RESOLVED FIXED 6249
REGRESSION: ~2% remaining from getter/setter change
https://bugs.webkit.org/show_bug.cgi?id=6249
Summary REGRESSION: ~2% remaining from getter/setter change
Maciej Stachowiak
Reported 2005-12-27 01:24:09 PST
On my machine, the getter/setter change caused a regression in JavaScript iBench speed from 1980ms or so to ~2150ms. After the fix for 6083, it is down to 2020. That's way better, but still a regression.
Attachments
patch that gets a bunch of speedups by passing this stuff for activation objects (9.13 KB, patch)
2006-01-05 23:55 PST, Maciej Stachowiak
eric: review+
Darin Adler
Comment 1 2005-12-27 09:56:01 PST
Some of the suggestions in bug 6083 might help.
Maciej Stachowiak
Comment 2 2006-01-05 23:55:49 PST
Created attachment 5509 [details] patch that gets a bunch of speedups by passing this stuff for activation objects
Maciej Stachowiak
Comment 3 2006-01-05 23:56:15 PST
Also fixes 6249
Maciej Stachowiak
Comment 4 2006-01-05 23:56:36 PST
I meant: Also fixes 6395
Maciej Stachowiak
Comment 5 2006-01-05 23:57:31 PST
*** Bug 6395 has been marked as a duplicate of this bug. ***
Eric Seidel (no email)
Comment 6 2006-01-06 00:11:30 PST
Comment on attachment 5509 [details] patch that gets a bunch of speedups by passing this stuff for activation objects Looks good. You missed an ending newline in the very last file in this patch.
Maciej Stachowiak
Comment 7 2006-01-06 00:32:34 PST
Cc'd Maks, this has some interesting general speedups along the lines of what we talked about as well as fixing the slowdown from the getter/setter change.
Alice Liu
Comment 8 2006-01-09 16:27:37 PST
Joost de Valk (AlthA)
Comment 9 2006-01-22 04:56:26 PST
Removing keyword(s) since bug is fixed.
Joost de Valk (AlthA)
Comment 10 2006-01-22 05:00:47 PST
Removing keyword(s) since bug is fixed.
Eric Seidel (no email)
Comment 11 2006-01-31 21:20:39 PST
Removing Regression keyword from bugs already fixed.
Note You need to log in before you can comment on or make changes to this bug.