[AArch64] Handle ISD::LROUND and ISD::LLROUND

This patch optimizes ISD::LROUND and ISD::LLROUND to fcvtas
instruction. It currently only handles the scalar version.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@360894 91177308-0d34-0410-b5e6-96231b3b80d8

diff --git a/include/llvm/Target/TargetSelectionDAG.td b/include/llvm/Target/TargetSelectionDAG.td
index 64b07862713e0bcdaa069e39106b28b0a33d0543..6b1ef477bfa14a2548b8d388590bd9ed6ffb4e82 100644 (file)
--- a/include/llvm/Target/TargetSelectionDAG.td
+++ b/include/llvm/Target/TargetSelectionDAG.td
@@ -450,6 +450,9 @@ def ffloor     : SDNode<"ISD::FFLOOR"     , SDTFPUnaryOp>;
 def fnearbyint : SDNode<"ISD::FNEARBYINT" , SDTFPUnaryOp>;
 def fround     : SDNode<"ISD::FROUND"     , SDTFPUnaryOp>;
 
+def lround     : SDNode<"ISD::LROUND"     , SDTFPToIntOp>;
+def llround    : SDNode<"ISD::LLROUND"    , SDTFPToIntOp>;
+
 def fpround    : SDNode<"ISD::FP_ROUND"   , SDTFPRoundOp>;
 def fpextend   : SDNode<"ISD::FP_EXTEND"  , SDTFPExtendOp>;
 def fcopysign  : SDNode<"ISD::FCOPYSIGN"  , SDTFPSignOp>;

diff --git a/lib/Target/AArch64/AArch64ISelLowering.cpp b/lib/Target/AArch64/AArch64ISelLowering.cpp
index 3ff4a225794388417fbf4f6a2100be56f610b101..43620f150b6390fc4b2901c142a2e643807bca9c 100644 (file)
--- a/lib/Target/AArch64/AArch64ISelLowering.cpp
+++ b/lib/Target/AArch64/AArch64ISelLowering.cpp
@@ -457,6 +457,8 @@ AArch64TargetLowering::AArch64TargetLowering(const TargetMachine &TM,
     setOperationAction(ISD::FMAXNUM, Ty, Legal);
     setOperationAction(ISD::FMINIMUM, Ty, Legal);
     setOperationAction(ISD::FMAXIMUM, Ty, Legal);
+    setOperationAction(ISD::LROUND, Ty, Legal);
+    setOperationAction(ISD::LLROUND, Ty, Legal);
   }
 
   if (Subtarget->hasFullFP16()) {

diff --git a/lib/Target/AArch64/AArch64InstrInfo.td b/lib/Target/AArch64/AArch64InstrInfo.td
index 2c7124bcf9f6d64898561f19018d45b00e203d04..bf6be6761a9badb1086a43dedd5712046bb28e55 100644 (file)
--- a/lib/Target/AArch64/AArch64InstrInfo.td
+++ b/lib/Target/AArch64/AArch64InstrInfo.td
@@ -3083,6 +3083,15 @@ defm : FPToIntegerPats<fp_to_uint, ftrunc, "FCVTZU">;
 defm : FPToIntegerPats<fp_to_sint, fround, "FCVTAS">;
 defm : FPToIntegerPats<fp_to_uint, fround, "FCVTAU">;
 
+def : Pat<(i64 (lround f32:$Rn)),
+          (!cast<Instruction>(FCVTASUXSr) f32:$Rn)>;
+def : Pat<(i64 (lround f64:$Rn)),
+          (!cast<Instruction>(FCVTASUXDr) f64:$Rn)>;
+def : Pat<(i64 (llround f32:$Rn)),
+          (!cast<Instruction>(FCVTASUXSr) f32:$Rn)>;
+def : Pat<(i64 (llround f64:$Rn)),
+          (!cast<Instruction>(FCVTASUXDr) f64:$Rn)>;
+
 //===----------------------------------------------------------------------===//
 // Scaled integer to floating point conversion instructions.
 //===----------------------------------------------------------------------===//

