Bug 21994 - r37894 causes a 10% regression on richards.
Summary: r37894 causes a 10% regression on richards.
Status: RESOLVED INVALID
Alias: None
Product: WebKit
Classification: Unclassified
Component: JavaScriptCore (show other bugs)
Version: 528+ (Nightly build)
Hardware: Mac OS X 10.5
: P2 Normal
Assignee: Nobody
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2008-10-31 01:42 PDT by Gavin Barraclough
Modified: 2008-12-02 14:57 PST (History)
1 user (show)

See Also:

Attachments
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 Gavin Barraclough 2008-10-31 01:42:04 PDT 
r37894 causes a 10% regression on richards.  Caused by:
-    #define CTI_SAMPLER
+    #define CTI_SAMPLER 0
Measured on my Mac Pro.
Comment 1 Geoffrey Garen 2008-10-31 11:19:31 PDT 
I am an idiot.
Comment 2 Gavin Barraclough 2008-10-31 11:46:20 PDT 
lol
Comment 3 Geoffrey Garen 2008-12-02 14:53:41 PST 
Gavin mentioned in person that he saw this effect disappear in later revisions.
Comment 4 Geoffrey Garen 2008-12-02 14:55:07 PST 
I tried testing against TOT, but changing to "#define CTI_SAMPLER" causes the infamous GCC 4.2 hang.
Comment 5 Geoffrey Garen 2008-12-02 14:57:57 PST 
It might be nice to do a little more investigation when GCC 4.2 is in a more cooperative mood. For now, though, I think we can close this bug based on Gavin's existing data.