Implement AArch64 TTI interface isAsCheapAsAMove.

llvm-svn: 214159

4 files changed, 54 insertions, 3 deletions

diff --git a/llvm/lib/Target/AArch64/AArch64InstrFormats.td b/llvm/lib/Target/AArch64/AArch64InstrFormats.td
index e88c0c038c3..106f2158909 100644
--- a/llvm/lib/Target/AArch64/AArch64InstrFormats.td
+++ b/llvm/lib/Target/AArch64/AArch64InstrFormats.td
@@ -1636,7 +1636,7 @@ class AddSubRegAlias<string asm, Instruction inst, RegisterClass dstRegtype,
 
 multiclass AddSub<bit isSub, string mnemonic,
                   SDPatternOperator OpNode = null_frag> {
-  let hasSideEffects = 0 in {
+  let hasSideEffects = 0, isReMaterializable = 1, isAsCheapAsAMove = 1 in {
   // Add/Subtract immediate
   def Wri  : BaseAddSubImm<isSub, 0, GPR32sp, GPR32sp, addsub_shifted_imm32,
                            mnemonic, OpNode> {
@@ -1961,14 +1961,14 @@ class LogicalRegAlias<string asm, Instruction inst, RegisterClass regtype>
 
 multiclass LogicalImm<bits<2> opc, string mnemonic, SDNode OpNode,
                       string Alias> {
-  let AddedComplexity = 6 in
+  let AddedComplexity = 6, isReMaterializable = 1, isAsCheapAsAMove = 1 in
   def Wri : BaseLogicalImm<opc, GPR32sp, GPR32, logical_imm32, mnemonic,
                            [(set GPR32sp:$Rd, (OpNode GPR32:$Rn,
                                                logical_imm32:$imm))]> {
     let Inst{31} = 0;
     let Inst{22} = 0; // 64-bit version has an additional bit of immediate.
   }
-  let AddedComplexity = 6 in
+  let AddedComplexity = 6, isReMaterializable = 1, isAsCheapAsAMove = 1 in
   def Xri : BaseLogicalImm<opc, GPR64sp, GPR64, logical_imm64, mnemonic,
                            [(set GPR64sp:$Rd, (OpNode GPR64:$Rn,
                                                logical_imm64:$imm))]> {
@@ -2013,8 +2013,10 @@ class BaseLogicalRegPseudo<RegisterClass regtype, SDPatternOperator OpNode>
 // Split from LogicalImm as not all instructions have both.
 multiclass LogicalReg<bits<2> opc, bit N, string mnemonic,
                       SDPatternOperator OpNode> {
+  let isReMaterializable = 1, isAsCheapAsAMove = 1 in {
   def Wrr : BaseLogicalRegPseudo<GPR32, OpNode>;
   def Xrr : BaseLogicalRegPseudo<GPR64, OpNode>;
+  }
 
   def Wrs : BaseLogicalSReg<opc, N, GPR32, logical_shifted_reg32, mnemonic,
                             [(set GPR32:$Rd, (OpNode GPR32:$Rn,
diff --git a/llvm/lib/Target/AArch64/AArch64InstrInfo.cpp b/llvm/lib/Target/AArch64/AArch64InstrInfo.cpp
index 9465d90d84f..946c000cdcc 100644
--- a/llvm/lib/Target/AArch64/AArch64InstrInfo.cpp
+++ b/llvm/lib/Target/AArch64/AArch64InstrInfo.cpp
@@ -541,6 +541,51 @@ void AArch64InstrInfo::insertSelect(MachineBasicBlock &MBB,
       CC);
 }
 
+// FIXME: this implementation should be micro-architecture dependent, so a
+// micro-architecture target hook should be introduced here in future.
+bool AArch64InstrInfo::isAsCheapAsAMove(const MachineInstr *MI) const {
+  if (!Subtarget.isCortexA57() && !Subtarget.isCortexA53())
+    return MI->isAsCheapAsAMove();
+
+  switch (MI->getOpcode()) {
+  default:
+    return false;
+
+  // add/sub on register without shift
+  case AArch64::ADDWri:
+  case AArch64::ADDXri:
+  case AArch64::SUBWri:
+  case AArch64::SUBXri:
+    return (MI->getOperand(3).getImm() == 0);
+
+  // logical ops on immediate
+  case AArch64::ANDWri:
+  case AArch64::ANDXri:
+  case AArch64::EORWri:
+  case AArch64::EORXri:
+  case AArch64::ORRWri:
+  case AArch64::ORRXri:
+    return true;
+
+  // logical ops on register without shift
+  case AArch64::ANDWrr:
+  case AArch64::ANDXrr:
+  case AArch64::BICWrr:
+  case AArch64::BICXrr:
+  case AArch64::EONWrr:
+  case AArch64::EONXrr:
+  case AArch64::EORWrr:
+  case AArch64::EORXrr:
+  case AArch64::ORNWrr:
+  case AArch64::ORNXrr:
+  case AArch64::ORRWrr:
+  case AArch64::ORRXrr:
+    return true;
+  }
+
+  llvm_unreachable("Unknown opcode to check as cheap as a move!");
+}
+
 bool AArch64InstrInfo::isCoalescableExtInstr(const MachineInstr &MI,
                                              unsigned &SrcReg, unsigned &DstReg,
                                              unsigned &SubIdx) const {
diff --git a/llvm/lib/Target/AArch64/AArch64InstrInfo.h b/llvm/lib/Target/AArch64/AArch64InstrInfo.h
index 3c496869008..400367f5358 100644
--- a/llvm/lib/Target/AArch64/AArch64InstrInfo.h
+++ b/llvm/lib/Target/AArch64/AArch64InstrInfo.h
@@ -46,6 +46,8 @@ public:
 
   unsigned GetInstSizeInBytes(const MachineInstr *MI) const;
 
+  bool isAsCheapAsAMove(const MachineInstr *MI) const override;
+
   bool isCoalescableExtInstr(const MachineInstr &MI, unsigned &SrcReg,
                              unsigned &DstReg, unsigned &SubIdx) const override;
 
diff --git a/llvm/lib/Target/AArch64/AArch64Subtarget.h b/llvm/lib/Target/AArch64/AArch64Subtarget.h
index e08372f8267..8e570d5f06a 100644
--- a/llvm/lib/Target/AArch64/AArch64Subtarget.h
+++ b/llvm/lib/Target/AArch64/AArch64Subtarget.h
@@ -100,6 +100,8 @@ public:
   bool isTargetMachO() const { return TargetTriple.isOSBinFormatMachO(); }
 
   bool isCyclone() const { return CPUString == "cyclone"; }
+  bool isCortexA57() const { return CPUString == "cortex-a57"; }
+  bool isCortexA53() const { return CPUString == "cortex-a53"; }
 
   /// getMaxInlineSizeThreshold - Returns the maximum memset / memcpy size
   /// that still makes it profitable to inline the call.