Extend the BigJs benchmark: add a YouTube page.

Extend the BigJs benchmark: add a YouTube page.

The page uses a big parsing blocking script and is thus interesting for the
BigJs benchmark.

R=skyostil@chromium.org
BUG=

Committed: https://crrev.com/d14ac22f2e83cc15556d02e944b80955421172c8
Cr-Commit-Position: refs/heads/master@{#324038}

[marja, 2015-04-07 10:52:50]

ptal

How does the cloud storage thing for the wpr work nowadays? I remember it doing
something magical (uploading files to the cloud storage) when I "git cl upload",
but now it didn't.

[Sami, 2015-04-07 11:07:53]

https://codereview.chromium.org/1060563005/diff/20001/tools/perf/page_sets/bi...
File tools/perf/page_sets/big_js.py (right):

https://codereview.chromium.org/1060563005/diff/20001/tools/perf/page_sets/bi...
tools/perf/page_sets/big_js.py:21: # crbug.com/403735).
Hmm, interesting, I didn't know about this! Do you think this is a problem other
benchmarks like [1] or are we doing something special with big_js?

[1]
https://code.google.com/p/chromium/codesearch#chromium/src/tools/perf/page_se...

https://codereview.chromium.org/1060563005/diff/20001/tools/perf/page_sets/da...
File tools/perf/page_sets/data/big_js.json (right):

https://codereview.chromium.org/1060563005/diff/20001/tools/perf/page_sets/da...
tools/perf/page_sets/data/big_js.json:8: "big_js_001.wpr": [
I think we'll need to figure out why big_js_001.wpr.sha1 isn't making it into
this patch and whether it's even getting generated for you locally?

[marja, 2015-04-07 11:59:58]

Ahh, now that I add the sha1 I also get prompted to upload stuff to the cloud
storage. Got it.

https://codereview.chromium.org/1060563005/diff/20001/tools/perf/page_sets/bi...
File tools/perf/page_sets/big_js.py (right):

https://codereview.chromium.org/1060563005/diff/20001/tools/perf/page_sets/bi...
tools/perf/page_sets/big_js.py:21: # crbug.com/403735).
On 2015/04/07 11:07:53, Sami wrote:
> Hmm, interesting, I didn't know about this! Do you think this is a problem
other
> benchmarks like [1] or are we doing something special with big_js?
> 
> [1]
>
https://code.google.com/p/chromium/codesearch#chromium/src/tools/perf/page_se...


It's generally probably not a problem since most of the benchmarks contain a
bunch of pages and run all of them.

BigJs used to contain only 1, in which case this was needed, and I want to keep
it there, so that I can run "youtube + foo" for example and I don't need to run
"youtube + someotherslowpage".

https://codereview.chromium.org/1060563005/diff/20001/tools/perf/page_sets/da...
File tools/perf/page_sets/data/big_js.json (right):

https://codereview.chromium.org/1060563005/diff/20001/tools/perf/page_sets/da...
tools/perf/page_sets/data/big_js.json:8: "big_js_001.wpr": [
On 2015/04/07 11:07:53, Sami wrote:
> I think we'll need to figure out why big_js_001.wpr.sha1 isn't making it into
> this patch and whether it's even getting generated for you locally?

User error.

[Sami, 2015-04-07 12:59:05]

Great, lgtm.

https://codereview.chromium.org/1060563005/diff/20001/tools/perf/page_sets/bi...
File tools/perf/page_sets/big_js.py (right):

https://codereview.chromium.org/1060563005/diff/20001/tools/perf/page_sets/bi...
tools/perf/page_sets/big_js.py:21: # crbug.com/403735).
On 2015/04/07 11:59:58, marja wrote:
> On 2015/04/07 11:07:53, Sami wrote:
> > Hmm, interesting, I didn't know about this! Do you think this is a problem
> other
> > benchmarks like [1] or are we doing something special with big_js?
> > 
> > [1]
> >
>
https://code.google.com/p/chromium/codesearch#chromium/src/tools/perf/page_se...
> 
> 
> It's generally probably not a problem since most of the benchmarks contain a
> bunch of pages and run all of them.
> 
> BigJs used to contain only 1, in which case this was needed, and I want to
keep
> it there, so that I can run "youtube + foo" for example and I don't need to
run
> "youtube + someotherslowpage".

Gotcha, thanks.

[marja, 2015-04-07 13:03:41]

The CQ bit was checked by marja@chromium.org

[commit-bot: I haz the power, 2015-04-07 13:03:53]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1060563005/40001

[marja, 2015-04-07 13:07:05]

That said, the renderer reuse problem might occur in file:// based benchmarks -
you might want to check whether tough_scheduling_cases is affected. Maybe we
just use one renderer all the time and cold runs are not really cold.

[commit-bot: I haz the power, 2015-04-07 14:15:04]

Committed patchset #3 (id:40001)

[commit-bot: I haz the power, 2015-04-07 14:15:50]

Patchset 3 (id:??) landed as
https://crrev.com/d14ac22f2e83cc15556d02e944b80955421172c8
Cr-Commit-Position: refs/heads/master@{#324038}