On Fri, 8 May 2020, Richard Biener wrote: > > Currently we fail to optimize those which are used when MIN/MAX_EXPR > cannot be used for FP values but the target has IEEE conforming > implementations. i386 ieee_s{min,max} patterns are definitely not IEEE-compliant, their comment alludes to that: ;; These versions of the min/max patterns implement exactly the operations ;; min = (op1 < op2 ? op1 : op2) ;; max = (!(op1 < op2) ? op1 : op2) ;; Their operands are not commutative, and thus they may be used in the ;; presence of -0.0 and NaN. I don't understand why the patch is correct if the IFNs refer to fully IEEE-compliant operations (which is in itself a bit ambiguous w.r.t behavior when exactly one operand is a NaN). Am I missing something? Alexander
Re: [PATCH] make minmax detection work with FMIN/FMAX IFNs
Alexander Monakov via Gcc-patches Fri, 08 May 2020 06:46:25 -0700
- [PATCH] make minmax detection work with ... Richard Biener
- Re: [PATCH] make minmax detection w... Uros Bizjak via Gcc-patches
- Re: [PATCH] make minmax detection w... Alexander Monakov via Gcc-patches
- Re: [PATCH] make minmax detecti... Uros Bizjak via Gcc-patches
- Re: [PATCH] make minmax det... Alexander Monakov via Gcc-patches
- Re: [PATCH] make minmax... Richard Biener
- Re: [PATCH] make m... Uros Bizjak via Gcc-patches
- Re: [PATCH] ma... Alexander Monakov via Gcc-patches
- Re: [PATCH] make m... Richard Sandiford
- Re: [PATCH] ma... Alexander Monakov via Gcc-patches
- Re: [PATCH... Richard Biener
- Re: [PATCH... Richard Sandiford
- Re: [PATCH] make m... Joseph Myers