[Webkit-unassigned] [Bug 16777] eliminate JSC::NaN and JSC::Inf

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=16777





--- Comment #10 from Gavin Barraclough <barraclough at apple.com>  2011-06-10 23:44:50 PST ---
(In reply to comment #8)
> (From update of attachment 96833 [details])
> View in context: https://bugs.webkit.org/attachment.cgi?id=96833&action=review
> 
> So the nonInlineNaN() optimization was not helpful?

I'm not measuring a performance difference now.  I imagine this may be the kind of thing that changed with the introduction of the JIT.  Inlining methods that accessed globals used to have a significant impact on Interpreter::privateExecute.

> > Source/JavaScriptCore/runtime/JSGlobalObjectFunctions.cpp:53
> > +#define NaN std::numeric_limits<double>::quiet_NaN()
> > +#define Inf std::numeric_limits<double>::infinity()
> 
> Really?

Copy & paste throughout the file just seemed a bit code bloaty, since it was used a number of times - but I'm happy to expand these out - will do.

G.

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