[InstrProfiling] Mark __llvm_profile_instrument_target last parameter as i32 zeroext or signext if appropriate.

[• koriakin]

Summary

On some architectures (s390x, ppc64, sparc64, mips), C-level int is passed
as i32 signext instead of plain i32. Likewise, unsigned int may be passed
as i32, i32 signext, or i32 zeroext depending on the platform. Mark
__llvm_profile_instrument_target properly (its last parameter is unsigned
int).

This (together with the clang change) makes compiler-rt profile testsuite pass
on s390x.

[vsk]

@davidxl would be a good reviewer for this.

include/llvm/Analysis/TargetLibraryInfo.h

178 ↗

(On Diff #61913)

Could you split this up into two setters, to better match the getters below?

include/llvm/Transforms/InstrProfiling.h

35 ↗

(On Diff #61913)

It appears that TLI is assumed to be non-null. Can you make this a reference?

lib/Transforms/Instrumentation/InstrProfiling.cpp

177 ↗

(On Diff #61913)

Unrelated to this patch, but: it seems weird that this is called every time we do a lowerValueProfileInst(). Can we just do it once per run?

[vsk]

Also, this is missing a test.

[• koriakin]

In D21736#467527, @vsk wrote:

Also, this is missing a test.

True. I'm going to add a clang test in a separate patch (this way it will cover D21737 as well). Should I make an llvm test as well?

[vsk]

In D21736#467529, @koriakin wrote:

In D21736#467527, @vsk wrote:

Also, this is missing a test.

True. I'm going to add a clang test in a separate patch (this way it will cover D21737 as well). Should I make an llvm test as well?

Yes, that would be helpful.

[davidxl]

The TLI change is independent of InstProfile changes. Please split the InstProf change into a separate patch.

[• koriakin]

Split off the TLI part, changed pointer to reference.

[• koriakin]

Added a test.

[vsk]

LGTM.

[davidxl]

One nit. Can you save TLI pointer in InstProfiling instead of passing it around?

[• koriakin]

In D21736#467986, @davidxl wrote:

One nit. Can you save TLI pointer in InstProfiling instead of passing it around?

Save where? I thought run is not supposed to modify the Pass object, and I don't have TLI before that.

[• koriakin]

Saved TLI as a field of InstProfiling.

[davidxl]

lgtm

[• koriakin]

Changed to use TargetTransformInfo, per comment on D21739. D21737 is now no longer necessary.

I'm not sure if keeping TTI on the pass object is a good idea now...

[• koriakin]

Added another case for mips, which uses i32 signext.

[• koriakin]

Split off the tests (they'll come together with target-specific implementations for D21739).

[• koriakin]

Updated for new revision of D21739 - now aborts if knownTarget() is false.

[• koriakin]

Moved the functions to TLI, as per ML: http://lists.llvm.org/pipermail/llvm-dev/2016-September/104641.html.

[• koriakin]

It seems I have accidentally swapped the diffs of this revision and D21739. Here comes the correct diff.

[• koriakin]

Readded missing test.

[• koriakin]

Changed to composite getExtAttrForI32Param/Return functions as for D21739.

[hfinkel]

LGTM too.

llvm/trunk/include/llvm/Transforms/InstrProfiling.h

 
 #include "llvm/IR/IntrinsicInst.h"
 #include "llvm/IR/PassManager.h"
 #include "llvm/ProfileData/InstrProf.h"
 #include "llvm/Transforms/Instrumentation.h"
 
 namespace llvm {
 
+class TargetLibraryInfo;
+
 /// Instrumentation based profiling lowering pass. This pass lowers
 /// the profile instrumented code generated by FE or the IR based
 /// instrumentation pass.
 class InstrProfiling : public PassInfoMixin<InstrProfiling> {
 public:
   InstrProfiling() {}
   InstrProfiling(const InstrProfOptions &Options) : Options(Options) {}
 
   PreservedAnalyses run(Module &M, ModuleAnalysisManager &AM);
-  bool run(Module &M);
+  bool run(Module &M, const TargetLibraryInfo &TLI);
 
 private:
   InstrProfOptions Options;
   Module *M;
+  const TargetLibraryInfo *TLI;
   struct PerFunctionProfileData {
     uint32_t NumValueSites[IPVK_Last + 1];
     GlobalVariable *RegionCounters;
     GlobalVariable *DataVar;
     PerFunctionProfileData() : RegionCounters(nullptr), DataVar(nullptr) {
       memset(NumValueSites, 0, sizeof(uint32_t) * (IPVK_Last + 1));
     }
   };

llvm/trunk/lib/Transforms/Instrumentation/InstrProfiling.cpp

 // This pass lowers instrprof_* intrinsics emitted by a frontend for profiling.
 // It also builds the data structures and initialization code needed for
 // updating execution counts and emitting the profile at runtime.
 //
 //===----------------------------------------------------------------------===//
 
