Index: llvm/lib/Target/AMDGPU/AMDGPUCodeGenPrepare.cpp
===================================================================
--- llvm/lib/Target/AMDGPU/AMDGPUCodeGenPrepare.cpp
+++ llvm/lib/Target/AMDGPU/AMDGPUCodeGenPrepare.cpp
@@ -525,12 +525,64 @@
   return true;
 }
 
-static bool shouldKeepFDivF32(Value *Num, bool UnsafeDiv, bool HasDenormals) {
+// Perform RCP optimizations:
+//
+// 1/x -> rcp(x) when fast unsafe rcp is legal or fpmath >= 2.5ULP with
+//                                                denormals flushed.
+//
+// a/b -> a*rcp(b) when fast unsafe rcp is legal.
+static Value *performRCPOpt(Value *Num, Value *Den, bool FastUnsafeRcpLegal,
+                            IRBuilder<> Builder, MDNode *FPMath, Module *Mod,
+                            bool HasDenormals, bool NeedHighAccuracy) {
+
+  Type *Ty = Den->getType();
+  if (!FastUnsafeRcpLegal && Ty->isFloatTy() &&
+                             (HasDenormals || NeedHighAccuracy))
+    return nullptr;
+
+  Function *Decl = Intrinsic::getDeclaration(Mod, Intrinsic::amdgcn_rcp, Ty);
+  if (const ConstantFP *CLHS = dyn_cast<ConstantFP>(Num)) {
+    if (FastUnsafeRcpLegal || Ty->isFloatTy() || Ty->isHalfTy()) {
+      if (CLHS->isExactlyValue(1.0)) {
+        // v_rcp_f32 and v_rsq_f32 do not support denormals, and according to
+        // the CI documentation has a worst case error of 1 ulp.
+        // OpenCL requires <= 2.5 ulp for 1.0 / x, so it should always be OK to
+        // use it as long as we aren't trying to use denormals.
+        //
+        // v_rcp_f16 and v_rsq_f16 DO support denormals.
+
+        // NOTE: v_sqrt and v_rcp will be combined to v_rsq later. So we don't
+        //       insert rsq intrinsic here.
+
+        // 1.0 / x -> rcp(x)
+        return Builder.CreateCall(Decl, { Den });
+       }
+
+       // Same as for 1.0, but expand the sign out of the constant.
+       if (CLHS->isExactlyValue(-1.0)) {
+         // -1.0 / x -> rcp (fneg x)
+         Value *FNeg = Builder.CreateFNeg(Den);
+         return Builder.CreateCall(Decl, { FNeg });
+       }
+    }
+  }
+
+  if (FastUnsafeRcpLegal) {
+    // Turn into multiply by the reciprocal.
+    // x / y -> x * (1.0 / y)
+    Value *Recip = Builder.CreateCall(Decl, { Den });
+    return Builder.CreateFMul(Num, Recip, "", FPMath);
+  }
+  return nullptr;
+}
+
+static bool shouldKeepFDivF32(Value *Num, bool FastUnsafeRcpLegal,
+                              bool HasDenormals) {
   const ConstantFP *CNum = dyn_cast<ConstantFP>(Num);
   if (!CNum)
     return HasDenormals;
 
-  if (UnsafeDiv)
+  if (FastUnsafeRcpLegal)
     return true;
 
   bool IsOne = CNum->isExactlyValue(+1.0) || CNum->isExactlyValue(-1.0);
@@ -539,44 +591,57 @@
   return HasDenormals ^ IsOne;
 }
 
-// Insert an intrinsic for fast fdiv for safe math situations where we can
-// reduce precision. Leave fdiv for situations where the generic node is
-// expected to be optimized.
+
+// Optimizations is performed based on fpmath, fast math flags as wells as
+// denormals to lower fdiv using either rcp or fdiv.fast.
+//
+// FastUnsafeRcpLegal: We determine whether it is legal to use rcp based on
+//                     unsafe-fp-math, fast math flags, denormals and fpmath
+//                     accuracy request.
+//
+// RCP Optimizations:
+//   1/x -> rcp(x) when fast unsafe rcp is legal or fpmath >= 2.5ULP with
+//                                                  denormals flushed.
+//   a/b -> a*rcp(b) when fast unsafe rcp is legal.
+//
+// Use fdiv.fast:
+//   a/b -> fdiv.fast(a, b) when RCP optimization is not performed and
+//                          fpmath >= 2.5ULP with denormals flushed.
+//
+//   1/x -> fdiv.fast(1,x)  when RCP optimization is not performed and
+//                          fpmath >= 2.5ULP with denormals.
 bool AMDGPUCodeGenPrepare::visitFDiv(BinaryOperator &FDiv) {
-  Type *Ty = FDiv.getType();
 
-  if (!Ty->getScalarType()->isFloatTy())
-    return false;
+  Type *Ty = FDiv.getType()->getScalarType();
 
-  MDNode *FPMath = FDiv.getMetadata(LLVMContext::MD_fpmath);
-  if (!FPMath)
+  // No intrinsic for fdiv16 if target does not support f16.
+  if (Ty->isHalfTy() && !ST->has16BitInsts())
     return false;
 
   const FPMathOperator *FPOp = cast<const FPMathOperator>(&FDiv);
-  float ULP = FPOp->getFPAccuracy();
-  if (ULP < 2.5f)
-    return false;
+  MDNode *FPMath = FDiv.getMetadata(LLVMContext::MD_fpmath);
+  const bool NeedHighAccuracy = !FPMath || FPOp->getFPAccuracy() < 2.5f;
 
   FastMathFlags FMF = FPOp->getFastMathFlags();
-  bool UnsafeDiv = HasUnsafeFPMath || FMF.isFast() ||
-                                      FMF.allowReciprocal();
+  // Determine whether it is ok to use rcp based on unsafe-fp-math,
+  // fast math flags, denormals and accuracy request.
+  const bool FastUnsafeRcpLegal = HasUnsafeFPMath || FMF.isFast() ||
+          (FMF.allowReciprocal() && ((!HasFP32Denormals && !NeedHighAccuracy)
+                                     || FMF.approxFunc()));
 
-  // With UnsafeDiv node will be optimized to just rcp and mul.
-  if (UnsafeDiv)
-    return false;
+  // Use fdiv.fast for only f32, fpmath >= 2.5ULP and rcp is not used.
+  const bool UseFDivFast = Ty->isFloatTy() && !NeedHighAccuracy &&
+                           !FastUnsafeRcpLegal;
 
-  IRBuilder<> Builder(FDiv.getParent(), std::next(FDiv.getIterator()), FPMath);
+  IRBuilder<> Builder(FDiv.getParent(), std::next(FDiv.getIterator()));
   Builder.setFastMathFlags(FMF);
   Builder.SetCurrentDebugLocation(FDiv.getDebugLoc());
 
-  Function *Decl = Intrinsic::getDeclaration(Mod, Intrinsic::amdgcn_fdiv_fast);
-
   Value *Num = FDiv.getOperand(0);
   Value *Den = FDiv.getOperand(1);
 
   Value *NewFDiv = nullptr;
-
-  if (VectorType *VT = dyn_cast<VectorType>(Ty)) {
+  if (VectorType *VT = dyn_cast<VectorType>(FDiv.getType())) {
     NewFDiv = UndefValue::get(VT);
 
     // FIXME: Doesn't do the right thing for cases where the vector is partially
@@ -584,19 +649,32 @@
     for (unsigned I = 0, E = VT->getNumElements(); I != E; ++I) {
       Value *NumEltI = Builder.CreateExtractElement(Num, I);
       Value *DenEltI = Builder.CreateExtractElement(Den, I);
-      Value *NewElt;
-
-      if (shouldKeepFDivF32(NumEltI, UnsafeDiv, HasFP32Denormals)) {
-        NewElt = Builder.CreateFDiv(NumEltI, DenEltI);
-      } else {
-        NewElt = Builder.CreateCall(Decl, { NumEltI, DenEltI });
+      Value *NewElt = nullptr;
+      if (UseFDivFast && !shouldKeepFDivF32(NumEltI, FastUnsafeRcpLegal,
+                                           HasFP32Denormals)) {
+        Function *Decl =
+                 Intrinsic::getDeclaration(Mod, Intrinsic::amdgcn_fdiv_fast);
+        NewElt = Builder.CreateCall(Decl, { NumEltI, DenEltI }, "", FPMath);
       }
+      if (!NewElt) // Try rcp.
+        NewElt = performRCPOpt(NumEltI, DenEltI, FastUnsafeRcpLegal, Builder,
+                               FPMath, Mod, HasFP32Denormals, NeedHighAccuracy);
+      if (!NewElt)
+        NewElt = Builder.CreateFDiv(NumEltI, DenEltI, "", FPMath);
 
       NewFDiv = Builder.CreateInsertElement(NewFDiv, NewElt, I);
     }
-  } else {
-    if (!shouldKeepFDivF32(Num, UnsafeDiv, HasFP32Denormals))
-      NewFDiv = Builder.CreateCall(Decl, { Num, Den });
+  } else { // Scalar.
+    if (UseFDivFast && !shouldKeepFDivF32(Num, FastUnsafeRcpLegal,
+                                          HasFP32Denormals)) {
+      Function *Decl =
+               Intrinsic::getDeclaration(Mod, Intrinsic::amdgcn_fdiv_fast);
+      NewFDiv = Builder.CreateCall(Decl, { Num, Den }, "", FPMath);
+    }
+    if (!NewFDiv) { // Try rcp.
+      NewFDiv = performRCPOpt(Num, Den, FastUnsafeRcpLegal, Builder, FPMath,
+                              Mod, HasFP32Denormals, NeedHighAccuracy);
+    }
   }
 
   if (NewFDiv) {
Index: llvm/lib/Target/AMDGPU/SIISelLowering.cpp
===================================================================
--- llvm/lib/Target/AMDGPU/SIISelLowering.cpp
+++ llvm/lib/Target/AMDGPU/SIISelLowering.cpp
@@ -7537,49 +7537,54 @@
   SDValue RHS = Op.getOperand(1);
   EVT VT = Op.getValueType();
   const SDNodeFlags Flags = Op->getFlags();
-  bool Unsafe = DAG.getTarget().Options.UnsafeFPMath || Flags.hasAllowReciprocal();
 
