[all-commits] [llvm/llvm-project] 76e9c2: [FPEnv] Generate constrained FP comparisons from c...

[Ulrich Weigand via All-commits]

  Branch: refs/heads/master
  Home:   https://github.com/llvm/llvm-project
  Commit: 76e9c2a9870e36415eb343d28942a42296f85597
      https://github.com/llvm/llvm-project/commit/76e9c2a9870e36415eb343d28942a42296f85597
  Author: Ulrich Weigand <ulrich.weigand at de.ibm.com>
  Date:   2020-01-10 (Fri, 10 Jan 2020)

  Changed paths:
    M clang/lib/CodeGen/CGExprScalar.cpp
    A clang/test/CodeGen/fpconstrained-cmp-double.c
    A clang/test/CodeGen/fpconstrained-cmp-float.c
    M llvm/include/llvm/IR/IRBuilder.h

  Log Message:
  -----------
  [FPEnv] Generate constrained FP comparisons from clang

Update the IRBuilder to generate constrained FP comparisons in
CreateFCmp when IsFPConstrained is true, similar to the other
places in the IRBuilder.

Also, add a new CreateFCmpS to emit signaling FP comparisons,
and use it in clang where comparisons are supposed to be signaling
(currently, only when emitting code for the <, <=, >, >= operators).

Note that there is currently no way to add fast-math flags to a
constrained FP comparison, since this is implemented as an intrinsic
call that returns a boolean type, and FMF are only allowed for calls
returning a floating-point type. However, given the discussion around
https://bugs.llvm.org/show_bug.cgi?id=42179, it seems that FCmp itself
really shouldn't have any FMF either, so this is probably OK.

Reviewed by: craig.topper

Differential Revision: https://reviews.llvm.org/D71467