 #include "llvm/Transforms/InstrProfiling.h"
 #include "llvm/ADT/Triple.h"
+#include "llvm/Analysis/TargetLibraryInfo.h"
 #include "llvm/IR/IRBuilder.h"
 #include "llvm/IR/IntrinsicInst.h"
 #include "llvm/IR/Module.h"
 #include "llvm/ProfileData/InstrProf.h"
 #include "llvm/Transforms/Utils/ModuleUtils.h"
 
 using namespace llvm;
 
public:
   static char ID;
   InstrProfilingLegacyPass() : ModulePass(ID), InstrProf() {}
   InstrProfilingLegacyPass(const InstrProfOptions &Options)
       : ModulePass(ID), InstrProf(Options) {}
   StringRef getPassName() const override {
     return "Frontend instrumentation-based coverage lowering";
   }
 
-  bool runOnModule(Module &M) override { return InstrProf.run(M); }
+  bool runOnModule(Module &M) override {
+    return InstrProf.run(M, getAnalysis<TargetLibraryInfoWrapperPass>().getTLI());
+  }
 
   void getAnalysisUsage(AnalysisUsage &AU) const override {
     AU.setPreservesCFG();
+    AU.addRequired<TargetLibraryInfoWrapperPass>();
   }
 };
 
 } // anonymous namespace
 
 PreservedAnalyses InstrProfiling::run(Module &M, ModuleAnalysisManager &AM) {
-  if (!run(M))
+  auto &TLI = AM.getResult<TargetLibraryAnalysis>(M);
+  if (!run(M, TLI))
     return PreservedAnalyses::all();
 
   return PreservedAnalyses::none();
 }
 
 char InstrProfilingLegacyPass::ID = 0;
-INITIALIZE_PASS(InstrProfilingLegacyPass, "instrprof",
-                "Frontend instrumentation-based coverage lowering.", false,
-                false)
+INITIALIZE_PASS_BEGIN(
+    InstrProfilingLegacyPass, "instrprof",
+    "Frontend instrumentation-based coverage lowering.", false, false)
+INITIALIZE_PASS_DEPENDENCY(TargetLibraryInfoWrapperPass)
+INITIALIZE_PASS_END(
+    InstrProfilingLegacyPass, "instrprof",
+    "Frontend instrumentation-based coverage lowering.", false, false)
 
 ModulePass *
 llvm::createInstrProfilingLegacyPass(const InstrProfOptions &Options) {
   return new InstrProfilingLegacyPass(Options);
 }
 
 bool InstrProfiling::isMachO() const {
   return Triple(M->getTargetTriple()).isOSBinFormatMachO();

 
 static InstrProfIncrementInst *castToIncrementInst(Instruction *Instr) {
   InstrProfIncrementInst *Inc = dyn_cast<InstrProfIncrementInstStep>(Instr);
   if (Inc)
     return Inc;
   return dyn_cast<InstrProfIncrementInst>(Instr);
 }
 
-bool InstrProfiling::run(Module &M) {
+bool InstrProfiling::run(Module &M, const TargetLibraryInfo &TLI) {
   bool MadeChange = false;
 
   this->M = &M;
+  this->TLI = &TLI;
   NamesVar = nullptr;
   NamesSize = 0;
   ProfileDataMap.clear();
   UsedVars.clear();
 
   // We did not know how many value sites there would be inside
   // the instrumented function. This is counting the number of instrumented
   // target value sites to enter it as field in the profile data variable.
bool InstrProfiling::run(Module &M, const TargetLibraryInfo &TLI) {
   emitNameData();
   emitRegistration();
   emitRuntimeHook();
   emitUses();
   emitInitialization();
   return true;
 }
 
-static Constant *getOrInsertValueProfilingCall(Module &M) {
+static Constant *getOrInsertValueProfilingCall(Module &M,
+                                               const TargetLibraryInfo &TLI) {
   LLVMContext &Ctx = M.getContext();
   auto *ReturnTy = Type::getVoidTy(M.getContext());
   Type *ParamTypes[] = {
 #define VALUE_PROF_FUNC_PARAM(ParamType, ParamName, ParamLLVMType) ParamLLVMType
 #include "llvm/ProfileData/InstrProfData.inc"
   };
   auto *ValueProfilingCallTy =
       FunctionType::get(ReturnTy, makeArrayRef(ParamTypes), false);
-  return M.getOrInsertFunction(getInstrProfValueProfFuncName(),
+  Constant *Res = M.getOrInsertFunction(getInstrProfValueProfFuncName(),
                                         ValueProfilingCallTy);
+  if (Function *FunRes = dyn_cast<Function>(Res)) {
+    if (auto AK = TLI.getExtAttrForI32Param(false))
+      FunRes->addAttribute(3, AK);
+  }
+  return Res;
 }
 
 void InstrProfiling::computeNumValueSiteCounts(InstrProfValueProfileInst *Ind) {
 
   GlobalVariable *Name = Ind->getName();
   uint64_t ValueKind = Ind->getValueKind()->getZExtValue();
   uint64_t Index = Ind->getIndex()->getZExtValue();
   auto It = ProfileDataMap.find(Name);
void InstrProfiling::lowerValueProfileInst(InstrProfValueProfileInst *Ind) {
   uint64_t Index = Ind->getIndex()->getZExtValue();
   for (uint32_t Kind = IPVK_First; Kind < ValueKind; ++Kind)
     Index += It->second.NumValueSites[Kind];
 