-  if (!Unsafe && VT == MVT::f32 && hasFP32Denormals(DAG.getMachineFunction()))
+  bool FastUnsafeRcpLegal = DAG.getTarget().Options.UnsafeFPMath ||
+         (Flags.hasAllowReciprocal() &&
+          ((VT == MVT::f32 && hasFP32Denormals(DAG.getMachineFunction())) ||
+            VT == MVT::f16 ||
+            Flags.hasApproximateFuncs()));
+
+  // Do rcp optimization only when fast unsafe rcp is legal here.
+  // NOTE: We already performed RCP optimization to insert intrinsics in
+  // AMDGPUCodeGenPrepare. Ideally there should have no opportunity here to
+  // rcp optimization.
+  //   However, there are cases like FREM, which is expended into a sequence
+  // of instructions including FDIV, which may expose new opportunities.
+  if (!FastUnsafeRcpLegal)
     return SDValue();
 
   if (const ConstantFPSDNode *CLHS = dyn_cast<ConstantFPSDNode>(LHS)) {
-    if (Unsafe || VT == MVT::f32 || VT == MVT::f16) {
-      if (CLHS->isExactlyValue(1.0)) {
-        // v_rcp_f32 and v_rsq_f32 do not support denormals, and according to
-        // the CI documentation has a worst case error of 1 ulp.
-        // OpenCL requires <= 2.5 ulp for 1.0 / x, so it should always be OK to
-        // use it as long as we aren't trying to use denormals.
-        //
-        // v_rcp_f16 and v_rsq_f16 DO support denormals.
-
-        // 1.0 / sqrt(x) -> rsq(x)
-
-        // XXX - Is UnsafeFPMath sufficient to do this for f64? The maximum ULP
-        // error seems really high at 2^29 ULP.
-        if (RHS.getOpcode() == ISD::FSQRT)
-          return DAG.getNode(AMDGPUISD::RSQ, SL, VT, RHS.getOperand(0));
-
-        // 1.0 / x -> rcp(x)
-        return DAG.getNode(AMDGPUISD::RCP, SL, VT, RHS);
-      }
+    if (CLHS->isExactlyValue(1.0)) {
+      // v_rcp_f32 and v_rsq_f32 do not support denormals, and according to
+      // the CI documentation has a worst case error of 1 ulp.
+      // OpenCL requires <= 2.5 ulp for 1.0 / x, so it should always be OK to
+      // use it as long as we aren't trying to use denormals.
+      //
+      // v_rcp_f16 and v_rsq_f16 DO support denormals.
 
-      // Same as for 1.0, but expand the sign out of the constant.
-      if (CLHS->isExactlyValue(-1.0)) {
-        // -1.0 / x -> rcp (fneg x)
-        SDValue FNegRHS = DAG.getNode(ISD::FNEG, SL, VT, RHS);
-        return DAG.getNode(AMDGPUISD::RCP, SL, VT, FNegRHS);
-      }
+      // 1.0 / sqrt(x) -> rsq(x)
+
+      // XXX - Is UnsafeFPMath sufficient to do this for f64? The maximum ULP
+      // error seems really high at 2^29 ULP.
+      if (RHS.getOpcode() == ISD::FSQRT)
+        return DAG.getNode(AMDGPUISD::RSQ, SL, VT, RHS.getOperand(0));
+
+      // 1.0 / x -> rcp(x)
+      return DAG.getNode(AMDGPUISD::RCP, SL, VT, RHS);
     }
-  }
 
-  if (Unsafe) {
-    // Turn into multiply by the reciprocal.
-    // x / y -> x * (1.0 / y)
-    SDValue Recip = DAG.getNode(AMDGPUISD::RCP, SL, VT, RHS);
-    return DAG.getNode(ISD::FMUL, SL, VT, LHS, Recip, Flags);
+    // Same as for 1.0, but expand the sign out of the constant.
+    if (CLHS->isExactlyValue(-1.0)) {
+      // -1.0 / x -> rcp (fneg x)
+      SDValue FNegRHS = DAG.getNode(ISD::FNEG, SL, VT, RHS);
+      return DAG.getNode(AMDGPUISD::RCP, SL, VT, FNegRHS);
+    }
   }
 
-  return SDValue();
+  // Turn into multiply by the reciprocal.
+  // x / y -> x * (1.0 / y)
+  SDValue Recip = DAG.getNode(AMDGPUISD::RCP, SL, VT, RHS);
+  return DAG.getNode(ISD::FMUL, SL, VT, LHS, Recip, Flags);
 }
 
 static SDValue getFPBinOp(SelectionDAG &DAG, unsigned Opcode, const SDLoc &SL,
@@ -8725,6 +8730,11 @@
                            N->getFlags());
   }
 
+  if ((VT == MVT::f32 || VT == MVT::f16) && N0.getOpcode() == ISD::FSQRT) {
+    return DCI.DAG.getNode(AMDGPUISD::RSQ, SDLoc(N), VT,
+                           N0.getOperand(0), N->getFlags());
+  }
+
   return AMDGPUTargetLowering::performRcpCombine(N, DCI);
 }
 
Index: llvm/test/CodeGen/AMDGPU/amdgpu-codegenprepare-fdiv.ll
===================================================================
--- llvm/test/CodeGen/AMDGPU/amdgpu-codegenprepare-fdiv.ll
+++ llvm/test/CodeGen/AMDGPU/amdgpu-codegenprepare-fdiv.ll
@@ -16,8 +16,10 @@
 ; CHECK: %md.1ulp = fdiv float %a, %b, !fpmath !2
 ; CHECK: %md.25ulp = call float @llvm.amdgcn.fdiv.fast(float %a, float %b), !fpmath !0
 ; CHECK: %md.3ulp = call float @llvm.amdgcn.fdiv.fast(float %a, float %b), !fpmath !3
-; CHECK: %fast.md.25ulp = fdiv fast float %a, %b, !fpmath !0
-; CHECK: arcp.md.25ulp = fdiv arcp float %a, %b, !fpmath !0
+; CHECK: %[[FAST_RCP:[0-9]+]] = call fast float @llvm.amdgcn.rcp.f32(float %b)
+; CHECK: %fast.md.25ulp = fmul fast float %a, %[[FAST_RCP]], !fpmath !0
+; CHECK: %[[ARCP_RCP:[0-9]+]] = call arcp float @llvm.amdgcn.rcp.f32(float %b)
+; CHECK: arcp.md.25ulp = fmul arcp float %a, %[[ARCP_RCP]], !fpmath !0
 define amdgpu_kernel void @fdiv_fpmath(float addrspace(1)* %out, float %a, float %b) #1 {
   %no.md = fdiv float %a, %b
   store volatile float %no.md, float addrspace(1)* %out
@@ -45,12 +47,12 @@
 
 ; CHECK-LABEL: @rcp_fdiv_fpmath(
 ; CHECK: %no.md = fdiv float 1.000000e+00, %x{{$}}
-; CHECK: %md.25ulp = fdiv float 1.000000e+00, %x, !fpmath !0
+; CHECK: %md.25ulp = call float @llvm.amdgcn.rcp.f32(float %x)
 ; CHECK: %md.half.ulp = fdiv float 1.000000e+00, %x, !fpmath !1
-; CHECK: %arcp.no.md = fdiv arcp float 1.000000e+00, %x{{$}}
-; CHECK: %arcp.25ulp = fdiv arcp float 1.000000e+00, %x, !fpmath !0
-; CHECK: %fast.no.md = fdiv fast float 1.000000e+00, %x{{$}}
-; CHECK: %fast.25ulp = fdiv fast float 1.000000e+00, %x, !fpmath !0
+; CHECK: %arcp.no.md = fdiv arcp float 1.000000e+00, %x
+; CHECK: %arcp.25ulp = call arcp float @llvm.amdgcn.rcp.f32(float %x)
+; CHECK: %fast.no.md = call fast float @llvm.amdgcn.rcp.f32(float %x)
+; CHECK: %fast.25ulp = call fast float @llvm.amdgcn.rcp.f32(float %x)
 define amdgpu_kernel void @rcp_fdiv_fpmath(float addrspace(1)* %out, float %x) #1 {
   %no.md = fdiv float 1.0, %x
   store volatile float %no.md, float addrspace(1)* %out
@@ -76,10 +78,58 @@
   ret void
 }
 
