Bug 85011 - Remove RegisterFile::end()/m_end
Summary: Remove RegisterFile::end()/m_end
Status: REOPENED
Alias: None
Product: WebKit
Classification: Unclassified
Component: New Bugs (show other bugs)
Version: 528+ (Nightly build)
Hardware: Unspecified Unspecified
: P2 Normal
Assignee: Oliver Hunt
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2012-04-26 15:26 PDT by Oliver Hunt
Modified: 2012-05-02 12:19 PDT (History)
4 users (show)

See Also:

Attachments
Patch (28.34 KB, patch)
2012-04-26 15:54 PDT, Oliver Hunt 		no flags Details | Formatted Diff | Diff
Remove now bogus NDEBUG ifdef (29.06 KB, patch)
2012-04-26 16:18 PDT, Oliver Hunt 		no flags Details | Formatted Diff | Diff
Show Obsolete (2) 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 Oliver Hunt 2012-04-26 15:26:15 PDT 
Remove RegisterFile::end()/m_end
Comment 1 Oliver Hunt 2012-04-26 15:54:40 PDT 
Created attachment 139087 [details]
Patch
Comment 2 Early Warning System Bot 2012-04-26 16:07:45 PDT 
Comment on attachment 139087 [details]
Patch

Attachment 139087 [details] did not pass qt-wk2-ews (qt):
Output: http://queues.webkit.org/results/12534029
Comment 3 Build Bot 2012-04-26 16:10:44 PDT 
Comment on attachment 139087 [details]
Patch

Attachment 139087 [details] did not pass mac-ews (mac):
Output: http://queues.webkit.org/results/12548026
Comment 4 Philippe Normand 2012-04-26 16:15:44 PDT 
Comment on attachment 139087 [details]
Patch

Attachment 139087 [details] did not pass gtk-ews (gtk):
Output: http://queues.webkit.org/results/12552041
Comment 5 Oliver Hunt 2012-04-26 16:18:30 PDT 
Created attachment 139090 [details]
Remove now bogus NDEBUG ifdef
Comment 6 Oliver Hunt 2012-04-26 17:06:19 PDT 
Committed r115388: <http://trac.webkit.org/changeset/115388>
Comment 8 Oliver Hunt 2012-04-28 11:47:47 PDT 
(In reply to comment #7)
> This patch seems to have caused massive swings in number of tests in http://webkit-perf.appspot.com on all JSC using (Mac Lion and Qt) bots.
> 
> 10% regression in Dromaeo/dom-traverse:
> 
> http://webkit-perf.appspot.com/graph.html#tests=[[40020,2001,32196]]&sel=1335481319023.587,1335500151119.3953,1212.7659574468084,2680.851063829787&displayrange=7&datatype=running
> 
> 50% regression in DOM/Template
> 
> http://webkit-perf.appspot.com/graph.html#tests=[[110,2001,32196]]&sel=1335480730398.3186,1335502455543.162,6.808510638297875,30.638297872340424&displayrange=7&datatype=running
> 
> 50% progression in Bindings/event-target-wrapper (Lion bot only, not Qt bots)
> 
> http://webkit-perf.appspot.com/graph.html#tests=[[8062,2001,32196]]&sel=1335478825629.3184,1335502278508.2844,0,1468.0851063829787&displayrange=7&datatype=running

That is certainly sad making.  Could you file a separate bug on that, assign it to me -- i'll look into it on monday.  Doesn't make sense that this would cause any variation (if anything it should reduce variation, and improve overall performance.  grumble.
Comment 9 Antti Koivisto 2012-04-29 01:38:03 PDT 
Filed bug 85155.
Comment 10 Oliver Hunt 2012-05-02 12:17:36 PDT 
Reverted r115388 for reason:

This caused many weird performance problems, and needs to be landed in pieces.

Committed r115861: <http://trac.webkit.org/changeset/115861>