Bug 5687 - inline more ScopeChain functions for speed
Summary: inline more ScopeChain functions for speed
Status: RESOLVED FIXED
Alias: None
Product: WebKit
Classification: Unclassified
Component: JavaScriptCore (show other bugs)
Version: 420+
Hardware: Mac OS X 10.4
: P2 Normal
Assignee: Darin Adler
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2005-11-09 08:48 PST by Darin Adler
Modified: 2005-11-26 17:27 PST (History)
0 users

See Also:


Attachments
make more functions inline for speed (4.06 KB, patch)
2005-11-09 08:48 PST, Darin Adler
eric: review+
Details | Formatted Diff | Diff

Note You need to log in before you can comment on or make changes to this bug.
Description Darin Adler 2005-11-09 08:48:15 PST
 
Comment 1 Darin Adler 2005-11-09 08:48:39 PST
Created attachment 4650 [details]
make more functions inline for speed
Comment 2 Maciej Stachowiak 2005-11-17 00:12:43 PST
I suspect this might not make a speed difference any more, now that an iterator with inline operations is 
used to traverse the scope chain during execution, instead of making a copy and popping.
Comment 3 Darin Adler 2005-11-23 17:28:30 PST
Comment on attachment 4650 [details]
make more functions inline for speed

While I understand that these functions are no longer as hot as they used to
be, I still think inlining them makes sense -- they are only used in 1 or 2
places and they are performance-critiical places.
Comment 4 Maciej Stachowiak 2005-11-24 00:45:03 PST
Upon further reading I think inlining mark() will almost certainly make a difference but I am not really sure 
about the others - worth measuring and profiling.
Comment 5 Eric Seidel (no email) 2005-11-24 12:12:26 PST
Comment on attachment 4650 [details]
make more functions inline for speed

The code looks fine.  However while you're in there, you might as well clean up
the formatting of ScopeChain::bottom() (spacing) and ScopeChain::pop() (one
line if) to follow our published coding guidelines.