Bug 113528 - REGRESSION: run-perf-tests --profiler= seems to have broken
Summary: REGRESSION: run-perf-tests --profiler= seems to have broken
Status: RESOLVED FIXED
Alias: None
Product: WebKit
Classification: Unclassified
Component: Tools / Tests (show other bugs)
Version: 528+ (Nightly build)
Hardware: Unspecified Unspecified
: P2 Normal
Assignee: Sergio Villar Senin
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2013-03-28 12:33 PDT by Eric Seidel (no email)
Modified: 2015-01-23 06:56 PST (History)
11 users (show)

See Also:

Attachments
Patch (1.87 KB, patch)
2015-01-23 04:11 PST, Sergio Villar Senin 		no flags Details | Formatted Diff | Diff
Patch (3.06 KB, patch)
2015-01-23 05:39 PST, Sergio Villar Senin 		dbates: review+
Details | Formatted Diff | Diff
Show Obsolete (1) View All Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this bug.
Description Eric Seidel (no email) 2013-03-28 12:33:06 PDT 
REGRESSION: run-perf-tests --profiler seems to have broken

--profiler=pprof used to let you use the pprof profiler on linux for example.  Now it no longer seems to work.  Something (probably my fault) broke at some point in recent memory.
Comment 1 Sergio Villar Senin 2015-01-23 04:06:32 PST 
Not sure if the failure is the same than the one reported by Eric but it definitely does not work in GTK (and I guess that also in EFL) when using jhbuild to build.

The problem is that "perf --record" is never called because we just wrap the command with jhbuild-run.
Comment 2 Sergio Villar Senin 2015-01-23 04:11:07 PST 
Created attachment 245221 [details]
Patch
Comment 3 Sergio Villar Senin 2015-01-23 05:39:05 PST 
Created attachment 245224 [details]
Patch

Now with unit test
Comment 4 Daniel Bates 2015-01-23 06:45:04 PST 
Comment on attachment 245224 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=245224&action=review

> Tools/Scripts/webkitpy/port/driver.py:228
> +        wrapper = []

Maybe a better name for this variable would be wrapper_arguments.
Comment 5 Sergio Villar Senin 2015-01-23 06:48:53 PST 
(In reply to comment #4)
> Comment on attachment 245224 [details]
> Patch
> 
> View in context:
> https://bugs.webkit.org/attachment.cgi?id=245224&action=review
> 
> > Tools/Scripts/webkitpy/port/driver.py:228
> > +        wrapper = []
> 
> Maybe a better name for this variable would be wrapper_arguments.

Okidoki
Comment 6 Sergio Villar Senin 2015-01-23 06:56:57 PST 
Committed r179008: <http://trac.webkit.org/changeset/179008>