+; CHECK-LABEL: @rcp_fdiv_arcp_denormal(
+; CHECK: %arcp.low.accuracy = call arcp float @llvm.amdgcn.fdiv.fast(float 1.000000e+00, float %x), !fpmath !0
+; CHECK: %arcp.high.accuracy = fdiv arcp float 1.000000e+00, %x, !fpmath !2
+; CHECK: %arcp.low.afn = call arcp afn float @llvm.amdgcn.rcp.f32(float %x)
+; CHECK: %arcp.high.afn = call arcp afn float @llvm.amdgcn.rcp.f32(float %x)
+define amdgpu_kernel void @rcp_fdiv_arcp_denormal(float addrspace(1)* %out, float %x) #2 {
+
+  %arcp.low.accuracy = fdiv arcp float 1.0, %x, !fpmath !0
+  store volatile float %arcp.low.accuracy, float addrspace(1)* %out
+
+  %arcp.high.accuracy = fdiv arcp float 1.0, %x, !fpmath !2
+  store volatile float %arcp.high.accuracy, float addrspace(1)* %out
+
+  %arcp.low.afn = fdiv arcp afn float 1.0, %x, !fpmath !0
+  store volatile float %arcp.low.afn, float addrspace(1)* %out
+
+  %arcp.high.afn = fdiv arcp afn float 1.0, %x, !fpmath !2
+  store volatile float %arcp.high.afn, float addrspace(1)* %out
+
+  ret void
+}
+
 ; CHECK-LABEL: @fdiv_fpmath_vector(
-; CHECK: %no.md = fdiv <2 x float> %a, %b{{$}}
-; CHECK: %md.half.ulp = fdiv <2 x float> %a, %b, !fpmath !1
-; CHECK: %md.1ulp = fdiv <2 x float> %a, %b, !fpmath !2
+; CHECK: %[[NO_A0:[0-9]+]] = extractelement <2 x float> %a, i64 0
+; CHECK: %[[NO_B0:[0-9]+]] = extractelement <2 x float> %b, i64 0
+; CHECK: %[[NO_FDIV0:[0-9]+]] = fdiv float %[[NO_A0]], %[[NO_B0]]
+; CHECK: %[[NO_INS0:[0-9]+]] = insertelement <2 x float> undef, float %[[NO_FDIV0]], i64 0
+; CHECK: %[[NO_A1:[0-9]+]] = extractelement <2 x float> %a, i64 1
+; CHECK: %[[NO_B1:[0-9]+]] = extractelement <2 x float> %b, i64 1
+; CHECK: %[[NO_FDIV1:[0-9]+]] = fdiv float %[[NO_A1]], %[[NO_B1]]
+; CHECK: %no.md = insertelement <2 x float> %[[NO_INS0]], float %[[NO_FDIV1]], i64 1
+; CHECK: store volatile <2 x float> %no.md, <2 x float> addrspace(1)* %out
+
+; CHECK: %[[HALF_A0:[0-9]+]] = extractelement <2 x float> %a, i64 0
+; CHECK: %[[HALF_B0:[0-9]+]] = extractelement <2 x float> %b, i64 0
+; CHECK: %[[HALF_FDIV0:[0-9]+]] = fdiv float %[[HALF_A0]], %[[HALF_B0]], !fpmath !1
+; CHECK: %[[HALF_INS0:[0-9]+]] = insertelement <2 x float> undef, float %[[HALF_FDIV0]], i64 0
+; CHECK: %[[HALF_A1:[0-9]+]] = extractelement <2 x float> %a, i64 1
+; CHECK: %[[HALF_B1:[0-9]+]] = extractelement <2 x float> %b, i64 1
+; CHECK: %[[HALF_FDIV1:[0-9]+]] = fdiv float %[[HALF_A1]], %[[HALF_B1]], !fpmath !1
+; CHECK: %md.half.ulp = insertelement <2 x float> %[[HALF_INS0]], float %[[HALF_FDIV1]], i64 1
+; CHECK: store volatile <2 x float> %md.half.ulp, <2 x float> addrspace(1)* %out
+
+; CHECK: %[[ONE_A0:[0-9]+]] = extractelement <2 x float> %a, i64 0
+; CHECK: %[[ONE_B0:[0-9]+]] = extractelement <2 x float> %b, i64 0
+; CHECK: %[[ONE_FDIV0:[0-9]+]] = fdiv float %[[ONE_A0]], %[[ONE_B0]], !fpmath !2
+; CHECK: %[[ONE_INS0:[0-9]+]] = insertelement <2 x float> undef, float %[[ONE_FDIV0]], i64 0
+; CHECK: %[[ONE_A1:[0-9]+]] = extractelement <2 x float> %a, i64 1
+; CHECK: %[[ONE_B1:[0-9]+]] = extractelement <2 x float> %b, i64 1
+; CHECK: %[[ONE_FDIV1:[0-9]+]] = fdiv float %[[ONE_A1]], %[[ONE_B1]], !fpmath !2
+; CHECK: %md.1ulp = insertelement <2 x float> %[[ONE_INS0]], float %[[ONE_FDIV1]], i64 1
+; CHECK: store volatile <2 x float> %md.1ulp, <2 x float> addrspace(1)* %out
 
 ; CHECK: %[[A0:[0-9]+]] = extractelement <2 x float> %a, i64 0
 ; CHECK: %[[B0:[0-9]+]] = extractelement <2 x float> %b, i64 0
@@ -106,12 +156,52 @@
 }
 
 ; CHECK-LABEL: @rcp_fdiv_fpmath_vector(
-; CHECK: %no.md = fdiv <2 x float> <float 1.000000e+00, float 1.000000e+00>, %x{{$}}
-; CHECK: %md.half.ulp = fdiv <2 x float> <float 1.000000e+00, float 1.000000e+00>, %x, !fpmath !1
-; CHECK: %arcp.no.md = fdiv arcp <2 x float> <float 1.000000e+00, float 1.000000e+00>, %x{{$}}
-; CHECK: %fast.no.md = fdiv fast <2 x float> <float 1.000000e+00, float 1.000000e+00>, %x{{$}}
-; CHECK: %arcp.25ulp = fdiv arcp <2 x float> <float 1.000000e+00, float 1.000000e+00>, %x, !fpmath !0
-; CHECK: %fast.25ulp = fdiv fast <2 x float> <float 1.000000e+00, float 1.000000e+00>, %x, !fpmath !0
+; CHECK: %[[NO0:[0-9]+]] =  extractelement <2 x float> %x, i64 0
+; CHECK: %[[NO_FDIV0:[0-9]+]] = fdiv float 1.000000e+00, %[[NO0]]
+; CHECK: %[[NO_INS0:[0-9]+]] = insertelement <2 x float> undef, float %[[NO_FDIV0]], i64 0
+; CHECK: %[[NO1:[0-9]+]] = extractelement <2 x float> %x, i64 1
+; CHECK: %[[NO_FDIV1:[0-9]+]] = fdiv float 1.000000e+00, %[[NO1]]
+; CHECK: %no.md = insertelement <2 x float> %[[NO_INS0]], float %[[NO_FDIV1]], i64 1
+; CHECK: store volatile <2 x float> %no.md, <2 x float> addrspace(1)* %out
+
+; CHECK: %[[HALF0:[0-9]+]] =  extractelement <2 x float> %x, i64 0
+; CHECK: %[[HALF_FDIV0:[0-9]+]] = fdiv float 1.000000e+00, %[[HALF0]], !fpmath !1
+; CHECK: %[[HALF_INS0:[0-9]+]] = insertelement <2 x float> undef, float %[[HALF_FDIV0]], i64 0
+; CHECK: %[[HALF1:[0-9]+]] = extractelement <2 x float> %x, i64 1
+; CHECK: %[[HALF_FDIV1:[0-9]+]] =  fdiv float 1.000000e+00, %[[HALF1]], !fpmath !1
+; CHECK: %md.half.ulp = insertelement <2 x float> %[[HALF_INS0]], float %[[HALF_FDIV1]], i64 1
+; CHECK: store volatile <2 x float> %md.half.ulp, <2 x float> addrspace(1)* %out
+
+; CHECK: %[[ARCP_NO0:[0-9]+]] =  extractelement <2 x float> %x, i64 0
+; CHECK: %[[ARCP_NO_FDIV0:[0-9]+]] = fdiv arcp float 1.000000e+00, %[[ARCP_NO0]]
+; CHECK: %[[ARCP_NO_INS0:[0-9]+]] = insertelement <2 x float> undef, float %[[ARCP_NO_FDIV0]], i64 0
+; CHECK: %[[ARCP_NO1:[0-9]+]] = extractelement <2 x float> %x, i64 1
+; CHECK: %[[ARCP_NO_FDIV1:[0-9]+]] =  fdiv arcp float 1.000000e+00, %[[ARCP_NO1]]
+; CHECK: %arcp.no.md = insertelement <2 x float> %[[ARCP_NO_INS0]], float %[[ARCP_NO_FDIV1]], i64 1
+; CHECK: store volatile <2 x float> %arcp.no.md, <2 x float> addrspace(1)* %out
+
+; CHECK: %[[FAST_NO0:[0-9]+]] =  extractelement <2 x float> %x, i64 0
+; CHECK: %[[FAST_NO_RCP0:[0-9]+]] = call fast float @llvm.amdgcn.rcp.f32(float %[[FAST_NO0]])
+; CHECK: %[[FAST_NO_INS0:[0-9]+]] = insertelement <2 x float> undef, float %[[FAST_NO_RCP0]], i64 0
+; CHECK: %[[FAST_NO1:[0-9]+]] = extractelement <2 x float> %x, i64 1
+; CHECK: %[[FAST_NO_RCP1:[0-9]+]] =  call fast float @llvm.amdgcn.rcp.f32(float %[[FAST_NO1]])
+; CHECK: %fast.no.md = insertelement <2 x float> %[[FAST_NO_INS0]], float %[[FAST_NO_RCP1]], i64 1
+; CHECK: store volatile <2 x float> %fast.no.md, <2 x float> addrspace(1)* %out
+
+; CHECK: %[[ARCP_250:[0-9]+]] =  extractelement <2 x float> %x, i64 0
+; CHECK: %[[ARCP_25_RCP0:[0-9]+]] = call arcp float @llvm.amdgcn.rcp.f32(float %[[ARCP_250]])
+; CHECK: %[[ARCP_25_INS0:[0-9]+]] = insertelement <2 x float> undef, float %[[ARCP_25_RCP0]], i64 0
+; CHECK: %[[ARCP_251:[0-9]+]] = extractelement <2 x float> %x, i64 1
+; CHECK: %[[ARCP_25_RCP1:[0-9]+]] =  call arcp float @llvm.amdgcn.rcp.f32(float %[[ARCP_251]])
+; CHECK: %arcp.25ulp = insertelement <2 x float> %[[ARCP_25_INS0]], float %[[ARCP_25_RCP1]], i64 1
+; CHECK: store volatile <2 x float> %arcp.25ulp, <2 x float> addrspace(1)* %out
+
+; CHECK: %[[FAST_250:[0-9]+]] =  extractelement <2 x float> %x, i64 0
+; CHECK: %[[FAST_25_RCP0:[0-9]+]] = call fast float @llvm.amdgcn.rcp.f32(float %[[FAST_250]])
+; CHECK: %[[FAST_25_INS0:[0-9]+]] = insertelement <2 x float> undef, float %[[FAST_25_RCP0]], i64 0
+; CHECK: %[[FAST_251:[0-9]+]] = extractelement <2 x float> %x, i64 1
+; CHECK: %[[FAST_25_RCP1:[0-9]+]] =  call fast float @llvm.amdgcn.rcp.f32(float %[[FAST_251]])
+; CHECK: %fast.25ulp = insertelement <2 x float> %[[FAST_25_INS0]], float %[[FAST_25_RCP1]], i64 1
 ; CHECK: store volatile <2 x float> %fast.25ulp, <2 x float> addrspace(1)* %out
 define amdgpu_kernel void @rcp_fdiv_fpmath_vector(<2 x float> addrspace(1)* %out, <2 x float> %x) #1 {
   %no.md = fdiv <2 x float> <float 1.0, float 1.0>, %x
@@ -136,12 +226,48 @@
 }
 
