Sami wrote:
> > > lgtm, thanks for the fix.
> > > 
> > > I guess document.scrollingElement has been shipped for a while now so this
> > won't
> > > cause problems when run on older reference builds (typically whatever is
the
> > > current stable channel)?
> > 
> > Yes it shipped in M44.  How can I check to see what the current reference
> build
> > is?
> 
> I don't exactly know where it's configured, but a good rule of thumb so far
> seems to be the ~current stable release.

Ok, I guess the extra trybots should fail if we're wrong and the ref builds
don't have scrollingElement yet.  Let's try...

[Rick Byers, 2015-10-29 20:01:51]

The CQ bit was checked by rbyers@chromium.org

[commit-bot: I haz the power, 2015-10-29 20:02:43]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1428733002/1
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1428733002/1

[commit-bot: I haz the power, 2015-10-29 21:11:09]

Committed patchset #1 (id:1)

[commit-bot: I haz the power, 2015-10-29 21:11:58]

Patchset 1 (id:??) landed as
https://crrev.com/a9b90544c07f7e54781c91856d0019e1f3a1567a
Cr-Commit-Position: refs/heads/master@{#356947}