[x86] enable machine combiner reassociations for scalar single-precision multiplies

llvm-svn: 241752

4 files changed, 31 insertions, 8 deletions

diff --git a/llvm/lib/Target/X86/X86InstrInfo.cpp b/llvm/lib/Target/X86/X86InstrInfo.cpp
index dd2d04dbe49..5484ae91855 100644
--- a/llvm/lib/Target/X86/X86InstrInfo.cpp
+++ b/llvm/lib/Target/X86/X86InstrInfo.cpp
@@ -6405,11 +6405,13 @@ static bool hasReassocSibling(const MachineInstr &Inst, bool &Commuted) {
 
 // TODO: There are many more machine instruction opcodes to match:
 //       1. Other data types (double, integer, vectors)
-//       2. Other math / logic operations (mul, and, or)
+//       2. Other math / logic operations (and, or)
 static bool isAssociativeAndCommutative(unsigned Opcode) {
   switch (Opcode) {
-  case X86::VADDSSrr:
   case X86::ADDSSrr:
+  case X86::VADDSSrr:
+  case X86::MULSSrr:
+  case X86::VMULSSrr:
     return true;
   default:
     return false;
diff --git a/llvm/test/CodeGen/X86/fdiv-combine.ll b/llvm/test/CodeGen/X86/fdiv-combine.ll
index 279bb0624ac..8adfde7f3f3 100644
--- a/llvm/test/CodeGen/X86/fdiv-combine.ll
+++ b/llvm/test/CodeGen/X86/fdiv-combine.ll
@@ -1,4 +1,4 @@
-; RUN: llc < %s -mtriple=x86_64-unknown-unknown | FileCheck %s
+; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mcpu=x86-64 | FileCheck %s
 
 ; Anything more than one division using a single divisor operand
 ; should be converted into a reciprocal and multiplication.
@@ -17,9 +17,9 @@ define float @div2_arcp(float %x, float %y, float %z) #0 {
 ; CHECK:       # BB#0:
 ; CHECK-NEXT:    movss {{.*#+}} xmm3 = mem[0],zero,zero,zero
 ; CHECK-NEXT:    divss %xmm2, %xmm3
-; CHECK-NEXT:    mulss %xmm3, %xmm0
 ; CHECK-NEXT:    mulss %xmm1, %xmm0
 ; CHECK-NEXT:    mulss %xmm3, %xmm0
+; CHECK-NEXT:    mulss %xmm3, %xmm0
 ; CHECK-NEXT:    retq
   %div1 = fdiv arcp float %x, %z
   %mul = fmul arcp float %div1, %y
diff --git a/llvm/test/CodeGen/X86/machine-combiner.ll b/llvm/test/CodeGen/X86/machine-combiner.ll
index 545decb0311..2286da7e94d 100644
--- a/llvm/test/CodeGen/X86/machine-combiner.ll
+++ b/llvm/test/CodeGen/X86/machine-combiner.ll
@@ -144,3 +144,24 @@ define float @reassociate_adds6(float %x0, float %x1, float %x2, float %x3) {
   ret float %t2
 }
 
+; Verify that SSE and AVX scalar single precison multiplies are reassociated.
+
+define float @reassociate_muls1(float %x0, float %x1, float %x2, float %x3) {
+; SSE-LABEL: reassociate_muls1:
+; SSE:       # BB#0:
+; SSE-NEXT:    divss %xmm1, %xmm0
+; SSE-NEXT:    mulss %xmm3, %xmm2
+; SSE-NEXT:    mulss %xmm2, %xmm0
+; SSE-NEXT:    retq
+;
+; AVX-LABEL: reassociate_muls1:
+; AVX:       # BB#0:
+; AVX-NEXT:    vdivss %xmm1, %xmm0, %xmm0
+; AVX-NEXT:    vmulss %xmm3, %xmm2, %xmm1
+; AVX-NEXT:    vmulss %xmm1, %xmm0, %xmm0
+; AVX-NEXT:    retq
+  %t0 = fdiv float %x0, %x1
+  %t1 = fmul float %x2, %t0
+  %t2 = fmul float %x3, %t1
+  ret float %t2
+}
diff --git a/llvm/test/CodeGen/X86/sqrt-fastmath.ll b/llvm/test/CodeGen/X86/sqrt-fastmath.ll
index 0f8d9f4d713..9b851db8121 100644
--- a/llvm/test/CodeGen/X86/sqrt-fastmath.ll
+++ b/llvm/test/CodeGen/X86/sqrt-fastmath.ll
@@ -34,11 +34,11 @@ define float @ff(float %f) #0 {
 ; ESTIMATE:       # BB#0:
 ; ESTIMATE-NEXT:    vrsqrtss %xmm0, %xmm0, %xmm1
 ; ESTIMATE-NEXT:    vmulss {{.*}}(%rip), %xmm1, %xmm2
-; ESTIMATE-NEXT:    vmulss %xmm1, %xmm1, %xmm1
-; ESTIMATE-NEXT:    vmulss %xmm0, %xmm1, %xmm1
+; ESTIMATE-NEXT:    vmulss %xmm0, %xmm1, %xmm3
+; ESTIMATE-NEXT:    vmulss %xmm3, %xmm1, %xmm1
 ; ESTIMATE-NEXT:    vaddss {{.*}}(%rip), %xmm1, %xmm1
+; ESTIMATE-NEXT:    vmulss %xmm0, %xmm2, %xmm2
 ; ESTIMATE-NEXT:    vmulss %xmm2, %xmm1, %xmm1
-; ESTIMATE-NEXT:    vmulss %xmm1, %xmm0, %xmm1
 ; ESTIMATE-NEXT:    vxorps %xmm2, %xmm2, %xmm2
 ; ESTIMATE-NEXT:    vcmpeqss %xmm2, %xmm0, %xmm0
 ; ESTIMATE-NEXT:    vandnps %xmm1, %xmm0, %xmm0
@@ -78,7 +78,7 @@ define float @reciprocal_square_root(float %x) #0 {
 ; ESTIMATE:       # BB#0:
 ; ESTIMATE-NEXT:    vrsqrtss %xmm0, %xmm0, %xmm1
 ; ESTIMATE-NEXT:    vmulss {{.*}}(%rip), %xmm1, %xmm2
-; ESTIMATE-NEXT:    vmulss %xmm1, %xmm1, %xmm1
+; ESTIMATE-NEXT:    vmulss %xmm0, %xmm1, %xmm0
 ; ESTIMATE-NEXT:    vmulss %xmm0, %xmm1, %xmm0
 ; ESTIMATE-NEXT:    vaddss {{.*}}(%rip), %xmm0, %xmm0
 ; ESTIMATE-NEXT:    vmulss %xmm2, %xmm0, %xmm0