Richard Biener <rguent...@suse.de> writes: > On Mon, 11 May 2020, Alexander Monakov wrote: > >> On Mon, 11 May 2020, Richard Sandiford wrote: >> >> > Like you say, the idea is that since the operation is commutative and >> > is the same in both vector and scalar form, there's no reason to require >> > any -ffast-math flags. >> >> Note that PR88540 that Richard is referencing uses open-coded x < y ? x : y >> (non-commutative) and we want to use SSE minpd even without -ffast-math, as >> SSE min/max insns match semantics of open-coded ternary operators. >> >> (unlike Arm SIMD, SSE does not have a way to compute fmin/fmax with a >> single instruction in presence of NaNs) > > Indeed. So it looks like for SSE we eventually want phiopt to generate > a COND_EXPR here and a new optabs cond_fmin cond_fmax that could be > used to vectorize this? cond_fmin and cond_fmax can neither be > treated as MIN_EXPR or MAX_EXPR nor fmin/fmax since it is not > commutative.
I know I'm taking the name too literally, but: "cond_foo" in the sense of "IFN_COND_FOO (cond, ..., fallback)" is always supposed to be the equivalent of: cond ? IFN_FOO (...) : fallback So cond_fmin would be the conditional form of fmin. Thanks, Richard
