[Webkit-unassigned] [Bug 27762] [v8] Cache atomic strings in externalized v8 strings

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=27762





--- Comment #19 from Christian Plesner Hansen <christian.plesner.hansen at gmail.com>  2009-07-29 01:04:00 PDT ---
(In reply to comment #17)
> (From update of attachment 33628 [details])
> If what you say is true, and that add() would show up just as much for JSC, it
> seems we should fix this in a different way which fixes it for all JS engines.
> :)  Hard to know without seeing the profile in question.

I agree, a solution that covers both JSC and v8 would be better.  Just as an
experiment I tried replacing my changes with a naive atomic string cache in
StringImpls and it makes my getElementById microbenchmark ~8% faster.  So that
might be worth doing if the cost of an extra word per StringImpl is acceptable.

I've only tried it in webkit/v8 though because webkit/jsc doesn't run on linux
(as far as I know).

-- 
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.