[llvm] r176292 - Brag about function call vectorization in the docs.

[Hal Finkel]

----- Original Message -----
> From: "Duncan Sands" <baldrick at free.fr>
> To: llvm-commits at cs.uiuc.edu
> Sent: Friday, March 1, 2013 8:59:56 AM
> Subject: Re: [llvm] r176292 - Brag about function call vectorization in the	docs.
> 
> Hi Hal,
> 
> >>> On that thought, we might want some kind of analysis that can
> >>> determine if errno from some math function is unused so that we
> >>> can optimize regardless of the conformance mode when possible.
> >>
> >> That's tricky though, errno is always a global variable but the
> >> way
> >> it's accessed varies from operating system to operating system.
> >
> > I know, this is why I've not yet done this myself ;) -- We'd
> > probably need some kind of errno-recognition framework. Thinking
> > about it, it might make more sense for clang to recognize errno
> > access and tag it somehow.
> 
> one traditional method that LLVM front-ends use to indicate that a
> math function
> does not set errno, is to plunk the readonly or readnone attribute on
> its
> declaration.  This is a bit naughty but works pretty well in practice
> IIRC, and
> is independent of fast-math.

Hi Duncan,

Good point; and I think this is fine for intrinsics (and when compiling in fast-math mode more generally), but is not quite right for general libm calls.

Thanks again,
Hal

> 
> Ciao, Duncan.
> _______________________________________________
> llvm-commits mailing list
> llvm-commits at cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits
>