[llvm] r318629 - [LibCallSimplifier] allow splat vectors for pow(x, 0.5) -> sqrt() transforms

[Sanjay Patel via llvm-commits]

Author: spatel
Date: Sun Nov 19 08:42:27 2017
New Revision: 318629

URL: http://llvm.org/viewvc/llvm-project?rev=318629&view=rev
Log:
[LibCallSimplifier] allow splat vectors for pow(x, 0.5) -> sqrt() transforms

Modified:
    llvm/trunk/lib/Transforms/Utils/SimplifyLibCalls.cpp
    llvm/trunk/test/Transforms/InstCombine/pow-sqrt.ll

Modified: llvm/trunk/lib/Transforms/Utils/SimplifyLibCalls.cpp
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Transforms/Utils/SimplifyLibCalls.cpp?rev=318629&r1=318628&r2=318629&view=diff
==============================================================================
--- llvm/trunk/lib/Transforms/Utils/SimplifyLibCalls.cpp (original)
+++ llvm/trunk/lib/Transforms/Utils/SimplifyLibCalls.cpp Sun Nov 19 08:42:27 2017
@@ -1081,11 +1081,10 @@ Value *LibCallSimplifier::replacePowWith
   if (!Pow->isFast())
     return nullptr;
 
-  // TODO: This should use m_APFloat to allow vector splats.
-  ConstantFP *Op2C = dyn_cast<ConstantFP>(Pow->getArgOperand(1));
-  if (!Op2C)
+  const APFloat *Arg1C;
+  if (!match(Pow->getArgOperand(1), m_APFloat(Arg1C)))
     return nullptr;
-  if (!Op2C->isExactlyValue(0.5) && !Op2C->isExactlyValue(-0.5))
+  if (!Arg1C->isExactlyValue(0.5) && !Arg1C->isExactlyValue(-0.5))
     return nullptr;
 
   // Fast-math flags from the pow() are propagated to all replacement ops.
@@ -1114,7 +1113,7 @@ Value *LibCallSimplifier::replacePowWith
   }
 
   // If this is pow(x, -0.5), get the reciprocal.
-  if (Op2C->isExactlyValue(-0.5))
+  if (Arg1C->isExactlyValue(-0.5))
     Sqrt = B.CreateFDiv(ConstantFP::get(Ty, 1.0), Sqrt);
 
   return Sqrt;
@@ -1170,6 +1169,9 @@ Value *LibCallSimplifier::optimizePow(Ca
     }
   }
 
+  if (Value *Sqrt = replacePowWithSqrt(CI, B))
+    return Sqrt;
+
   ConstantFP *Op2C = dyn_cast<ConstantFP>(Op2);
   if (!Op2C)
     return Ret;
@@ -1177,9 +1179,6 @@ Value *LibCallSimplifier::optimizePow(Ca
   if (Op2C->getValueAPF().isZero()) // pow(x, 0.0) -> 1.0
     return ConstantFP::get(CI->getType(), 1.0);
 
-  if (Value *Sqrt = replacePowWithSqrt(CI, B))
-    return Sqrt;
-
   // FIXME: Correct the transforms and pull this into replacePowWithSqrt().
   if (Op2C->isExactlyValue(0.5) &&
       hasUnaryFloatFn(TLI, Op2->getType(), LibFunc_sqrt, LibFunc_sqrtf,

Modified: llvm/trunk/test/Transforms/InstCombine/pow-sqrt.ll
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/test/Transforms/InstCombine/pow-sqrt.ll?rev=318629&r1=318628&r2=318629&view=diff
==============================================================================
--- llvm/trunk/test/Transforms/InstCombine/pow-sqrt.ll (original)
+++ llvm/trunk/test/Transforms/InstCombine/pow-sqrt.ll Sun Nov 19 08:42:27 2017
@@ -32,8 +32,9 @@ define double @pow_libcall_half_approx(d
 
 define <2 x double> @pow_intrinsic_neghalf_fast(<2 x double> %x) {
 ; CHECK-LABEL: @pow_intrinsic_neghalf_fast(
-; CHECK-NEXT:    [[POW:%.*]] = call fast <2 x double> @llvm.pow.v2f64(<2 x double> %x, <2 x double> <double -5.000000e-01, double -5.000000e-01>)
-; CHECK-NEXT:    ret <2 x double> [[POW]]
+; CHECK-NEXT:    [[TMP1:%.*]] = call fast <2 x double> @llvm.sqrt.v2f64(<2 x double> %x)
+; CHECK-NEXT:    [[TMP2:%.*]] = fdiv fast <2 x double> <double 1.000000e+00, double 1.000000e+00>, [[TMP1]]
+; CHECK-NEXT:    ret <2 x double> [[TMP2]]
 ;
   %pow = call fast <2 x double> @llvm.pow.v2f64(<2 x double> %x, <2 x double> <double -5.0e-01, double -5.0e-01>)
   ret <2 x double> %pow