 ; CHECK-LABEL: @rcp_fdiv_fpmath_vector_nonsplat(
-; CHECK: %no.md = fdiv <2 x float> <float 1.000000e+00, float 2.000000e+00>, %x
-; CHECK: %arcp.no.md = fdiv arcp <2 x float> <float 1.000000e+00, float 2.000000e+00>, %x
-; CHECK: %fast.no.md = fdiv fast <2 x float> <float 1.000000e+00, float 2.000000e+00>, %x{{$}}
-; CHECK: %arcp.25ulp = fdiv arcp <2 x float> <float 1.000000e+00, float 2.000000e+00>, %x, !fpmath !0
-; CHECK: %fast.25ulp = fdiv fast <2 x float> <float 1.000000e+00, float 2.000000e+00>, %x, !fpmath !0
-; CHECK: store volatile <2 x float> %fast.25ulp
+; CHECK: %[[NO0:[0-9]+]] =  extractelement <2 x float> %x, i64 0
+; CHECK: %[[NO_FDIV0:[0-9]+]] = fdiv float 1.000000e+00, %[[NO0]]
+; CHECK: %[[NO_INS0:[0-9]+]] = insertelement <2 x float> undef, float %[[NO_FDIV0]], i64 0
+; CHECK: %[[NO1:[0-9]+]] = extractelement <2 x float> %x, i64 1
+; CHECK: %[[NO_FDIV1:[0-9]+]] = fdiv float 2.000000e+00, %[[NO1]]
+; CHECK: %no.md = insertelement <2 x float> %[[NO_INS0]], float %[[NO_FDIV1]], i64 1
+; CHECK: store volatile <2 x float> %no.md, <2 x float> addrspace(1)* %out
+
+; CHECK: %[[ARCP_NO0:[0-9]+]] =  extractelement <2 x float> %x, i64 0
+; CHECK: %[[ARCP_NO_FDIV0:[0-9]+]] = fdiv arcp float 1.000000e+00, %[[ARCP_NO0]]
+; CHECK: %[[ARCP_NO_INS0:[0-9]+]] = insertelement <2 x float> undef, float %[[ARCP_NO_FDIV0]], i64 0
+; CHECK: %[[ARCP_NO1:[0-9]+]] = extractelement <2 x float> %x, i64 1
+; CHECK: %[[ARCP_NO_FDIV1:[0-9]+]] =  fdiv arcp float 2.000000e+00, %[[ARCP_NO1]]
+; CHECK: %arcp.no.md = insertelement <2 x float> %[[ARCP_NO_INS0]], float %[[ARCP_NO_FDIV1]], i64 1
+; CHECK: store volatile <2 x float> %arcp.no.md, <2 x float> addrspace(1)* %out
+
+; CHECK: %[[FAST_NO0:[0-9]+]] =  extractelement <2 x float> %x, i64 0
+; CHECK: %[[FAST_NO_RCP0:[0-9]+]] = call fast float @llvm.amdgcn.rcp.f32(float %[[FAST_NO0]])
+; CHECK: %[[FAST_NO_INS0:[0-9]+]] = insertelement <2 x float> undef, float %[[FAST_NO_RCP0]], i64 0
+; CHECK: %[[FAST_NO1:[0-9]+]] = extractelement <2 x float> %x, i64 1
+; CHECK: %[[FAST_NO_RCP1:[0-9]+]] =  call fast float @llvm.amdgcn.rcp.f32(float %[[FAST_NO1]])
+; CHECK: %[[FAST_NO_MUL1:[0-9]+]] = fmul fast float 2.000000e+00, %[[FAST_NO_RCP1]]
+; CHECK: %fast.no.md = insertelement <2 x float> %[[FAST_NO_INS0]], float %[[FAST_NO_MUL1]], i64 1
+; CHECK: store volatile <2 x float> %fast.no.md, <2 x float> addrspace(1)* %out
+
+; CHECK: %[[ARCP_250:[0-9]+]] =  extractelement <2 x float> %x, i64 0
+; CHECK: %[[ARCP_25_RCP0:[0-9]+]] = call arcp float @llvm.amdgcn.rcp.f32(float %[[ARCP_250]])
+; CHECK: %[[ARCP_25_INS0:[0-9]+]] = insertelement <2 x float> undef, float %[[ARCP_25_RCP0]], i64 0
+; CHECK: %[[ARCP_251:[0-9]+]] = extractelement <2 x float> %x, i64 1
+; CHECK: %[[ARCP_25_RCP1:[0-9]+]] =  call arcp float @llvm.amdgcn.rcp.f32(float %[[ARCP_251]])
+; CHECK: %[[ARCP_25_MUL1:[0-9]+]] = fmul arcp float 2.000000e+00, %[[ARCP_25_RCP1]]
+; CHECK: %arcp.25ulp = insertelement <2 x float> %[[ARCP_25_INS0]], float %[[ARCP_25_MUL1]], i64 1
+; CHECK: store volatile <2 x float> %arcp.25ulp, <2 x float> addrspace(1)* %out
+
+; CHECK: %[[FAST_250:[0-9]+]] =  extractelement <2 x float> %x, i64 0
+; CHECK: %[[FAST_25_RCP0:[0-9]+]] = call fast float @llvm.amdgcn.rcp.f32(float %[[FAST_250]])
+; CHECK: %[[FAST_25_INS0:[0-9]+]] = insertelement <2 x float> undef, float %[[FAST_25_RCP0]], i64 0
+; CHECK: %[[FAST_251:[0-9]+]] = extractelement <2 x float> %x, i64 1
+; CHECK: %[[FAST_25_RCP1:[0-9]+]] =  call fast float @llvm.amdgcn.rcp.f32(float %[[FAST_251]])
+; CHECK: %[[FAST_25_MUL1:[0-9]+]] = fmul fast float 2.000000e+00, %[[FAST_25_RCP1]]
+; CHECK: %fast.25ulp = insertelement <2 x float> %[[FAST_25_INS0]], float %[[FAST_25_MUL1]], i64 1
+; CHECK: store volatile <2 x float> %fast.25ulp, <2 x float> addrspace(1)* %out
 define amdgpu_kernel void @rcp_fdiv_fpmath_vector_nonsplat(<2 x float> addrspace(1)* %out, <2 x float> %x) #1 {
   %no.md = fdiv <2 x float> <float 1.0, float 2.0>, %x
   store volatile <2 x float> %no.md, <2 x float> addrspace(1)* %out
@@ -161,12 +287,29 @@
   ret void
 }
 
