Add 1 second delay before gestures in smoothness.top_25_smooth

Add 1 second delay before gestures in smoothness.top_25_smooth

With GPU rasterization there's a one-time startup cost of doing shader
compiles for skia. The tests were starting their gestures almost
immediately after load, so they were sometimes forced to wait for the
shader compiles to happen before the first gesture.

This test was intended to test scrolling smoothness, not time to first
paint, so adding a delay makes them more consistent. In the real world
these shaders are most likely cached on disk or in memory, so any load
after the browser was running a while should be better.

BUG=648644
Committed: https://crrev.com/7ac9b1b7052cc76fa7b3fef51d2dc447f55991d3
Cr-Commit-Position: refs/heads/master@{#423886}

[jbauman, 2016-09-29 21:12:36]

CL for src perf tryjob to run smoothness.top_25_smooth benchmark on winx64-zen
platform(s)

[jbauman, 2016-09-30 01:28:01]

The CQ bit was checked by jbauman@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-09-30 01:28:21]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-09-30 03:08:08]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-09-30 03:08:09]

Dry run: This issue passed the CQ dry run.

[jbauman, 2016-09-30 20:03:34]

Description was changed from

==========
CL for src perf tryjob to run smoothness.top_25_smooth benchmark on winx64-zen
platform(s)
==========

to

==========
Add 1 second delay before gestures in smoothness.top_25_smooth

With GPU rasterization there's a one-time startup cost of doing shader
compiles for skia. The tests were starting their gestures almost
immediately after load, so they were sometimes forced to wait for the
shader compiles to happen before the first gesture.

This test was intended to test scrolling smoothness, not time to first
paint, so adding a delay makes them more consistent. In the real world
these shaders are most likely cached on disk or in memory, so any load
after the browser was running a while should be better.

BUG=648644
==========

[jbauman, 2016-09-30 20:06:51]

jbauman@chromium.org changed reviewers:
+ sullivan@chromium.org

[sullivan, 2016-09-30 20:23:55]

sullivan@chromium.org changed reviewers:
+ tdresser@chromium.org, vmiura@chromium.org

[sullivan, 2016-09-30 20:23:56]

Adding test owner vmiura and tdresser from input: is this okay with you?

[tdresser, 2016-09-30 20:26:59]

If I understand correctly, the cost that we're avoiding here will be paid in
other metrics, right?

Assuming that's the case (and we'll still see a regression in shader compile
time), this LGTM.

[jbauman, 2016-09-30 22:21:26]

On 2016/09/30 20:26:59, tdresser wrote:
> If I understand correctly, the cost that we're avoiding here will be paid in
> other metrics, right?
> 
> Assuming that's the case (and we'll still see a regression in shader compile
> time), this LGTM.

Yeah, I think shader compile times should be caught be the page_cyclers and
startup.*

[vmiura, 2016-10-04 23:59:45]

On 2016/09/30 22:21:26, jbauman wrote:
> On 2016/09/30 20:26:59, tdresser wrote:
> > If I understand correctly, the cost that we're avoiding here will be paid in
> > other metrics, right?
> > 
> > Assuming that's the case (and we'll still see a regression in shader compile
> > time), this LGTM.
> 
> Yeah, I think shader compile times should be caught be the page_cyclers and
> startup.*

Also LGTM.

[jbauman, 2016-10-07 00:31:07]

sullivan@: OWNERS review?

[sullivan, 2016-10-07 11:46:52]

lgtm

[jbauman, 2016-10-07 15:46:15]

The CQ bit was checked by jbauman@chromium.org

[commit-bot: I haz the power, 2016-10-07 15:46:32]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-10-07 16:51:32]

Description was changed from

==========
Add 1 second delay before gestures in smoothness.top_25_smooth

With GPU rasterization there's a one-time startup cost of doing shader
compiles for skia. The tests were starting their gestures almost
immediately after load, so they were sometimes forced to wait for the
shader compiles to happen before the first gesture.

This test was intended to test scrolling smoothness, not time to first
paint, so adding a delay makes them more consistent. In the real world
these shaders are most likely cached on disk or in memory, so any load
after the browser was running a while should be better.

BUG=648644
==========

to

==========
Add 1 second delay before gestures in smoothness.top_25_smooth

With GPU rasterization there's a one-time startup cost of doing shader
compiles for skia. The tests were starting their gestures almost
immediately after load, so they were sometimes forced to wait for the
shader compiles to happen before the first gesture.

This test was intended to test scrolling smoothness, not time to first
paint, so adding a delay makes them more consistent. In the real world
these shaders are most likely cached on disk or in memory, so any load
after the browser was running a while should be better.

BUG=648644
==========

[commit-bot: I haz the power, 2016-10-07 16:51:34]

Committed patchset #3 (id:40001)

[commit-bot: I haz the power, 2016-10-07 16:53:50]

Description was changed from

==========
Add 1 second delay before gestures in smoothness.top_25_smooth

With GPU rasterization there's a one-time startup cost of doing shader
compiles for skia. The tests were starting their gestures almost
immediately after load, so they were sometimes forced to wait for the
shader compiles to happen before the first gesture.

This test was intended to test scrolling smoothness, not time to first
paint, so adding a delay makes them more consistent. In the real world
these shaders are most likely cached on disk or in memory, so any load
after the browser was running a while should be better.

BUG=648644
==========

to

==========
Add 1 second delay before gestures in smoothness.top_25_smooth

With GPU rasterization there's a one-time startup cost of doing shader
compiles for skia. The tests were starting their gestures almost
immediately after load, so they were sometimes forced to wait for the
shader compiles to happen before the first gesture.

This test was intended to test scrolling smoothness, not time to first
paint, so adding a delay makes them more consistent. In the real world
these shaders are most likely cached on disk or in memory, so any load
after the browser was running a while should be better.

BUG=648644

Committed: https://crrev.com/7ac9b1b7052cc76fa7b3fef51d2dc447f55991d3
Cr-Commit-Position: refs/heads/master@{#423886}
==========

[commit-bot: I haz the power, 2016-10-07 16:53:52]

Patchset 3 (id:??) landed as
https://crrev.com/7ac9b1b7052cc76fa7b3fef51d2dc447f55991d3
Cr-Commit-Position: refs/heads/master@{#423886}