[PATCH] [ARM] generate VMAXNM/VMINNM for a compare followed by a select (in safe math mode too)

Renato Golin renato.golin at linaro.org
Wed May 6 02:42:03 PDT 2015


On 5 May 2015 at 20:48, Artyom Skrobov <artyom.skrobov at arm.com> wrote:
> Yes, I did put together some back-of-the-envelope tables of the possibilities, to check that the semantics are preserved by each of the switch blocks.

Excellent, thanks!


> What matters for the transformability is whether the operands are NaN, it doesn't matter whether they're constants.
> You're right that my comments there are slightly inaccurate -- they should say "the non-NaN should be RHS" and "the non-NaN should be LHS" instead.
> Thanks for catching that!  I'll fix the comments.

Makes sense.

cheers,
--renato



More information about the llvm-commits mailing list