[InstSimplify] sqrt(X) * sqrt(X) --> X

This was misplaced in InstCombine. We can loosen the FMF as a follow-up step.

llvm-svn: 325965

llvm/lib/Analysis/InstructionSimplify.cpp

@@ -4251,6 +4251,12 @@ static Value *SimplifyFMulInst(Value *Op0, Value *Op1, FastMathFlags FMF,
   if (FMF.noNaNs() && FMF.noSignedZeros() && match(Op1, m_AnyZero()))
     return Op1;
 
+  // sqrt(X) * sqrt(X) --> X
+  Value *X;
+  if (FMF.isFast() && Op0 == Op1 &&
+      match(Op0, m_Intrinsic<Intrinsic::sqrt>(m_Value(X))))
+    return X;
+
   return nullptr;
 }
 

llvm/lib/Transforms/InstCombine/InstCombineMulDivRem.cpp

@@ -620,10 +620,6 @@ Instruction *InstCombiner::visitFMul(BinaryOperator &I) {
 
   if (Op0 == Op1) {
     if (IntrinsicInst *II = dyn_cast<IntrinsicInst>(Op0)) {
-      // sqrt(X) * sqrt(X) -> X
-      if (AllowReassociate && II->getIntrinsicID() == Intrinsic::sqrt)
-        return replaceInstUsesWith(I, II->getOperand(0));
-
       // fabs(X) * fabs(X) -> X * X
       if (II->getIntrinsicID() == Intrinsic::fabs) {
         Instruction *FMulVal = BinaryOperator::CreateFMul(II->getOperand(0),

llvm/test/Transforms/InstCombine/fmul.ll

@@ -172,19 +172,8 @@ define float @test11(float %x, float %y) {
   ret float %c
 }
 
-; PR21126: http://llvm.org/bugs/show_bug.cgi?id=21126
-; With unsafe/fast math, sqrt(X) * sqrt(X) is just X.
 declare double @llvm.sqrt.f64(double)
 
-define double @sqrt_squared1(double %f) {
-; CHECK-LABEL: @sqrt_squared1(
-; CHECK-NEXT:    ret double [[F:%.*]]
-;
-  %sqrt = call double @llvm.sqrt.f64(double %f)
-  %mul = fmul fast double %sqrt, %sqrt
-  ret double %mul
-}
-
 ; With unsafe/fast math, sqrt(X) * sqrt(X) is just X,
 ; but make sure another use of the sqrt is intact.
 ; Note that the remaining fmul is altered but is not 'fast'

llvm/test/Transforms/InstSimplify/fast-math.ll

@@ -203,3 +203,18 @@ define float @fdiv_neg_swapped2(float %f) {
   %div = fdiv nnan float %f, %neg
   ret float %div
 }
+
+; PR21126: http://llvm.org/bugs/show_bug.cgi?id=21126
+; With unsafe/fast math, sqrt(X) * sqrt(X) is just X.
+
+declare double @llvm.sqrt.f64(double)
+
+define double @sqrt_squared(double %f) {
+; CHECK-LABEL: @sqrt_squared(
+; CHECK-NEXT:    ret double [[F:%.*]]
+;
+  %sqrt = call double @llvm.sqrt.f64(double %f)
+  %mul = fmul fast double %sqrt, %sqrt
+  ret double %mul
+}
+