   IRBuilder<> Builder(Ind);
   Value *Args[3] = {Ind->getTargetValue(),
                     Builder.CreateBitCast(DataVar, Builder.getInt8PtrTy()),
                     Builder.getInt32(Index)};
-  Ind->replaceAllUsesWith(
-      Builder.CreateCall(getOrInsertValueProfilingCall(*M), Args));
+  CallInst *Call = Builder.CreateCall(getOrInsertValueProfilingCall(*M, *TLI),
+                                      Args);
+  if (auto AK = TLI->getExtAttrForI32Param(false))
+    Call->addAttribute(3, AK);
+  Ind->replaceAllUsesWith(Call);
   Ind->eraseFromParent();
 }
 
 void InstrProfiling::lowerIncrement(InstrProfIncrementInst *Inc) {
   GlobalVariable *Counters = getOrCreateRegionCounters(Inc);
 
   IRBuilder<> Builder(Inc);
   uint64_t Index = Inc->getIndex()->getZExtValue();

llvm/trunk/test/Instrumentation/InstrProfiling/icall.ll

 ;; Check that static counters are allocated for value profiler
 
 ; RUN: opt < %s -mtriple=x86_64-apple-macosx10.10.0 -vp-static-alloc=true -instrprof -S | FileCheck %s --check-prefix=STATIC
 ; RUN: opt < %s -mtriple=x86_64-unknown-linux -instrprof -vp-static-alloc=true -S | FileCheck %s --check-prefix=STATIC
+; RUN: opt < %s -mtriple=powerpc-unknown-linux -instrprof -vp-static-alloc=true -S | FileCheck %s --check-prefix=STATIC
+; RUN: opt < %s -mtriple=sparc-unknown-linux -instrprof -vp-static-alloc=true -S | FileCheck %s --check-prefix=STATIC
+; RUN: opt < %s -mtriple=s390x-unknown-linux -instrprof -vp-static-alloc=true -S | FileCheck %s --check-prefix=STATIC-EXT
+; RUN: opt < %s -mtriple=powerpc64-unknown-linux -instrprof -vp-static-alloc=true -S | FileCheck %s --check-prefix=STATIC-EXT
+; RUN: opt < %s -mtriple=sparc64-unknown-linux -instrprof -vp-static-alloc=true -S | FileCheck %s --check-prefix=STATIC-EXT
+; RUN: opt < %s -mtriple=mips-unknown-linux -instrprof -vp-static-alloc=true -S | FileCheck %s --check-prefix=STATIC-SEXT
+; RUN: opt < %s -mtriple=mips64-unknown-linux -instrprof -vp-static-alloc=true -S | FileCheck %s --check-prefix=STATIC-SEXT
 ; RUN: opt < %s -mtriple=x86_64-apple-macosx10.10.0 -vp-static-alloc=false -instrprof -S | FileCheck %s --check-prefix=DYN
 ; RUN: opt < %s -mtriple=x86_64-unknown-linux -instrprof -vp-static-alloc=false -S | FileCheck %s --check-prefix=DYN
 
 
 @__profn_foo = private constant [3 x i8] c"foo"
 
 define i32 @foo(i32 ()* ) {
   call void @llvm.instrprof.increment(i8* getelementptr inbounds ([3 x i8], [3 x i8]* @__profn_foo, i32 0, i32 0), i64 12884901887, i32 1, i32 0)

 
 attributes #0 = { nounwind }
 
 ; STATIC: @__profvp_foo
 ; STATIC: @__llvm_prf_vnodes
 
 ; DYN-NOT: @__profvp_foo
 ; DYN-NOT: @__llvm_prf_vnodes
+
+; STATIC: call void @__llvm_profile_instrument_target(i64 %3, i8* bitcast ({ i64, i64, i64*, i8*, i8*, i32, [1 x i16] }* @__profd_foo to i8*), i32 0)
+; STATIC-EXT: call void @__llvm_profile_instrument_target(i64 %3, i8* bitcast ({ i64, i64, i64*, i8*, i8*, i32, [1 x i16] }* @__profd_foo to i8*), i32 zeroext 0)
+; STATIC-SEXT: call void @__llvm_profile_instrument_target(i64 %3, i8* bitcast ({ i64, i64, i64*, i8*, i8*, i32, [1 x i16] }* @__profd_foo to i8*), i32 signext 0)
+
+; STATIC: declare void @__llvm_profile_instrument_target(i64, i8*, i32)
+; STATIC-EXT: declare void @__llvm_profile_instrument_target(i64, i8*, i32 zeroext)
+; STATIC-SEXT: declare void @__llvm_profile_instrument_target(i64, i8*, i32 signext)