Add Chrome Endure graph plotting code.

Add Chrome Endure graph plotting code.

Chrome Endure refers to an endurance testing effort to study the effect
of memory bloat in webapps running in chrome.

This code is a fork of chrome's existing code for displaying performance
graphs.  It has become very specific to Chrome Endure, containing features
requested by users of Chrome Endure, and having the ability to display
performance results across time within a given test run (as compared to
the existing chrome perf graphs, which display data across test runs).

In addition to adding new features specific to Chrome Endure, the current
code removes some features that originally existed from chrome's perf
graphing code, which haven't been used for Chrome Endure.

This code has been running for awhile to display Chrome Endure perf
graphs, but hasn't been checked in until now.  Other users want to build on
top of this codebase to add new graphing features specific to Chrome Endure.

BUG=chromium-os:32302
TEST=Verified that graphs work well enough for current Chrome Endure purposes.


Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=153612

[cmp, 2012-08-27 02:42:01]

lgtm with comments, nits, and an offline conversation

Hi Dennis, I understand the overall purpose here.  This is a good change so I
want to see it landed in the perf/ directory asap.  There are some minor nits
that I saw while reviewing that can be fixed.  These issues may appear in the
original files you based your version on, it is still good to fix your copy.

Comments:
- we appear headed for a large 3-way merge.  assuming there are others that want
to modify your code, and your code is based on other code in perf/, is it
possible to avoid a 3-way merge by doing a 2-way merge sooner with your code and
the code already in perf/, so that when others' code lands, we can do another
2-way merge there?

- can we plan to make the endure test results available on our public dashboard?
 i'm interested in helping find a way to make that possible.  it probably means
running your tests on other equipment.  this seems doable.

- i want to speak to you offline about your patch.  you're marked as busy
monday+tuesday, but if you're free then let's discuss for 15 mins or so after
10AM.  otherwise, let's talk on wednesday.

Feel free to land your patch after the nits are fixed.  Can you communicate to
the others that when they land their changes to your code, they should send
reviews to both of us?  Thanks.

http://codereview.chromium.org/10832403/diff/1/chrome/test/functional/perf/en...
File chrome/test/functional/perf/endure_graphs/endure_plotter.js (right):

http://codereview.chromium.org/10832403/diff/1/chrome/test/functional/perf/en...
chrome/test/functional/perf/endure_graphs/endure_plotter.js:232: 
remove blank line

http://codereview.chromium.org/10832403/diff/1/chrome/test/functional/perf/en...
chrome/test/functional/perf/endure_graphs/endure_plotter.js:292: traceName = t
append a semicolon

http://codereview.chromium.org/10832403/diff/1/chrome/test/functional/perf/en...
chrome/test/functional/perf/endure_graphs/endure_plotter.js:304: return true
append a semicolon

http://codereview.chromium.org/10832403/diff/1/chrome/test/functional/perf/en...
chrome/test/functional/perf/endure_graphs/endure_plotter.js:374: errorMessages
+= 'No data for the measurement to compare against.<br>'
append a semicolon

[dennis_jeffrey, 2012-08-27 23:37:13]

Chase,

Thank you very much for the review of this really large CL, and the insightful
comments.  I addressed the nits and will commit this now.  Some more comments
inline (which we already discussed in our offline meeting earlier today):

On 2012/08/27 02:42:01, cmp wrote:
> lgtm with comments, nits, and an offline conversation
> 
> Hi Dennis, I understand the overall purpose here.  This is a good change so I
> want to see it landed in the perf/ directory asap.  There are some minor nits
> that I saw while reviewing that can be fixed.  These issues may appear in the
> original files you based your version on, it is still good to fix your copy.
> 
> Comments:
> - we appear headed for a large 3-way merge.  assuming there are others that
want
> to modify your code, and your code is based on other code in perf/, is it
> possible to avoid a 3-way merge by doing a 2-way merge sooner with your code
and
> the code already in perf/, so that when others' code lands, we can do another
> 2-way merge there?
> 

I agree that having the current fork of the graphing code is going to lead to
problematic 3-way merges (since there is still some shared code between this
fork and the original codebase, even though there are lots of new chrome
endure-specific features in this CL).  I'll start planning how to integrate
these changes into the existing codebase so that we can get rid of this fork
(http://code.google.com/p/chromium-os/issues/detail?id=33934).

