[llvm-commits] Please review: FMA3 instructions set

[Evan Cheng]

On May 22, 2012, at 9:53 AM, Anton Korobeynikov <anton at korobeynikov.info> wrote:

>> Do I understand correctly that this patch lowers fadd + fmul to fma by default?  We want it to be easy for LLVM to generate fma when it is beneficial to performance, but we can't simply naively lower to it everywhere and still conform to language semantics.  Someone else can speak to what policy should be here, but at the very least we will need to have an option to block fma formation.
> One can check how fma is implemented on ARM. In general - stuff should
> be guarded by NoExcessPrecision flag or something like this

Some of us have serious concerns whether basing on that predicate is correct, especially since NoExcessPrecision is false by default (which allows fma). We need to add support for #pragma STDC FP_CONTRACT to Clang to allow fmul + fadd within expressions to be translated into fma's.

Evan

> 
> -- 
> With best regards, Anton Korobeynikov
> Faculty of Mathematics and Mechanics, Saint Petersburg State University
> 
> _______________________________________________
> llvm-commits mailing list
> llvm-commits at cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits