PowerPC: CTR shouldn't fire if a TLS call is in the loop

[majnemer]

Summary

Determining the address of a TLS variable results in a function call in
certain TLS models. This means that a simple ICmpInst might actually
result in invalidating the CTR register.

In such cases, do not attempt to rely on the CTR register for loop
optimization purposes.

This fixes PR22034.

[hfinkel]

LGTM, thanks!

lib/Target/PowerPC/PPCCTRLoops.cpp

207 ↗

(On Diff #17644)

Please add a comment here explaining that determining the address of a TLS variable results in a function call in these TLS models.

llvm/trunk/lib/Target/PowerPC/PPCCTRLoops.cpp

 
 static bool isLargeIntegerTy(bool Is32Bit, Type *Ty) {
   if (IntegerType *ITy = dyn_cast<IntegerType>(Ty))
     return ITy->getBitWidth() > (Is32Bit ? 32U : 64U);
 
   return false;
 }
 
+// Determining the address of a TLS variable results in a function call in
+// certain TLS models.
+static bool memAddrUsesCTR(const PPCTargetMachine *TM,
+                           const llvm::Value *MemAddr) {
+  const auto *GV = dyn_cast<GlobalValue>(MemAddr);
+  if (!GV)
+    return false;
+  if (!GV->isThreadLocal())
+    return false;
+  if (!TM)
+    return true;
+  TLSModel::Model Model = TM->getTLSModel(GV);
+  return Model == TLSModel::GeneralDynamic || Model == TLSModel::LocalDynamic;
+}
+
 bool PPCCTRLoops::mightUseCTR(const Triple &TT, BasicBlock *BB) {
   for (BasicBlock::iterator J = BB->begin(), JE = BB->end();
        J != JE; ++J) {
     if (CallInst *CI = dyn_cast<CallInst>(J)) {
       if (InlineAsm *IA = dyn_cast<InlineAsm>(CI->getCalledValue())) {
         // Inline ASM is okay, unless it clobbers the ctr register.
         InlineAsm::ConstraintInfoVector CIV = IA->ParseConstraints();
         for (unsigned i = 0, ie = CIV.size(); i < ie; ++i) {
#endif
     } else if (SwitchInst *SI = dyn_cast<SwitchInst>(J)) {
       if (!TM)
         return true;
       const TargetLowering *TLI = TM->getSubtargetImpl()->getTargetLowering();
 
       if (SI->getNumCases() + 1 >= (unsigned)TLI->getMinimumJumpTableEntries())
         return true;
     }
+    for (Value *Operand : J->operands())
+      if (memAddrUsesCTR(TM, Operand))
+        return true;
   }
 
   return false;
 }
 
 bool PPCCTRLoops::convertToCTRLoop(Loop *L) {
   bool MadeChange = false;
 

llvm/trunk/test/CodeGen/PowerPC/ctrloops.ll

 target datalayout = "E-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v128:128:128-n32:64"
 target triple = "powerpc64-unknown-freebsd10.0"
-; RUN: llc < %s -march=ppc64 | FileCheck %s
+; RUN: llc < %s -march=ppc64 -relocation-model=pic | FileCheck %s
 
 @a = common global i32 0, align 4
 
 define void @test1(i32 %c) nounwind {
 entry:
   br label %for.body
 
 for.body:                                         ; preds = %for.body, %entry

 for.end:                                          ; preds = %for.body, %entry
   ret void
 ; CHECK: @test3
 ; CHECK: mtctr
 ; CHECK-NOT: addi {[0-9]+}
 ; CHECK-NOT: cmplwi
 ; CHECK: bdnz
 }
+
+@tls_var = external thread_local global i8
+
+define i32 @test4() {
+entry:
+  br label %for.body
+
+for.body:                                         ; preds = %for.body, %entry
+  %phi = phi i32 [ %dec, %for.body ], [ undef, %entry ]
+  %load = ptrtoint i8* @tls_var to i32
+  %dec = add i32 %phi, -1
+  %cmp = icmp sgt i32 %phi, 1
+  br i1 %cmp, label %for.body, label %return
+
+return:                                           ; preds = %for.body
+  ret i32 %load
+; CHECK-LABEL: @test4
+; CHECK-NOT: mtctr
+; CHECK: addi {{[0-9]+}}
+; CHECK: cmpwi
+; CHECK-NOT: bdnz
+; CHECK: bgt
+}