-; FIXME: Should be able to get fdiv for 1.0 component
 ; CHECK-LABEL: @rcp_fdiv_fpmath_vector_partial_constant(
-; CHECK: %arcp.25ulp = fdiv arcp <2 x float> %x.insert, %y, !fpmath !0
+; CHECK: %[[ARCP_A0:[0-9]+]] = extractelement <2 x float> %x.insert, i64 0
+; CHECK: %[[ARCP_B0:[0-9]+]] = extractelement <2 x float> %y, i64 0
+; CHECK: %[[ARCP_RCP0:[0-9]+]] = call arcp float @llvm.amdgcn.rcp.f32(float %[[ARCP_B0]])
+; CHECK: %[[ARCP_MUL0:[0-9]+]] = fmul arcp float %[[ARCP_A0]], %[[ARCP_RCP0]], !fpmath !0
+; CHECK: %[[ARCP_INS0:[0-9]+]] = insertelement <2 x float> undef, float %[[ARCP_MUL0]], i64 0
+; CHECK: %[[ARCP_A1:[0-9]+]] = extractelement <2 x float> %x.insert, i64 1
+; CHECK: %[[ARCP_B1:[0-9]+]] = extractelement <2 x float> %y, i64 1
+; CHECK: %[[ARCP_RCP1:[0-9]+]] = call arcp float @llvm.amdgcn.rcp.f32(float %[[ARCP_B1]])
+; CHECK: %[[ARCP_MUL1:[0-9]+]] = fmul arcp float %[[ARCP_A1]], %[[ARCP_RCP1]], !fpmath !0
+; CHECK: %arcp.25ulp = insertelement <2 x float> %[[ARCP_INS0]], float %[[ARCP_MUL1]], i64 1
 ; CHECK: store volatile <2 x float> %arcp.25ulp
 
-; CHECK: %fast.25ulp = fdiv fast <2 x float> %x.insert, %y, !fpmath !0
+; CHECK: %[[FAST_A0:[0-9]+]] = extractelement <2 x float> %x.insert, i64 0
+; CHECK: %[[FAST_B0:[0-9]+]] = extractelement <2 x float> %y, i64 0
+; CHECK: %[[FAST_RCP0:[0-9]+]] = call fast float @llvm.amdgcn.rcp.f32(float %[[FAST_B0]])
+; CHECK: %[[FAST_MUL0:[0-9]+]] = fmul fast float %[[FAST_A0]], %[[FAST_RCP0]], !fpmath !0
+; CHECK: %[[FAST_INS0:[0-9]+]] = insertelement <2 x float> undef, float %[[FAST_MUL0]], i64 0
+; CHECK: %[[FAST_A1:[0-9]+]] = extractelement <2 x float> %x.insert, i64 1
+; CHECK: %[[FAST_B1:[0-9]+]] = extractelement <2 x float> %y, i64 1
+; CHECK: %[[FAST_RCP1:[0-9]+]] = call fast float @llvm.amdgcn.rcp.f32(float %[[FAST_B1]])
+; CHECK: %[[FAST_MUL1:[0-9]+]] = fmul fast float %[[FAST_A1]], %[[FAST_RCP1]], !fpmath !0
+; CHECK: %fast.25ulp = insertelement <2 x float> %[[FAST_INS0]], float %[[FAST_MUL1]], i64 1
 ; CHECK: store volatile <2 x float> %fast.25ulp
 define amdgpu_kernel void @rcp_fdiv_fpmath_vector_partial_constant(<2 x float> addrspace(1)* %out, <2 x float> %x, <2 x float> %y) #1 {
   %x.insert = insertelement <2 x float> %x, float 1.0, i32 0
@@ -186,8 +329,9 @@
 ; CHECK: %md.1ulp = fdiv float %a, %b, !fpmath !2
 ; CHECK: %md.25ulp = fdiv float %a, %b, !fpmath !0
 ; CHECK: %md.3ulp = fdiv float %a, %b, !fpmath !3
-; CHECK: %fast.md.25ulp = fdiv fast float %a, %b, !fpmath !0
-; CHECK: %arcp.md.25ulp = fdiv arcp float %a, %b, !fpmath !0
+; CHECK: %[[RCP_FAST:[0-9]+]] = call fast float @llvm.amdgcn.rcp.f32(float %b)
+; CHECK: %fast.md.25ulp = fmul fast float %a, %[[RCP_FAST]], !fpmath !0
+; CHECK: %arcp.md.25ulp  = fdiv arcp float %a, %b, !fpmath !0
 define amdgpu_kernel void @fdiv_fpmath_f32_denormals(float addrspace(1)* %out, float %a, float %b) #2 {
   %no.md = fdiv float %a, %b
   store volatile float %no.md, float addrspace(1)* %out
Index: llvm/test/CodeGen/AMDGPU/fdiv.ll
===================================================================
--- llvm/test/CodeGen/AMDGPU/fdiv.ll
+++ llvm/test/CodeGen/AMDGPU/fdiv.ll
@@ -284,6 +284,68 @@
   ret void
 }
 
+; FUNC-LABEL: {{^}}fdiv_f32_correctly_rounded_divide_sqrt:
+
+; GCN: v_div_scale_f32 [[NUM_SCALE:v[0-9]+]]
+; GCN-DAG: v_div_scale_f32 [[DEN_SCALE:v[0-9]+]]
+; GCN-DAG: v_rcp_f32_e32 [[NUM_RCP:v[0-9]+]], [[NUM_SCALE]]
+
+; PREGFX10: s_setreg_imm32_b32 hwreg(HW_REG_MODE, 4, 2), 3
+; GFX10: s_denorm_mode 15
+; GCN: v_fma_f32 [[A:v[0-9]+]], -[[NUM_SCALE]], [[NUM_RCP]], 1.0
+; GCN: v_fma_f32 [[B:v[0-9]+]], [[A]], [[NUM_RCP]], [[NUM_RCP]]
+; GCN: v_mul_f32_e32 [[C:v[0-9]+]], [[DEN_SCALE]], [[B]]
+; GCN: v_fma_f32 [[D:v[0-9]+]], -[[NUM_SCALE]], [[C]], [[DEN_SCALE]]
+; GCN: v_fma_f32 [[E:v[0-9]+]], [[D]], [[B]], [[C]]
+; GCN: v_fma_f32 [[F:v[0-9]+]], -[[NUM_SCALE]], [[E]], [[DEN_SCALE]]
+; PREGFX10: s_setreg_imm32_b32 hwreg(HW_REG_MODE, 4, 2), 0
+; GFX10: s_denorm_mode 12
+; GCN: v_div_fmas_f32 [[FMAS:v[0-9]+]], [[F]], [[B]], [[E]]
+; GCN: v_div_fixup_f32 v{{[0-9]+}}, [[FMAS]],
+
+define amdgpu_kernel void @fdiv_f32_correctly_rounded_divide_sqrt(float addrspace(1)* %out, float %a) #0 {
+entry:
+  %fdiv = fdiv float 1.000000e+00, %a
+  store float %fdiv, float addrspace(1)* %out
+  ret void
+}
+
+
+; FUNC-LABEL: {{^}}fdiv_f32_denorms_correctly_rounded_divide_sqrt:
+
+; GCN: v_div_scale_f32 [[NUM_SCALE:v[0-9]+]]
+; GCN-DAG: v_rcp_f32_e32 [[NUM_RCP:v[0-9]+]], [[NUM_SCALE]]
+
+; PREGFX10-DAG: v_div_scale_f32 [[DEN_SCALE:v[0-9]+]]
+; PREGFX10-NOT: s_setreg
+; PREGFX10: v_fma_f32 [[A:v[0-9]+]], -[[NUM_SCALE]], [[NUM_RCP]], 1.0
+; PREGFX10: v_fma_f32 [[B:v[0-9]+]], [[A]], [[NUM_RCP]], [[NUM_RCP]]
+; PREGFX10: v_mul_f32_e32 [[C:v[0-9]+]], [[DEN_SCALE]], [[B]]
+; PREGFX10: v_fma_f32 [[D:v[0-9]+]], -[[NUM_SCALE]], [[C]], [[DEN_SCALE]]
+; PREGFX10: v_fma_f32 [[E:v[0-9]+]], [[D]], [[B]], [[C]]
+; PREGFX10: v_fma_f32 [[F:v[0-9]+]], -[[NUM_SCALE]], [[E]], [[DEN_SCALE]]
+; PREGFX10-NOT: s_setreg
+
+; GFX10-NOT: s_denorm_mode
+; GFX10: v_fma_f32 [[A:v[0-9]+]], -[[NUM_SCALE]], [[NUM_RCP]], 1.0
+; GFX10: v_fmac_f32_e32 [[B:v[0-9]+]], [[A]], [[NUM_RCP]]
+; GFX10: v_div_scale_f32 [[DEN_SCALE:v[0-9]+]]
+; GFX10: v_mul_f32_e32 [[C:v[0-9]+]], [[DEN_SCALE]], [[B]]
+; GFX10: v_fma_f32 [[D:v[0-9]+]], [[C]], -[[NUM_SCALE]], [[DEN_SCALE]]
+; GFX10: v_fmac_f32_e32 [[E:v[0-9]+]], [[D]], [[B]]
+; GFX10: v_fmac_f32_e64 [[F:v[0-9]+]], -[[NUM_SCALE]], [[E]]
+; GFX10-NOT: s_denorm_mode
+
+; GCN: v_div_fmas_f32 [[FMAS:v[0-9]+]], [[F]], [[B]], [[E]]
+; GCN: v_div_fixup_f32 v{{[0-9]+}}, [[FMAS]],
+define amdgpu_kernel void @fdiv_f32_denorms_correctly_rounded_divide_sqrt(float addrspace(1)* %out, float %a) #2 {
+entry:
+  %fdiv = fdiv float 1.000000e+00, %a
+  store float %fdiv, float addrspace(1)* %out
+  ret void
+}
+
+
 attributes #0 = { nounwind "enable-unsafe-fp-math"="false" "target-features"="-fp32-denormals,+fp64-fp16-denormals,-flat-for-global" }
 attributes #1 = { nounwind "enable-unsafe-fp-math"="true" "target-features"="-fp32-denormals,-flat-for-global" }
 attributes #2 = { nounwind "enable-unsafe-fp-math"="false" "target-features"="+fp32-denormals,-flat-for-global" }
Index: llvm/test/CodeGen/AMDGPU/fdiv32-to-rcp-folding.ll
===================================================================
--- llvm/test/CodeGen/AMDGPU/fdiv32-to-rcp-folding.ll
+++ llvm/test/CodeGen/AMDGPU/fdiv32-to-rcp-folding.ll
@@ -348,7 +348,7 @@
 ; GCN: global_store_dword v[{{[0-9:]+}}], [[RCP]], off
 define amdgpu_kernel void @div_1_by_x_fast(float addrspace(1)* %arg) {
   %load = load float, float addrspace(1)* %arg, align 4
-  %div = fdiv fast float 1.000000e+00, %load
+  %div = fdiv fast float 1.000000e+00, %load, !fpmath !0
   store float %div, float addrspace(1)* %arg, align 4
   ret void
 }
@@ -359,7 +359,7 @@
 ; GCN: global_store_dword v[{{[0-9:]+}}], [[RCP]], off
 define amdgpu_kernel void @div_minus_1_by_x_fast(float addrspace(1)* %arg) {
   %load = load float, float addrspace(1)* %arg, align 4
-  %div = fdiv fast float -1.000000e+00, %load
+  %div = fdiv fast float -1.000000e+00, %load, !fpmath !0
   store float %div, float addrspace(1)* %arg, align 4
   ret void
 }
@@ -370,7 +370,7 @@
 ; GCN: global_store_dword v[{{[0-9:]+}}], [[RCP]], off
 define amdgpu_kernel void @div_1_by_minus_x_fast(float addrspace(1)* %arg) {
   %load = load float, float addrspace(1)* %arg, align 4
-  %neg = fsub float -0.000000e+00, %load
+  %neg = fsub float -0.000000e+00, %load, !fpmath !0
   %div = fdiv fast float 1.000000e+00, %neg
   store float %div, float addrspace(1)* %arg, align 4
   ret void
@@ -382,22 +382,18 @@
 ; GCN: global_store_dword v[{{[0-9:]+}}], [[RCP]], off
 define amdgpu_kernel void @div_minus_1_by_minus_x_fast(float addrspace(1)* %arg) {
   %load = load float, float addrspace(1)* %arg, align 4
-  %neg = fsub float -0.000000e+00, %load
+  %neg = fsub float -0.000000e+00, %load, !fpmath !0
   %div = fdiv fast float -1.000000e+00, %neg
   store float %div, float addrspace(1)* %arg, align 4
   ret void
 }
 