diff --git a/test/CodeGen/AArch64/llround-conv.ll b/test/CodeGen/AArch64/llround-conv.ll
index a36af0820d57ad90ac8095d632df4ae3a1be3d6d..797136037f0e942c9757e6ed35518a24ef12b768 100644 (file)
--- a/test/CodeGen/AArch64/llround-conv.ll
+++ b/test/CodeGen/AArch64/llround-conv.ll
@@ -1,7 +1,8 @@
 ; RUN: llc < %s -mtriple=aarch64 -mattr=+neon | FileCheck %s
 
 ; CHECK-LABEL: testmsws:
-; CHECK:       bl      llroundf
+; CHECK:       fcvtas  x0, s0
+; CHECK:       ret
 define i32 @testmsws(float %x) {
 entry:
   %0 = tail call i64 @llvm.llround.f32(float %x)
@@ -10,7 +11,8 @@ entry:
 }
 
 ; CHECK-LABEL: testmsxs:
-; CHECK:       b       llroundf
+; CHECK:       fcvtas  x0, s0
+; CHECK-NEXT:  ret
 define i64 @testmsxs(float %x) {
 entry:
   %0 = tail call i64 @llvm.llround.f32(float %x)
@@ -18,7 +20,8 @@ entry:
 }
 
 ; CHECK-LABEL: testmswd:
-; CHECK:       bl      llround
+; CHECK:       fcvtas  x0, d0
+; CHECK:       ret
 define i32 @testmswd(double %x) {
 entry:
   %0 = tail call i64 @llvm.llround.f64(double %x)
@@ -27,7 +30,8 @@ entry:
 }
 
 ; CHECK-LABEL: testmsxd:
-; CHECK:       b       llround
+; CHECK:       fcvtas  x0, d0
+; CHECK-NEXT:  ret
 define i64 @testmsxd(double %x) {
 entry:
   %0 = tail call i64 @llvm.llround.f64(double %x)

diff --git a/test/CodeGen/AArch64/lround-conv.ll b/test/CodeGen/AArch64/lround-conv.ll
index 483454a02f41c78bbfd38747f9ffcdf5dbd5c193..678d3149f20cccfbfbe082c1b40863973c1d4bcf 100644 (file)
--- a/test/CodeGen/AArch64/lround-conv.ll
+++ b/test/CodeGen/AArch64/lround-conv.ll
@@ -1,7 +1,8 @@
 ; RUN: llc < %s -mtriple=aarch64 -mattr=+neon | FileCheck %s
 
 ; CHECK-LABEL: testmsws:
-; CHECK:       bl      lroundf
+; CHECK:       fcvtas  x0, s0
+; CHECK:       ret
 define i32 @testmsws(float %x) {
 entry:
   %0 = tail call i64 @llvm.lround.i64.f32(float %x)
@@ -10,7 +11,8 @@ entry:
 }
 
 ; CHECK-LABEL: testmsxs:
-; CHECK:       b       lroundf
+; CHECK:       fcvtas  x0, s0
+; CHECK-NEXT:  ret
 define i64 @testmsxs(float %x) {
 entry:
   %0 = tail call i64 @llvm.lround.i64.f32(float %x)
@@ -18,7 +20,8 @@ entry:
 }
 
 ; CHECK-LABEL: testmswd:
-; CHECK:       bl      lround
+; CHECK:       fcvtas  x0, d0
+; CHECK:       ret
 define i32 @testmswd(double %x) {
 entry:
   %0 = tail call i64 @llvm.lround.i64.f64(double %x)
@@ -27,7 +30,8 @@ entry:
 }
 
 ; CHECK-LABEL: testmsxd:
-; CHECK:       b       lround
+; CHECK:       fcvtas  x0, d0
+; CHECK-NEXT:  ret
 define i64 @testmsxd(double %x) {
 entry:
   %0 = tail call i64 @llvm.lround.i64.f64(double %x)