From 7530a9f7d1e62b28e04d771158613c2954cc8d8c Mon Sep 17 00:00:00 2001
From: Hal Finkel <hfinkel@anl.gov>
Date: Thu, 4 Apr 2013 22:44:12 +0000
Subject: [PATCH] PPC: Improve code generation for mixed-precision reciprocal
 sqrt

The DAGCombine logic that recognized a/sqrt(b) and transformed it into
a multiplication by the reciprocal sqrt did not handle cases where the
sqrt and the division were separated by an fpext or fptrunc.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@178801 91177308-0d34-0410-b5e6-96231b3b80d8
---
 lib/Target/PowerPC/PPCISelLowering.cpp | 27 +++++++++++++++
 test/CodeGen/PowerPC/recipest.ll       | 48 ++++++++++++++++++++++++++
 2 files changed, 75 insertions(+)

diff --git a/lib/Target/PowerPC/PPCISelLowering.cpp b/lib/Target/PowerPC/PPCISelLowering.cpp
index e33d2cc7165..8195f5eb050 100644
--- a/lib/Target/PowerPC/PPCISelLowering.cpp
+++ b/lib/Target/PowerPC/PPCISelLowering.cpp
@@ -6849,6 +6849,33 @@ SDValue PPCTargetLowering::PerformDAGCombine(SDNode *N,
         return DAG.getNode(ISD::FMUL, dl, N->getValueType(0),
                            N->getOperand(0), RV);
       }
+    } else if (N->getOperand(1).getOpcode() == ISD::FP_EXTEND &&
+               N->getOperand(1).getOperand(0).getOpcode() == ISD::FSQRT) {
+      SDValue RV =
+        DAGCombineFastRecipFSQRT(N->getOperand(1).getOperand(0).getOperand(0),
+                                 DCI);
+      if (RV.getNode() != 0) {
+        DCI.AddToWorklist(RV.getNode());
+        RV = DAG.getNode(ISD::FP_EXTEND, N->getOperand(1).getDebugLoc(),
+                         N->getValueType(0), RV);
+        DCI.AddToWorklist(RV.getNode());
+        return DAG.getNode(ISD::FMUL, dl, N->getValueType(0),
+                           N->getOperand(0), RV);
+      }
+    } else if (N->getOperand(1).getOpcode() == ISD::FP_ROUND &&
+               N->getOperand(1).getOperand(0).getOpcode() == ISD::FSQRT) {
+      SDValue RV =
+        DAGCombineFastRecipFSQRT(N->getOperand(1).getOperand(0).getOperand(0),
+                                 DCI);
+      if (RV.getNode() != 0) {
+        DCI.AddToWorklist(RV.getNode());
+        RV = DAG.getNode(ISD::FP_ROUND, N->getOperand(1).getDebugLoc(),
+                         N->getValueType(0), RV,
+                         N->getOperand(1).getOperand(1));
+        DCI.AddToWorklist(RV.getNode());
+        return DAG.getNode(ISD::FMUL, dl, N->getValueType(0),
+                           N->getOperand(0), RV);
+      }
     }
 
     SDValue RV = DAGCombineFastRecip(N->getOperand(1), DCI);
diff --git a/test/CodeGen/PowerPC/recipest.ll b/test/CodeGen/PowerPC/recipest.ll
index cad98c5a352..89705faa46e 100644
--- a/test/CodeGen/PowerPC/recipest.ll
+++ b/test/CodeGen/PowerPC/recipest.ll
@@ -31,6 +31,54 @@ entry:
 ; CHECK-SAFE: blr
 }
 
+define double @foof(double %a, float %b) nounwind {
+entry:
+  %x = call float @llvm.sqrt.f32(float %b)
+  %y = fpext float %x to double
+  %r = fdiv double %a, %y
+  ret double %r
+
+; CHECK: @foof
+; CHECK: frsqrtes
+; CHECK: fnmsubs
+; CHECK: fmuls
+; CHECK: fmadds
+; CHECK: fmuls
+; CHECK: fmul
+; CHECK: blr
+
+; CHECK-SAFE: @foof
+; CHECK-SAFE: fsqrts
+; CHECK-SAFE: fdiv
+; CHECK-SAFE: blr
+}
+
+define float @food(float %a, double %b) nounwind {
+entry:
+  %x = call double @llvm.sqrt.f64(double %b)
+  %y = fptrunc double %x to float
+  %r = fdiv float %a, %y
+  ret float %r
+
+; CHECK: @foo
+; CHECK: frsqrte
+; CHECK: fnmsub
+; CHECK: fmul
+; CHECK: fmadd
+; CHECK: fmul
+; CHECK: fmul
+; CHECK: fmadd
+; CHECK: fmul
+; CHECK: frsp
+; CHECK: fmuls
+; CHECK: blr
+
+; CHECK-SAFE: @foo
+; CHECK-SAFE: fsqrt
+; CHECK-SAFE: fdivs
+; CHECK-SAFE: blr
+}
+
 define float @goo(float %a, float %b) nounwind {
 entry:
   %x = call float @llvm.sqrt.f32(float %b)