Add a RequireStructuredCFG Field to TargetMachine.

llvm-svn: 196634

5 files changed, 16 insertions, 2 deletions

diff --git a/llvm/lib/CodeGen/BranchFolding.cpp b/llvm/lib/CodeGen/BranchFolding.cpp
index 9cd4208d646..a4a3712de8b 100644
--- a/llvm/lib/CodeGen/BranchFolding.cpp
+++ b/llvm/lib/CodeGen/BranchFolding.cpp
@@ -83,7 +83,11 @@ INITIALIZE_PASS(BranchFolderPass, "branch-folder",
 
 bool BranchFolderPass::runOnMachineFunction(MachineFunction &MF) {
   TargetPassConfig *PassConfig = &getAnalysis<TargetPassConfig>();
-  BranchFolder Folder(PassConfig->getEnableTailMerge(), /*CommonHoist=*/true);
+  // TailMerge can create jump into if branches that make CFG irreducible for
+  // HW that requires structurized CFG.
+  bool EnableTailMerge = !MF.getTarget().requiresStructuredCFG() &&
+      PassConfig->getEnableTailMerge();
+  BranchFolder Folder(EnableTailMerge, /*CommonHoist=*/true);
   return Folder.OptimizeFunction(MF,
                                  MF.getTarget().getInstrInfo(),
                                  MF.getTarget().getRegisterInfo(),
diff --git a/llvm/lib/CodeGen/MachineBasicBlock.cpp b/llvm/lib/CodeGen/MachineBasicBlock.cpp
index ca71e3bf806..3d36dc18e38 100644
--- a/llvm/lib/CodeGen/MachineBasicBlock.cpp
+++ b/llvm/lib/CodeGen/MachineBasicBlock.cpp
@@ -677,6 +677,11 @@ MachineBasicBlock::SplitCriticalEdge(MachineBasicBlock *Succ, Pass *P) {
   MachineFunction *MF = getParent();
   DebugLoc dl;  // FIXME: this is nowhere
 
+  // Performance might be harmed on HW that implements branching using exec mask
+  // where both sides of the branches are always executed.
+  if (MF->getTarget().requiresStructuredCFG())
+    return NULL;
+
   // We may need to update this's terminator, but we can't do that if
   // AnalyzeBranch fails. If this uses a jump table, we won't touch it.
   const TargetInstrInfo *TII = MF->getTarget().getInstrInfo();
diff --git a/llvm/lib/CodeGen/Passes.cpp b/llvm/lib/CodeGen/Passes.cpp
index f4ffd03ec3e..db7021372bc 100644
--- a/llvm/lib/CodeGen/Passes.cpp
+++ b/llvm/lib/CodeGen/Passes.cpp
@@ -725,7 +725,10 @@ void TargetPassConfig::addMachineLateOptimization() {
     printAndVerify("After BranchFolding");
 
   // Tail duplication.
-  if (addPass(&TailDuplicateID))
+  // Note that duplicating tail just increases code size and degrades
+  // performance for targets that require Structured Control Flow.
+  // In addition it can also make CFG irreducible. Thus we disable it.
+  if (!TM->requiresStructuredCFG() && addPass(&TailDuplicateID))
     printAndVerify("After TailDuplicate");
 
   // Copy propagation.
diff --git a/llvm/lib/Target/R600/AMDGPUTargetMachine.cpp b/llvm/lib/Target/R600/AMDGPUTargetMachine.cpp
index bc4f5d720ae..b42728550aa 100644
--- a/llvm/lib/Target/R600/AMDGPUTargetMachine.cpp
+++ b/llvm/lib/Target/R600/AMDGPUTargetMachine.cpp
@@ -72,6 +72,7 @@ AMDGPUTargetMachine::AMDGPUTargetMachine(const Target &T, StringRef TT,
     InstrInfo.reset(new SIInstrInfo(*this));
     TLInfo.reset(new SITargetLowering(*this));
   }
+  setRequiresStructuredCFG(true);
   initAsmInfo();
 }
 
diff --git a/llvm/lib/Target/TargetMachine.cpp b/llvm/lib/Target/TargetMachine.cpp
index 9ba78bce7f9..a2350352071 100644
--- a/llvm/lib/Target/TargetMachine.cpp
+++ b/llvm/lib/Target/TargetMachine.cpp
@@ -55,6 +55,7 @@ TargetMachine::TargetMachine(const Target &T,
     MCUseLoc(true),
     MCUseCFI(true),
     MCUseDwarfDirectory(false),
+    RequireStructuredCFG(false),
     Options(Options) {
 }