 ; GCN-LABEL: {{^}}div_1_by_x_correctly_rounded:
-; GCN-DENORM-DAG: v_div_scale_f32
-; GCN-DENORM-DAG: v_rcp_f32_e32
-; GCN-DENORM-DAG: v_div_scale_f32
-; GCN-DENORM:     v_div_fmas_f32
-; GCN-DENORM:     v_div_fixup_f32
-
-; GCN-FLUSH: s_load_dword [[VAL:s[0-9]+]], s[0:1], 0x0
-; GCN-FLUSH: v_rcp_f32_e32 [[RCP:v[0-9]+]], [[VAL]]
-; GCN-FLUSH: global_store_dword v[{{[0-9:]+}}], [[RCP]], off
+; GCN-DAG: v_div_scale_f32
+; GCN-DAG: v_rcp_f32_e32
+; GCN-DAG: v_div_scale_f32
+; GCN:     v_div_fmas_f32
+; GCN:     v_div_fixup_f32
 define amdgpu_kernel void @div_1_by_x_correctly_rounded(float addrspace(1)* %arg) {
   %load = load float, float addrspace(1)* %arg, align 4
   %div = fdiv float 1.000000e+00, %load
@@ -406,15 +402,11 @@
 }
 
 ; GCN-LABEL: {{^}}div_minus_1_by_x_correctly_rounded:
-; GCN-DENORM-DAG: v_div_scale_f32
-; GCN-DENORM-DAG: v_rcp_f32_e32
-; GCN-DENORM-DAG: v_div_scale_f32
-; GCN-DENORM:     v_div_fmas_f32
-; GCN-DENORM:     v_div_fixup_f32
-
-; GCN-FLUSH: s_load_dword [[VAL:s[0-9]+]], s[0:1], 0x0
-; GCN-FLUSH: v_rcp_f32_e64 [[RCP:v[0-9]+]], -[[VAL]]
-; GCN-FLUSH: global_store_dword v[{{[0-9:]+}}], [[RCP]], off
+; GCN-DAG: v_div_scale_f32
+; GCN-DAG: v_rcp_f32_e32
+; GCN-DAG: v_div_scale_f32
+; GCN:     v_div_fmas_f32
+; GCN:     v_div_fixup_f32
 define amdgpu_kernel void @div_minus_1_by_x_correctly_rounded(float addrspace(1)* %arg) {
   %load = load float, float addrspace(1)* %arg, align 4
   %div = fdiv float -1.000000e+00, %load
@@ -423,15 +415,11 @@
 }
 
 ; GCN-LABEL: {{^}}div_1_by_minus_x_correctly_rounded:
-; GCN-DENORM-DAG: v_div_scale_f32
-; GCN-DENORM-DAG: v_rcp_f32_e32
-; GCN-DENORM-DAG: v_div_scale_f32
-; GCN-DENORM:     v_div_fmas_f32
-; GCN-DENORM:     v_div_fixup_f32
-
-; GCN-FLUSH: s_load_dword [[VAL:s[0-9]+]], s[0:1], 0x0
-; GCN-FLUSH: v_rcp_f32_e64 [[RCP:v[0-9]+]], -[[VAL]]
-; GCN-FLUSH: global_store_dword v[{{[0-9:]+}}], [[RCP]], off
+; GCN-DAG: v_div_scale_f32
+; GCN-DAG: v_rcp_f32_e32
+; GCN-DAG: v_div_scale_f32
+; GCN:     v_div_fmas_f32
+; GCN:     v_div_fixup_f32
 define amdgpu_kernel void @div_1_by_minus_x_correctly_rounded(float addrspace(1)* %arg) {
   %load = load float, float addrspace(1)* %arg, align 4
   %neg = fsub float -0.000000e+00, %load
@@ -441,15 +429,11 @@
 }
 
 ; GCN-LABEL: {{^}}div_minus_1_by_minus_x_correctly_rounded:
-; GCN-DENORM-DAG: v_div_scale_f32
-; GCN-DENORM-DAG: v_rcp_f32_e32
-; GCN-DENORM-DAG: v_div_scale_f32
-; GCN-DENORM:     v_div_fmas_f32
-; GCN-DENORM:     v_div_fixup_f32
-
-; GCN-FLUSH: s_load_dword [[VAL:s[0-9]+]], s[0:1], 0x0
-; GCN-FLUSH: v_rcp_f32_e32 [[RCP:v[0-9]+]], [[VAL]]
-; GCN-FLUSH: global_store_dword v[{{[0-9:]+}}], [[RCP]], off
+; GCN-DAG: v_div_scale_f32
+; GCN-DAG: v_rcp_f32_e32
+; GCN-DAG: v_div_scale_f32
+; GCN:     v_div_fmas_f32
+; GCN:     v_div_fixup_f32
 define amdgpu_kernel void @div_minus_1_by_minus_x_correctly_rounded(float addrspace(1)* %arg) {
   %load = load float, float addrspace(1)* %arg, align 4
   %neg = fsub float -0.000000e+00, %load
Index: llvm/test/CodeGen/AMDGPU/fneg-combines.ll
===================================================================
--- llvm/test/CodeGen/AMDGPU/fneg-combines.ll
+++ llvm/test/CodeGen/AMDGPU/fneg-combines.ll
@@ -219,13 +219,30 @@
 ; GCN-SAFE-DAG: v_mad_f32 [[A:v[0-9]+]],
 ; GCN-SAFE-DAG: v_cmp_ngt_f32_e32 {{.*}}, [[A]]
 ; GCN-SAFE-DAG: v_cndmask_b32_e64 v{{[0-9]+}}, -[[A]]
+define amdgpu_ps float @fneg_fadd_0(float inreg %tmp2, float inreg %tmp6, <4 x i32> %arg) local_unnamed_addr #0 {
+.entry:
+  %tmp7 = fdiv float 1.000000e+00, %tmp6
+  %tmp8 = fmul float 0.000000e+00, %tmp7
+  %tmp9 = fmul reassoc nnan arcp contract float 0.000000e+00, %tmp8
+  %.i188 = fadd float %tmp9, 0.000000e+00
+  %tmp10 = fcmp uge float %.i188, %tmp2
+  %tmp11 = fsub float -0.000000e+00, %.i188
+  %.i092 = select i1 %tmp10, float %tmp2, float %tmp11
+  %tmp12 = fcmp ule float %.i092, 0.000000e+00
+  %.i198 = select i1 %tmp12, float 0.000000e+00, float 0x7FF8000000000000
+  ret float %.i198
+}
+
+; This is a workaround because -enable-no-signed-zeros-fp-math does not set up
+; function attribute unsafe-fp-math automatically. Combine with the previous test
+; when that is done.
+; GCN-LABEL: {{^}}fneg_fadd_0_nsz:
 ; GCN-NSZ-DAG: v_rcp_f32_e32 [[A:v[0-9]+]],
 ; GCN-NSZ-DAG: v_mov_b32_e32 [[B:v[0-9]+]],
 ; GCN-NSZ-DAG: v_mov_b32_e32 [[C:v[0-9]+]],
 ; GCN-NSZ-DAG: v_mul_f32_e32 [[D:v[0-9]+]],
 ; GCN-NSZ-DAG: v_cmp_nlt_f32_e64 {{.*}}, -[[D]]
-
-define amdgpu_ps float @fneg_fadd_0(float inreg %tmp2, float inreg %tmp6, <4 x i32> %arg) local_unnamed_addr #0 {
+define amdgpu_ps float @fneg_fadd_0_nsz(float inreg %tmp2, float inreg %tmp6, <4 x i32> %arg) local_unnamed_addr #2 {
 .entry:
   %tmp7 = fdiv float 1.000000e+00, %tmp6
   %tmp8 = fmul float 0.000000e+00, %tmp7
@@ -2524,3 +2541,4 @@
 
