[llvm] [InstCombine] Do not require nsz for reassociating fmul (PR #93698)

[Jay Foad via llvm-commits]

https://github.com/jayfoad updated https://github.com/llvm/llvm-project/pull/93698

>From 64399ddd94053354567a04812d65639e42b2ebcd Mon Sep 17 00:00:00 2001
From: Jay Foad <jay.foad at amd.com>
Date: Wed, 29 May 2024 16:09:49 +0100
Subject: [PATCH 1/2] [InstCombine] Do not require nsz for reassociating fmul

The sign of the result of fmul is always the xor of the signs of the
operands. Reassociation should not affect that.
---
 llvm/lib/IR/Instruction.cpp                                  | 1 +
 llvm/test/Transforms/InstCombine/2006-10-26-VectorReassoc.ll | 3 +--
 llvm/test/Transforms/InstCombine/fdiv.ll                     | 3 +--
 3 files changed, 3 insertions(+), 4 deletions(-)

diff --git a/llvm/lib/IR/Instruction.cpp b/llvm/lib/IR/Instruction.cpp
index 29272e627a1d1..8ad97f34c1719 100644
--- a/llvm/lib/IR/Instruction.cpp
+++ b/llvm/lib/IR/Instruction.cpp
@@ -1204,6 +1204,7 @@ bool Instruction::isAssociative() const {
 
   switch (Opcode) {
   case FMul:
+    return cast<FPMathOperator>(this)->hasAllowReassoc();
   case FAdd:
     return cast<FPMathOperator>(this)->hasAllowReassoc() &&
            cast<FPMathOperator>(this)->hasNoSignedZeros();
diff --git a/llvm/test/Transforms/InstCombine/2006-10-26-VectorReassoc.ll b/llvm/test/Transforms/InstCombine/2006-10-26-VectorReassoc.ll
index fb860a5e7bdf3..c18e149a7f182 100644
--- a/llvm/test/Transforms/InstCombine/2006-10-26-VectorReassoc.ll
+++ b/llvm/test/Transforms/InstCombine/2006-10-26-VectorReassoc.ll
@@ -38,8 +38,7 @@ define <4 x float> @test_fmul_reassoc_nsz(<4 x float> %V) {
 ; TODO: This doesn't require 'nsz'.  It should fold to V * { 1.0, 4.0e+05, -9.0, 16.0 }
 define <4 x float> @test_fmul_reassoc(<4 x float> %V) {
 ; CHECK-LABEL: @test_fmul_reassoc(
-; CHECK-NEXT:     [[TMP1:%.*]] = fmul reassoc <4 x float> [[V:%.*]], <float 1.000000e+00, float 2.000000e+00, float 3.000000e+00, float 4.000000e+00>
-; CHECK-NEXT:     [[TMP2:%.*]] = fmul reassoc <4 x float> [[TMP1]], <float 1.000000e+00, float 2.000000e+05, float -3.000000e+00, float 4.000000e+00>
+; CHECK-NEXT:     [[TMP1:%.*]] = fmul reassoc <4 x float> [[V:%.*]], <float 1.000000e+00, float 4.000000e+05, float -9.000000e+00, float 1.600000e+01>
 ; CHECK-NEXT:     ret <4 x float> [[TMP2]]
         %Y = fmul reassoc <4 x float> %V, < float 1.000000e+00, float 2.000000e+00, float 3.000000e+00, float 4.000000e+00 >
         %Z = fmul reassoc <4 x float> %Y, < float 1.000000e+00, float 2.000000e+05, float -3.000000e+00, float 4.000000e+00 >
diff --git a/llvm/test/Transforms/InstCombine/fdiv.ll b/llvm/test/Transforms/InstCombine/fdiv.ll
index ca11685c98417..2e5ea29cecf6a 100644
--- a/llvm/test/Transforms/InstCombine/fdiv.ll
+++ b/llvm/test/Transforms/InstCombine/fdiv.ll
@@ -524,8 +524,7 @@ define <2 x float> @div_constant_dividend2_reassoc_only(<2 x float> %x) {
 
 define <2 x float> @div_constant_dividend3(<2 x float> %x) {
 ; CHECK-LABEL: @div_constant_dividend3(
-; CHECK-NEXT:    [[TMP1:%.*]] = fmul reassoc arcp <2 x float> [[X:%.*]], <float 1.500000e+01, float -7.000000e+00>
-; CHECK-NEXT:    [[T2:%.*]] = fmul reassoc arcp <2 x float> [[TMP1]], <float 0x3FD5555560000000, float 0x3FC24924A0000000>
+; CHECK-NEXT:    [[T2:%.*]] = fmul reassoc arcp <2 x float> [[X:%.*]], <float 5.000000e+00, float -1.000000e+00>
 ; CHECK-NEXT:    ret <2 x float> [[T2]]
 ;
   %t1 = fdiv <2 x float> <float 3.0e0, float 7.0e0>, %x

>From 1f1b12fec58a5397377deb61bd1cafb7ddd3c1e2 Mon Sep 17 00:00:00 2001
From: Jay Foad <jay.foad at amd.com>
Date: Thu, 30 May 2024 05:53:35 +0100
Subject: [PATCH 2/2] Remove TODO

---
 llvm/test/Transforms/InstCombine/2006-10-26-VectorReassoc.ll | 1 -
 1 file changed, 1 deletion(-)

diff --git a/llvm/test/Transforms/InstCombine/2006-10-26-VectorReassoc.ll b/llvm/test/Transforms/InstCombine/2006-10-26-VectorReassoc.ll
index c18e149a7f182..a9c8a1cab7d90 100644
--- a/llvm/test/Transforms/InstCombine/2006-10-26-VectorReassoc.ll
+++ b/llvm/test/Transforms/InstCombine/2006-10-26-VectorReassoc.ll
@@ -35,7 +35,6 @@ define <4 x float> @test_fmul_reassoc_nsz(<4 x float> %V) {
 }
 
 ; (V * C1) * C2 => V * (C1 * C2)
-; TODO: This doesn't require 'nsz'.  It should fold to V * { 1.0, 4.0e+05, -9.0, 16.0 }
 define <4 x float> @test_fmul_reassoc(<4 x float> %V) {
 ; CHECK-LABEL: @test_fmul_reassoc(
 ; CHECK-NEXT:     [[TMP1:%.*]] = fmul reassoc <4 x float> [[V:%.*]], <float 1.000000e+00, float 4.000000e+05, float -9.000000e+00, float 1.600000e+01>