> - can we plan to make the endure test results available on our public
dashboard?
>  i'm interested in helping find a way to make that possible.  it probably
means
> running your tests on other equipment.  this seems doable.
> 

Yes, I'd like to make the Chrome Endure results/graphs publicly viewable.  This
is being tracked in a general issue to get Chrome Endure fully open-source:
http://code.google.com/p/chromium-os/issues/detail?id=32302

> - i want to speak to you offline about your patch.  you're marked as busy
> monday+tuesday, but if you're free then let's discuss for 15 mins or so after
> 10AM.  otherwise, let's talk on wednesday.
> 
> Feel free to land your patch after the nits are fixed.  Can you communicate to
> the others that when they land their changes to your code, they should send
> reviews to both of us?  Thanks.

Yes, I'll make sure that you're copied on any code reviews pertaining to the
chrome endure graphs.

> 
>
http://codereview.chromium.org/10832403/diff/1/chrome/test/functional/perf/en...
> File chrome/test/functional/perf/endure_graphs/endure_plotter.js (right):
> 
>
http://codereview.chromium.org/10832403/diff/1/chrome/test/functional/perf/en...
> chrome/test/functional/perf/endure_graphs/endure_plotter.js:232: 
> remove blank line
> 
>
http://codereview.chromium.org/10832403/diff/1/chrome/test/functional/perf/en...
> chrome/test/functional/perf/endure_graphs/endure_plotter.js:292: traceName = t
> append a semicolon
> 
>
http://codereview.chromium.org/10832403/diff/1/chrome/test/functional/perf/en...
> chrome/test/functional/perf/endure_graphs/endure_plotter.js:304: return true
> append a semicolon
> 
>
http://codereview.chromium.org/10832403/diff/1/chrome/test/functional/perf/en...
> chrome/test/functional/perf/endure_graphs/endure_plotter.js:374: errorMessages
> += 'No data for the measurement to compare against.<br>'
> append a semicolon

[dennis_jeffrey, 2012-08-27 23:37:41]

http://codereview.chromium.org/10832403/diff/1/chrome/test/functional/perf/en...
File chrome/test/functional/perf/endure_graphs/endure_plotter.js (right):

http://codereview.chromium.org/10832403/diff/1/chrome/test/functional/perf/en...
chrome/test/functional/perf/endure_graphs/endure_plotter.js:232: 
On 2012/08/27 02:42:01, cmp wrote:
> remove blank line

Done.

http://codereview.chromium.org/10832403/diff/1/chrome/test/functional/perf/en...
chrome/test/functional/perf/endure_graphs/endure_plotter.js:292: traceName = t
On 2012/08/27 02:42:01, cmp wrote:
> append a semicolon

Done.

http://codereview.chromium.org/10832403/diff/1/chrome/test/functional/perf/en...
chrome/test/functional/perf/endure_graphs/endure_plotter.js:304: return true
On 2012/08/27 02:42:01, cmp wrote:
> append a semicolon

Done.

http://codereview.chromium.org/10832403/diff/1/chrome/test/functional/perf/en...
chrome/test/functional/perf/endure_graphs/endure_plotter.js:374: errorMessages
+= 'No data for the measurement to compare against.<br>'
On 2012/08/27 02:42:01, cmp wrote:
> append a semicolon

Done.

[commit-bot: I haz the power, 2012-08-27 23:42:55]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/dennisjeffrey@chromium.org/10832403/5001

[commit-bot: I haz the power, 2012-08-28 02:59:46]

Change committed as 153612