telemetry: extend key_silk_cases.json

telemetry: extend key_silk_cases.json

BUG=310389
NOTRY=True

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=243457

[ernstm, 2013-12-18 18:27:20]

I've extended the silk page set. I needed to do several workarounds to get it
recorded and running:
- patch record_wpr to work with file urls (crbug.com/329042)
- remove all "name" attributes (crbug.com/329050); can add back when bug is
fixed.
- move all file urls to the end of the page set (crbug.com/329349)
- disable one page that does not work on Chrome ToT

I have tested it with silk benchmark on Linux (z620), N7v2, N10, N4 and GN.

PTAL.

[tonyg, 2013-12-19 02:21:16]

lgtm

https://codereview.chromium.org/117833002/diff/1/tools/perf/page_sets/key_sil...
File tools/perf/page_sets/key_silk_cases.json (right):

https://codereview.chromium.org/117833002/diff/1/tools/perf/page_sets/key_sil...
tools/perf/page_sets/key_silk_cases.json:94: "smoothness": { "action": "wait",
"seconds": 4 }
These sleeps make me so sad. I can rest assured they are temporary and we have a
plan to get rid of them, right?

[nduca, 2013-12-19 04:09:12]

The sleeps lgtm. For many of these, you just want it to run awhile.

[ernstm, 2013-12-19 05:40:45]

https://codereview.chromium.org/117833002/diff/1/tools/perf/page_sets/key_sil...
File tools/perf/page_sets/key_silk_cases.json (right):

https://codereview.chromium.org/117833002/diff/1/tools/perf/page_sets/key_sil...
tools/perf/page_sets/key_silk_cases.json:94: "smoothness": { "action": "wait",
"seconds": 4 }
On 2013/12/19 02:21:16, tonyg wrote:
> These sleeps make me so sad. I can rest assured they are temporary and we have
a
> plan to get rid of them, right?

Those are like the wait action in tough_canvas cases: the pages are animating
all the time and we just need to wait a little to get frames.
Increasing/decreasing the wait time increases/decreases the number of frames. I
wouldn't expect this to cause flakiness.

[commit-bot: I haz the power, 2013-12-19 05:41:54]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/ernstm@chromium.org/117833002/1

[commit-bot: I haz the power, 2013-12-19 06:03:05]

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

[ernstm, 2013-12-19 18:25:38]

chromium_presubmit throws and error because of a license header mismatch in
polymer.min.js and web-animations.js. Looks like presubmit only accepts
Chromium's license header. Any ideas what to do?

[abarth-chromium, 2014-01-07 00:50:02]

On 2013/12/19 18:25:38, Manfred Ernst wrote:
> chromium_presubmit throws and error because of a license header mismatch in
> polymer.min.js and web-animations.js. Looks like presubmit only accepts
> Chromium's license header. Any ideas what to do?

What's the status on this CL?  I'd like to build upon it and add some more
polymer-based cases.

[nduca, 2014-01-07 19:39:07]

How about this?
1. remove the pages in this CL that need polymer so we can land the rest
2. take the polymer-coupled cases and host them on the internet somewhere, then
point the pageset at them in a second CL?

Abarth was actually working on (2) already for a drawer example. Maybe you two
can quickly chat in person (you know where each other sit right?) about how to
get that done? I know Adam, you were poking at getting your polymer-based
content hosted on gh pages, so maybe its just a matter of pushing the ink-button
demo and font wipe demos into your workflow...

[nduca, 2014-01-07 21:09:30]

The other option is to do what https://codereview.chromium.org/120753003/ and
use polymer's vulcanize tool (https://github.com/Polymer/vulcanize) to flatten
the polymer-coupled demos to a single compiled html file, then check that in.
Even if you go that route though, its probably worth separating that out to
another cl so the basic ones can land first.

[ernstm, 2014-01-07 23:54:30]

On 2014/01/07 21:09:30, nduca wrote:
> The other option is to do what https://codereview.chromium.org/120753003/ and
> use polymer's vulcanize tool (https://github.com/Polymer/vulcanize) to flatten
> the polymer-coupled demos to a single compiled html file, then check that in.
> Even if you go that route though, its probably worth separating that out to
> another cl so the basic ones can land first.

It was only the ink-button page that needed the polymer files. That page was
disabled anyway, because it didn't work on ToT (before the holidays). We can
take care of that page in a separate patch. Will commit now.

[commit-bot: I haz the power, 2014-01-07 23:57:05]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/ernstm@chromium.org/117833002/50001

[commit-bot: I haz the power, 2014-01-08 01:03:25]

Retried try job too often on win_rel for step(s) net_unittests
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=win_rel&nu...

[commit-bot: I haz the power, 2014-01-08 01:08:05]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/ernstm@chromium.org/117833002/50001

[commit-bot: I haz the power, 2014-01-08 01:10:58]

Change committed as 243457