Merge pull request #83 from arielb1/bad-arm-2

backport fixes to LLVM 4.0 ARM codegen bugs

Author: alexcrichton

include/llvm/CodeGen/MIRYamlMapping.h

@@ -381,6 +381,7 @@ struct MachineFunction {
   StringRef Name;
   unsigned Alignment = 0;
   bool ExposesReturnsTwice = false;
+  bool NoVRegs;
   // GISel MachineFunctionProperties.
   bool Legalized = false;
   bool RegBankSelected = false;
@@ -405,6 +406,7 @@ template <> struct MappingTraits<MachineFunction> {
     YamlIO.mapRequired("name", MF.Name);
     YamlIO.mapOptional("alignment", MF.Alignment);
     YamlIO.mapOptional("exposesReturnsTwice", MF.ExposesReturnsTwice);
+    YamlIO.mapOptional("noVRegs", MF.NoVRegs);
     YamlIO.mapOptional("legalized", MF.Legalized);
     YamlIO.mapOptional("regBankSelected", MF.RegBankSelected);
     YamlIO.mapOptional("selected", MF.Selected);

lib/CodeGen/IfConversion.cpp

@@ -1353,7 +1353,8 @@ static bool canFallThroughTo(MachineBasicBlock &MBB, MachineBasicBlock &ToMBB) {
       return false;
     PI = I++;
   }
-  return true;
+  // Finally see if the last I is indeed a successor to PI.
+  return PI->isSuccessor(&*I);
 }
 
 /// Invalidate predecessor BB info so it would be re-analyzed to determine if it

lib/CodeGen/MIRParser/MIRParser.cpp

@@ -330,6 +330,8 @@ bool MIRParserImpl::initializeMachineFunction(MachineFunction &MF) {
     MF.setAlignment(YamlMF.Alignment);
   MF.setExposesReturnsTwice(YamlMF.ExposesReturnsTwice);
 
+  if (YamlMF.NoVRegs)
+    MF.getProperties().set(MachineFunctionProperties::Property::NoVRegs);
   if (YamlMF.Legalized)
     MF.getProperties().set(MachineFunctionProperties::Property::Legalized);
   if (YamlMF.RegBankSelected)

lib/CodeGen/MIRPrinter.cpp

@@ -175,6 +175,8 @@ void MIRPrinter::print(const MachineFunction &MF) {
   YamlMF.Alignment = MF.getAlignment();
   YamlMF.ExposesReturnsTwice = MF.exposesReturnsTwice();
 
+  YamlMF.NoVRegs = MF.getProperties().hasProperty(
+      MachineFunctionProperties::Property::NoVRegs);
   YamlMF.Legalized = MF.getProperties().hasProperty(
       MachineFunctionProperties::Property::Legalized);
   YamlMF.RegBankSelected = MF.getProperties().hasProperty(

lib/Target/ARM/ARMConstantIslandPass.cpp

@@ -1715,6 +1715,13 @@ bool ARMConstantIslands::undoLRSpillRestore() {
       MI->eraseFromParent();
       MadeChange = true;
     }
+    if (MI->getOpcode() == ARM::tPUSH &&
+        MI->getOperand(2).getReg() == ARM::LR &&
+        MI->getNumExplicitOperands() == 3) {
+      // Just remove the push.
+      MI->eraseFromParent();
+      MadeChange = true;
+    }
   }
   return MadeChange;
 }
@@ -1984,6 +1991,16 @@ static bool jumpTableFollowsTB(MachineInstr *JTMI, MachineInstr *CPEMI) {
          &*MBB->begin() == CPEMI;
 }
 
+static bool registerDefinedBetween(unsigned Reg,
+                                   MachineBasicBlock::iterator From,
+                                   MachineBasicBlock::iterator To,
+                                   const TargetRegisterInfo *TRI) {
+  for (auto I = From; I != To; ++I)
+    if (I->modifiesRegister(Reg, TRI))
+      return true;
+  return false;
+}
+
 /// optimizeThumb2JumpTables - Use tbb / tbh instructions to generate smaller
 /// jumptables when it's possible.
 bool ARMConstantIslands::optimizeThumb2JumpTables() {
@@ -2033,7 +2050,7 @@ bool ARMConstantIslands::optimizeThumb2JumpTables() {
     unsigned DeadSize = 0;
     bool CanDeleteLEA = false;
     bool BaseRegKill = false;
-    
+
     unsigned IdxReg = ~0U;
     bool IdxRegKill = true;
     if (isThumb2) {
@@ -2061,6 +2078,12 @@ bool ARMConstantIslands::optimizeThumb2JumpTables() {
       IdxReg = Shift->getOperand(2).getReg();
       unsigned ShiftedIdxReg = Shift->getOperand(0).getReg();
 
+      // It's important that IdxReg is live until the actual TBB/TBH. Most of
+      // the range is checked later, but the LEA might still clobber it and not
+      // actually get removed.
+      if (BaseReg == IdxReg && !jumpTableFollowsTB(MI, User.CPEMI))
+        continue;
+
       MachineInstr *Load = User.MI->getNextNode();
       if (Load->getOpcode() != ARM::tLDRr)
         continue;
@@ -2070,6 +2093,16 @@ bool ARMConstantIslands::optimizeThumb2JumpTables() {
         continue;
 
       // If we're in PIC mode, there should be another ADD following.
+      auto *TRI = STI->getRegisterInfo();
+
+      // %base cannot be redefined after the load as it will appear before
+      // TBB/TBH like:
+      //      %base =
+      //      %base =
+      //      tBB %base, %idx
+      if (registerDefinedBetween(BaseReg, Load->getNextNode(), MBB->end(), TRI))
+        continue;
+
       if (isPositionIndependentOrROPI) {
         MachineInstr *Add = Load->getNextNode();
         if (Add->getOpcode() != ARM::tADDrr ||
@@ -2079,22 +2112,27 @@ bool ARMConstantIslands::optimizeThumb2JumpTables() {
           continue;
         if (Add->getOperand(0).getReg() != MI->getOperand(0).getReg())
           continue;
-
+        if (registerDefinedBetween(IdxReg, Add->getNextNode(), MI, TRI))
+          // IdxReg gets redefined in the middle of the sequence.
+          continue;
         Add->eraseFromParent();
         DeadSize += 2;
       } else {
         if (Load->getOperand(0).getReg() != MI->getOperand(0).getReg())
           continue;
+        if (registerDefinedBetween(IdxReg, Load->getNextNode(), MI, TRI))
+          // IdxReg gets redefined in the middle of the sequence.
+          continue;
       }
-      
-      
+
+
       // Now safe to delete the load and lsl. The LEA will be removed later.
       CanDeleteLEA = true;
       Shift->eraseFromParent();
       Load->eraseFromParent();
       DeadSize += 4;
     }
-    
+
     DEBUG(dbgs() << "Shrink JT: " << *MI);
     MachineInstr *CPEMI = User.CPEMI;
     unsigned Opc = ByteOk ? ARM::t2TBB_JT : ARM::t2TBH_JT;