 attributes #0 = { nounwind }
 attributes #1 = { nounwind readnone }
+attributes #2 = { nounwind "unsafe-fp-math"="true" }
Index: llvm/test/CodeGen/AMDGPU/known-never-snan.ll
===================================================================
--- llvm/test/CodeGen/AMDGPU/known-never-snan.ll
+++ llvm/test/CodeGen/AMDGPU/known-never-snan.ll
@@ -11,7 +11,7 @@
 ; GCN-NEXT:    v_rcp_f32_e32 v0, v0
 ; GCN-NEXT:    v_med3_f32 v0, |v0|, 2.0, 4.0
 ; GCN-NEXT:    s_setpc_b64 s[30:31]
-  %a.nnan.add = fdiv nnan float 1.0, %a
+  %a.nnan.add = fdiv nnan float 1.0, %a, !fpmath !0
   %known.not.snan = call float @llvm.fabs.f32(float %a.nnan.add)
   %max = call float @llvm.maxnum.f32(float %known.not.snan, float 2.0)
   %med = call float @llvm.minnum.f32(float %max, float 4.0)
@@ -22,10 +22,10 @@
 ; GCN-LABEL: v_test_known_not_snan_fneg_input_fmed3_r_i_i_f32:
 ; GCN:       ; %bb.0:
 ; GCN-NEXT:    s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
-; GCN-NEXT:    v_rcp_f32_e64 v0, -v0
-; GCN-NEXT:    v_med3_f32 v0, v0, 2.0, 4.0
+; GCN-NEXT:    v_rcp_f32_e32 v0, v0
+; GCN-NEXT:    v_med3_f32 v0, -v0, 2.0, 4.0
 ; GCN-NEXT:    s_setpc_b64 s[30:31]
-  %a.nnan.add = fdiv nnan float 1.0, %a
+  %a.nnan.add = fdiv nnan float 1.0, %a, !fpmath !0
   %known.not.snan = fsub float -0.0, %a.nnan.add
   %max = call float @llvm.maxnum.f32(float %known.not.snan, float 2.0)
   %med = call float @llvm.minnum.f32(float %max, float 4.0)
@@ -71,7 +71,7 @@
 ; GCN-NEXT:    v_bfi_b32 v0, s4, v0, v1
 ; GCN-NEXT:    v_med3_f32 v0, v0, 2.0, 4.0
 ; GCN-NEXT:    s_setpc_b64 s[30:31]
-  %a.nnan.add = fdiv nnan float 1.0, %a
+  %a.nnan.add = fdiv nnan float 1.0, %a, !fpmath !0
   %known.not.snan = call float @llvm.copysign.f32(float %a.nnan.add, float %sign)
   %max = call float @llvm.maxnum.f32(float %known.not.snan, float 2.0)
   %med = call float @llvm.minnum.f32(float %max, float 4.0)
@@ -101,7 +101,7 @@
 ; GCN-NEXT:    v_min_f32_e32 v0, v0, v1
 ; GCN-NEXT:    v_med3_f32 v0, v0, 2.0, 4.0
 ; GCN-NEXT:    s_setpc_b64 s[30:31]
-  %a.nnan.add = fdiv nnan float 1.0, %a
+  %a.nnan.add = fdiv nnan float 1.0, %a, !fpmath !0
   %b.nnan.add = fadd nnan float %b, 1.0
   %known.not.snan = call float @llvm.minnum.f32(float %a.nnan.add, float %b.nnan.add)
   %max = call float @llvm.maxnum.f32(float %known.not.snan, float 2.0)
@@ -166,7 +166,7 @@
 ; GCN-NEXT:    v_min_f32_e32 v0, v0, v1
 ; GCN-NEXT:    v_med3_f32 v0, v0, 2.0, 4.0
 ; GCN-NEXT:    s_setpc_b64 s[30:31]
-  %a.nnan.add = fdiv nnan float 1.0, %a
+  %a.nnan.add = fdiv nnan float 1.0, %a, !fpmath !0
   %known.not.snan = call float @llvm.minnum.f32(float %a.nnan.add, float %b)
   %max = call float @llvm.maxnum.f32(float %known.not.snan, float 2.0)
   %med = call float @llvm.minnum.f32(float %max, float 4.0)
@@ -182,7 +182,7 @@
 ; GCN-NEXT:    v_max_f32_e32 v0, v0, v1
 ; GCN-NEXT:    v_med3_f32 v0, v0, 2.0, 4.0
 ; GCN-NEXT:    s_setpc_b64 s[30:31]
-  %a.nnan.add = fdiv nnan float 1.0, %a
+  %a.nnan.add = fdiv nnan float 1.0, %a, !fpmath !0
   %b.nnan.add = fadd nnan float %b, 1.0
   %known.not.snan = call float @llvm.maxnum.f32(float %a.nnan.add, float %b.nnan.add)
   %max = call float @llvm.maxnum.f32(float %known.not.snan, float 2.0)
@@ -215,7 +215,7 @@
 ; GCN-NEXT:    v_max_f32_e32 v0, v0, v1
 ; GCN-NEXT:    v_med3_f32 v0, v0, 2.0, 4.0
 ; GCN-NEXT:    s_setpc_b64 s[30:31]
-  %a.nnan.add = fdiv nnan float 1.0, %a
+  %a.nnan.add = fdiv nnan float 1.0, %a, !fpmath !0
   %known.not.snan = call float @llvm.maxnum.f32(float %a.nnan.add, float %b)
   %max = call float @llvm.maxnum.f32(float %known.not.snan, float 2.0)
   %med = call float @llvm.minnum.f32(float %max, float 4.0)
@@ -232,7 +232,7 @@
 ; GCN-NEXT:    v_cndmask_b32_e32 v0, v1, v0, vcc
 ; GCN-NEXT:    v_med3_f32 v0, v0, 2.0, 4.0
 ; GCN-NEXT:    s_setpc_b64 s[30:31]
-  %a.nnan.add = fdiv nnan float 1.0, %a
+  %a.nnan.add = fdiv nnan float 1.0, %a, !fpmath !0
   %b.nnan.add = fadd nnan float %b, 1.0
   %cmp = icmp eq i32 %c, 0
   %known.not.snan = select i1 %cmp, float %a.nnan.add, float %b.nnan.add
@@ -269,7 +269,7 @@
 ; GCN-NEXT:    v_mul_f32_e32 v0, 1.0, v0
 ; GCN-NEXT:    v_med3_f32 v0, v0, 2.0, 4.0
 ; GCN-NEXT:    s_setpc_b64 s[30:31]
-  %a.nnan.add = fdiv nnan float 1.0, %a
+  %a.nnan.add = fdiv nnan float 1.0, %a, !fpmath !0
   %cmp = icmp eq i32 %c, 0
   %known.not.snan = select i1 %cmp, float %a.nnan.add, float %b
   %max = call float @llvm.maxnum.f32(float %known.not.snan, float 2.0)
@@ -669,3 +669,5 @@
 
 attributes #0 = { nounwind }
 attributes #1 = { nounwind readnone speculatable }
+
+!0 = !{float 2.500000e+00}
Index: llvm/test/CodeGen/AMDGPU/llvm.amdgcn.rcp.ll
===================================================================
--- llvm/test/CodeGen/AMDGPU/llvm.amdgcn.rcp.ll
+++ llvm/test/CodeGen/AMDGPU/llvm.amdgcn.rcp.ll
@@ -37,7 +37,7 @@
 ; SI-NOT: [[RESULT]]
 ; SI: buffer_store_dword [[RESULT]]
 define amdgpu_kernel void @safe_no_fp32_denormals_rcp_f32(float addrspace(1)* %out, float %src) #1 {
-  %rcp = fdiv float 1.0, %src
+  %rcp = fdiv float 1.0, %src, !fpmath !0
   store float %rcp, float addrspace(1)* %out, align 4
   ret void
 }
@@ -47,7 +47,7 @@
 ; SI-NOT: [[RESULT]]
 ; SI: buffer_store_dword [[RESULT]]
 define amdgpu_kernel void @safe_f32_denormals_rcp_pat_f32(float addrspace(1)* %out, float %src) #4 {
-  %rcp = fdiv float 1.0, %src
+  %rcp = fdiv float 1.0, %src, !fpmath !0
   store float %rcp, float addrspace(1)* %out, align 4
   ret void
 }
@@ -61,8 +61,7 @@
 }
 
