[AArch64][RegisterBankInfo] Use the statically computed mappings for COPY

We use to resort on the generic implementation to get the mappings for
COPYs. The generic implementation resorts on table lookup and
dynamically allocated objects to get the valid mappings.

Given we already know how to map G_BITCAST and have the static mappings
for them, use that code path for COPY as well. This is much more
efficient.

Improve the compile time of RegBankSelect by up to 20%.

Note: When we eventually generate all the mappings via TableGen, we
wouldn't have to do that dance to shave compile time. The intent of this
change was to make sure that moving to static structure really pays off.

NFC.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@315781 91177308-0d34-0410-b5e6-96231b3b80d8

diff --git a/lib/Target/AArch64/AArch64RegisterBankInfo.cpp b/lib/Target/AArch64/AArch64RegisterBankInfo.cpp
index 3c505fcec16d0ce3f88c526f866497053b4bd7ef..4406ceeaddddcf2dd61294491d10fe2a8795858a 100644 (file)
--- a/lib/Target/AArch64/AArch64RegisterBankInfo.cpp
+++ b/lib/Target/AArch64/AArch64RegisterBankInfo.cpp
@@ -415,12 +415,10 @@ AArch64RegisterBankInfo::getSameKindOfOperandsMapping(
 const RegisterBankInfo::InstructionMapping &
 AArch64RegisterBankInfo::getInstrMapping(const MachineInstr &MI) const {
   const unsigned Opc = MI.getOpcode();
-  const MachineFunction &MF = *MI.getParent()->getParent();
-  const MachineRegisterInfo &MRI = MF.getRegInfo();
 
   // Try the default logic for non-generic instructions that are either copies
   // or already have some operands assigned to banks.
-  if (!isPreISelGenericOpcode(Opc) ||
+  if ((Opc != TargetOpcode::COPY && !isPreISelGenericOpcode(Opc)) ||
       Opc == TargetOpcode::G_PHI) {
     const RegisterBankInfo::InstructionMapping &Mapping =
         getInstrMappingImpl(MI);
@@ -428,6 +426,11 @@ AArch64RegisterBankInfo::getInstrMapping(const MachineInstr &MI) const {
       return Mapping;
   }
 
+  const MachineFunction &MF = *MI.getParent()->getParent();
+  const MachineRegisterInfo &MRI = MF.getRegInfo();
+  const TargetSubtargetInfo &STI = MF.getSubtarget();
+  const TargetRegisterInfo &TRI = *STI.getRegisterInfo();
+
   switch (Opc) {
     // G_{F|S|U}REM are not listed because they are not legal.
     // Arithmetic ops.
@@ -451,6 +454,30 @@ AArch64RegisterBankInfo::getInstrMapping(const MachineInstr &MI) const {
   case TargetOpcode::G_FMUL:
   case TargetOpcode::G_FDIV:
     return getSameKindOfOperandsMapping(MI);
+  case TargetOpcode::COPY: {
+    unsigned DstReg = MI.getOperand(0).getReg();
+    unsigned SrcReg = MI.getOperand(1).getReg();
+    if (TargetRegisterInfo::isPhysicalRegister(DstReg) ||
+        TargetRegisterInfo::isPhysicalRegister(SrcReg)) {
+      const RegisterBank *DstRB = getRegBank(DstReg, MRI, TRI);
+      const RegisterBank *SrcRB = getRegBank(SrcReg, MRI, TRI);
+      if (!DstRB)
+        DstRB = SrcRB;
+      else if (!SrcRB)
+        SrcRB = DstRB;
+      // If both RB are null that means both registers are generic.
+      // We shouldn't be here.
+      assert(DstRB && SrcRB && "Both RegBank were nullptr");
+      unsigned Size = getSizeInBits(DstReg, MRI, TRI);
+      return getInstructionMapping(
+          DefaultMappingID, copyCost(*DstRB, *SrcRB, Size),
+          getCopyMapping(DstRB->getID(), SrcRB->getID(), Size),
+          // We only care about the mapping of the destination.
+          /*NumOperands*/ 1);
+    }
+    // Both registers are generic, use G_BITCAST.
+    LLVM_FALLTHROUGH;
+  }
   case TargetOpcode::G_BITCAST: {
     LLT DstTy = MRI.getType(MI.getOperand(0).getReg());
     LLT SrcTy = MRI.getType(MI.getOperand(1).getReg());
@@ -464,7 +491,8 @@ AArch64RegisterBankInfo::getInstrMapping(const MachineInstr &MI) const {
     return getInstructionMapping(
         DefaultMappingID, copyCost(DstRB, SrcRB, Size),
         getCopyMapping(DstRB.getID(), SrcRB.getID(), Size),
-        /*NumOperands*/ 2);
+        // We only care about the mapping of the destination for COPY.
+        /*NumOperands*/ Opc == TargetOpcode::G_BITCAST ? 2 : 1);
   }
   default:
     break;