 ; FUNC-LABEL: {{^}}safe_rsq_rcp_pat_f32:
-; SI: v_sqrt_f32_e32
-; SI: v_rcp_f32_e32
+; SI: v_rsq_f32_e32
 define amdgpu_kernel void @safe_rsq_rcp_pat_f32(float addrspace(1)* %out, float %src) #1 {
   %sqrt = call float @llvm.sqrt.f32(float %src)
   %rcp = call float @llvm.amdgcn.rcp.f32(float %sqrt)
@@ -144,3 +143,5 @@
 attributes #2 = { nounwind "unsafe-fp-math"="true" "target-features"="-fp32-denormals" }
 attributes #3 = { nounwind "unsafe-fp-math"="false" "target-features"="+fp32-denormals" }
 attributes #4 = { nounwind "unsafe-fp-math"="true" "target-features"="+fp32-denormals" }
+
+!0 = !{float 2.500000e+00}
Index: llvm/test/CodeGen/AMDGPU/mul24-pass-ordering.ll
===================================================================
--- llvm/test/CodeGen/AMDGPU/mul24-pass-ordering.ll
+++ llvm/test/CodeGen/AMDGPU/mul24-pass-ordering.ll
@@ -112,7 +112,7 @@
 
 bb19:                                             ; preds = %bb
   %tmp20 = uitofp i32 %arg6 to float
-  %tmp21 = fdiv float 1.000000e+00, %tmp20
+  %tmp21 = fdiv float 1.000000e+00, %tmp20, !fpmath !0
   %tmp22 = and i32 %arg6, 16777215
   br label %bb23
 
@@ -258,3 +258,5 @@
 
 attributes #0 = { nounwind willreturn }
 attributes #1 = { nounwind readnone speculatable }
+
+!0 = !{float 2.500000e+00}
Index: llvm/test/CodeGen/AMDGPU/rcp-pattern.ll
===================================================================
--- llvm/test/CodeGen/AMDGPU/rcp-pattern.ll
+++ llvm/test/CodeGen/AMDGPU/rcp-pattern.ll
@@ -10,7 +10,7 @@
 
 ; EG: RECIP_IEEE
 define amdgpu_kernel void @rcp_pat_f32(float addrspace(1)* %out, float %src) #0 {
-  %rcp = fdiv float 1.0, %src
+  %rcp = fdiv float 1.0, %src, !fpmath !0
   store float %rcp, float addrspace(1)* %out, align 4
   ret void
 }
@@ -71,7 +71,7 @@
 ; EG: RECIP_IEEE
 define amdgpu_kernel void @rcp_fabs_pat_f32(float addrspace(1)* %out, float %src) #0 {
   %src.fabs = call float @llvm.fabs.f32(float %src)
-  %rcp = fdiv float 1.0, %src.fabs
+  %rcp = fdiv float 1.0, %src.fabs, !fpmath !0
   store float %rcp, float addrspace(1)* %out, align 4
   ret void
 }
@@ -83,7 +83,7 @@
 
 ; EG: RECIP_IEEE
 define amdgpu_kernel void @neg_rcp_pat_f32(float addrspace(1)* %out, float %src) #0 {
-  %rcp = fdiv float -1.0, %src
+  %rcp = fdiv float -1.0, %src, !fpmath !0
   store float %rcp, float addrspace(1)* %out, align 4
   ret void
 }
@@ -95,7 +95,7 @@
 define amdgpu_kernel void @rcp_fabs_fneg_pat_f32(float addrspace(1)* %out, float %src) #0 {
   %src.fabs = call float @llvm.fabs.f32(float %src)
   %src.fabs.fneg = fsub float -0.0, %src.fabs
-  %rcp = fdiv float 1.0, %src.fabs.fneg
+  %rcp = fdiv float 1.0, %src.fabs.fneg, !fpmath !0
   store float %rcp, float addrspace(1)* %out, align 4
   ret void
 }
@@ -109,7 +109,7 @@
 define amdgpu_kernel void @rcp_fabs_fneg_pat_multi_use_f32(float addrspace(1)* %out, float %src) #0 {
   %src.fabs = call float @llvm.fabs.f32(float %src)
   %src.fabs.fneg = fsub float -0.0, %src.fabs
-  %rcp = fdiv float 1.0, %src.fabs.fneg
+  %rcp = fdiv float 1.0, %src.fabs.fneg, !fpmath !0
   store volatile float %rcp, float addrspace(1)* %out, align 4
 
   %other = fmul float %src, %src.fabs.fneg
Index: llvm/test/CodeGen/AMDGPU/rcp_iflag.ll
===================================================================
--- llvm/test/CodeGen/AMDGPU/rcp_iflag.ll
+++ llvm/test/CodeGen/AMDGPU/rcp_iflag.ll
@@ -5,7 +5,7 @@
 define amdgpu_kernel void @rcp_uint(i32 addrspace(1)* %in, float addrspace(1)* %out) {
   %load = load i32, i32 addrspace(1)* %in, align 4
   %cvt = uitofp i32 %load to float
-  %div = fdiv float 1.000000e+00, %cvt
+  %div = fdiv float 1.000000e+00, %cvt, !fpmath !0
   store float %div, float addrspace(1)* %out, align 4
   ret void
 }
@@ -15,7 +15,9 @@
 define amdgpu_kernel void @rcp_sint(i32 addrspace(1)* %in, float addrspace(1)* %out) {
   %load = load i32, i32 addrspace(1)* %in, align 4
   %cvt = sitofp i32 %load to float
-  %div = fdiv float 1.000000e+00, %cvt
+  %div = fdiv float 1.000000e+00, %cvt, !fpmath !0
   store float %div, float addrspace(1)* %out, align 4
   ret void
 }
+
+!0 = !{float 2.500000e+00}
Index: llvm/test/CodeGen/AMDGPU/rsq.ll
===================================================================
--- llvm/test/CodeGen/AMDGPU/rsq.ll
+++ llvm/test/CodeGen/AMDGPU/rsq.ll
@@ -11,7 +11,7 @@
 define amdgpu_kernel void @rsq_f32(float addrspace(1)* noalias %out, float addrspace(1)* noalias %in) nounwind {
   %val = load float, float addrspace(1)* %in, align 4
   %sqrt = call float @llvm.sqrt.f32(float %val) nounwind readnone
-  %div = fdiv float 1.0, %sqrt
+  %div = fdiv float 1.0, %sqrt, !fpmath !0
   store float %div, float addrspace(1)* %out, align 4
   ret void
 }
@@ -33,7 +33,7 @@
 ; SI: s_endpgm
 define amdgpu_kernel void @rsq_f32_sgpr(float addrspace(1)* noalias %out, float %val) nounwind {
   %sqrt = call float @llvm.sqrt.f32(float %val) nounwind readnone
-  %div = fdiv float 1.0, %sqrt
+  %div = fdiv float 1.0, %sqrt, !fpmath !0
   store float %div, float addrspace(1)* %out, align 4
   ret void
 }
@@ -78,13 +78,13 @@
 ; SI-SAFE: v_rcp_f32_e64 [[RSQ:v[0-9]+]], -[[SQRT]]
 ; SI-SAFE: buffer_store_dword [[RSQ]]
 
-; SI-UNSAFE: v_rsq_f32_e32 [[RSQ:v[0-9]+]], v{{[0-9]+}}
-; SI-UNSAFE: v_xor_b32_e32 [[NEG_RSQ:v[0-9]+]], 0x80000000, [[RSQ]]
-; SI-UNSAFE: buffer_store_dword [[NEG_RSQ]]
+; SI-UNSAFE: v_sqrt_f32_e32 [[SQRT:v[0-9]+]], v{{[0-9]+}}
+; SI-UNSAFE: v_rcp_f32_e64 [[RSQ:v[0-9]+]], -[[SQRT]]
+; SI-UNSAFE: buffer_store_dword [[RSQ]]
 define amdgpu_kernel void @neg_rsq_f32(float addrspace(1)* noalias %out, float addrspace(1)* noalias %in) nounwind {
   %val = load float, float addrspace(1)* %in, align 4
   %sqrt = call float @llvm.sqrt.f32(float %val)
-  %div = fdiv float -1.0, %sqrt
+  %div = fdiv float -1.0, %sqrt, !fpmath !0
   store float %div, float addrspace(1)* %out, align 4
   ret void
 }
@@ -109,14 +109,14 @@
 ; SI-SAFE: v_rcp_f32_e64 [[RSQ:v[0-9]+]], -[[SQRT]]
 ; SI-SAFE: buffer_store_dword [[RSQ]]
 
-; SI-UNSAFE: v_rsq_f32_e64 [[RSQ:v[0-9]+]], -v{{[0-9]+}}
-; SI-UNSAFE: v_xor_b32_e32 [[NEG_RSQ:v[0-9]+]], 0x80000000, [[RSQ]]
-; SI-UNSAFE: buffer_store_dword [[NEG_RSQ]]
+; SI-UNSAFE: v_sqrt_f32_e64 [[SQRT:v[0-9]+]], -v{{[0-9]+}}
+; SI-UNSAFE: v_rcp_f32_e64 [[RSQ:v[0-9]+]], -[[SQRT]]
+; SI-UNSAFE: buffer_store_dword [[RSQ]]
 define amdgpu_kernel void @neg_rsq_neg_f32(float addrspace(1)* noalias %out, float addrspace(1)* noalias %in) nounwind {
   %val = load float, float addrspace(1)* %in, align 4
   %val.fneg = fsub float -0.0, %val
   %sqrt = call float @llvm.sqrt.f32(float %val.fneg)
-  %div = fdiv float -1.0, %sqrt
+  %div = fdiv float -1.0, %sqrt, !fpmath !0
   store float %div, float addrspace(1)* %out, align 4
   ret void
 }
@@ -136,3 +136,5 @@
   store double %div, double addrspace(1)* %out, align 4
   ret void
 }
+
+!0 = !{float 2.500000e+00}