[InstrProfiling] Don't generate __llvm_profile_runtime_user

[phosek]

Summary

This is no longer needed, we can add __llvm_profile_runtime directly
to llvm.compiler.used or llvm.used to achieve the same effect.

[vsk]

Thanks, lgtm.

[hans]

This broke the check-profile tests on Mac, see e.g. http://green.lab.llvm.org/green/job/clang-stage1-RA/19301/
There are no Mac bots on the main buildbot waterfall, but I thought Apple were watching Green Dragon?

Reverted in f50aef745c3ba981f3d0bf118b809d0c3880a490 to unbreak the builds.

[phosek]

In D98325#2621884, @hans wrote:

This broke the check-profile tests on Mac, see e.g. http://green.lab.llvm.org/green/job/clang-stage1-RA/19301/
There are no Mac bots on the main buildbot waterfall, but I thought Apple were watching Green Dragon?

I'm getting ERR_CONNECTION_REFUSED when trying to open that URL, how I can see the error?

Reverted in f50aef745c3ba981f3d0bf118b809d0c3880a490 to unbreak the builds.

[phosek]

In D98325#2623072, @phosek wrote:

In D98325#2621884, @hans wrote:

This broke the check-profile tests on Mac, see e.g. http://green.lab.llvm.org/green/job/clang-stage1-RA/19301/
There are no Mac bots on the main buildbot waterfall, but I thought Apple were watching Green Dragon?

I'm getting ERR_CONNECTION_REFUSED when trying to open that URL, how I can see the error?

Turned out to be an issue with corporate network since Green Dragon doesn't use HTTPS, I've managed to open the page on a personal machine.

[phosek]

@vsk I managed to reproduce this on Mac and it seems like __llvm_profile_runtime_user is needed in Mach-O. With this change, I see an undefined reference to ___llvm_profile_runtime in the object file, but that doesn't cause the profile runtime to be linked and that symbol is stripped from the binary. I'm not familiar with Mach-O linking semantics so I don't know if this is working as intended or not.

[MaskRay]

On Linux, the clang driver passes -u__llvm_profile_runtime to the linker: Linux::addProfileRTLibs. Can Mach-O do something similar? The clang driver already does special things with ___llvm_profile_filename and ___llvm_profile_raw_version.

[phosek]

In D98325#2623886, @MaskRay wrote:

On Linux, the clang driver passes -u__llvm_profile_runtime to the linker: Linux::addProfileRTLibs. Can Mach-O do something similar? The clang driver already does special things with ___llvm_profile_filename and ___llvm_profile_raw_version.

We're actually trying to move away from that model and define __llvm_profile_runtime as needed in the backend, see D98061.

llvm/include/llvm/ProfileData/InstrProf.h

 
 /// Return the name of the hook variable defined in profile runtime library.
 /// A reference to the variable causes the linker to link in the runtime
 /// initialization module (which defines the hook variable).
 inline StringRef getInstrProfRuntimeHookVarName() {
   return INSTR_PROF_QUOTE(INSTR_PROF_PROFILE_RUNTIME_VAR);
 }
 
-/// Return the name of the compiler generated function that references the
-/// runtime hook variable. The function is a weak global.
-inline StringRef getInstrProfRuntimeHookVarUseFuncName() {
-  return "__llvm_profile_runtime_user";
-}
-
 inline StringRef getInstrProfCounterBiasVarName() {
   return "__llvm_profile_counter_bias";
 }
 
 /// Return the marker used to separate PGO names during serialization.
 inline StringRef getInstrProfNameSeparator() { return "\01"; }
 
 /// Return the modified name for function \c F suitable to be

llvm/lib/Transforms/Instrumentation/InstrProfiling.cpp

if (M->getGlobalVariable(getInstrProfRuntimeHookVarName()))
     return false;
 
   // Declare an external variable that will pull in the runtime initialization.
   auto *Int32Ty = Type::getInt32Ty(M->getContext());
   auto *Var =
       new GlobalVariable(*M, Int32Ty, false, GlobalValue::ExternalLinkage,
                          nullptr, getInstrProfRuntimeHookVarName());
 
-  // Make a function that uses it.
-  auto *User = Function::Create(FunctionType::get(Int32Ty, false),
-                                GlobalValue::LinkOnceODRLinkage,
-                                getInstrProfRuntimeHookVarUseFuncName(), M);
-  User->addFnAttr(Attribute::NoInline);
-  if (Options.NoRedZone)
-    User->addFnAttr(Attribute::NoRedZone);
-  User->setVisibility(GlobalValue::HiddenVisibility);
-  if (TT.supportsCOMDAT())
-    User->setComdat(M->getOrInsertComdat(User->getName()));
-
-  IRBuilder<> IRB(BasicBlock::Create(M->getContext(), "", User));
-  auto *Load = IRB.CreateLoad(Int32Ty, Var);
-  IRB.CreateRet(Load);
-
   // Mark the user variable as used so that it isn't stripped out.
-  CompilerUsedVars.push_back(User);
+  CompilerUsedVars.push_back(Var);
   return true;
 }
 
 void InstrProfiling::emitUses() {
   // The metadata sections are parallel arrays. Optimizers (e.g.
   // GlobalOpt/ConstantMerge) may not discard associated sections as a unit, so
   // we conservatively retain all unconditionally in the compiler.
   //

llvm/test/Instrumentation/InstrProfiling/linkage.ll

 ; COFF: @__profc_foo_extern = linkonce_odr hidden global {{.*}}section ".lprfc$M", comdat, align 8
 ; COFF: @__profd_foo_extern = linkonce_odr hidden global {{.*}}section ".lprfd$M", comdat, align 8
 define available_externally void @foo_extern() {
   call void @llvm.instrprof.increment(i8* getelementptr inbounds ([10 x i8], [10 x i8]* @__profn_foo_extern, i32 0, i32 0), i64 0, i32 1, i32 0)
   ret void
 }
 
 declare void @llvm.instrprof.increment(i8*, i64, i32, i32)
-
-; MACHO: define linkonce_odr hidden i32 @__llvm_profile_runtime_user() {{.*}} {
-; MACHO:   %[[REG:.*]] = load i32, i32* @__llvm_profile_runtime
-; MACHO:   ret i32 %[[REG]]
-; MACHO: }
-; COFF: define linkonce_odr hidden i32 @__llvm_profile_runtime_user() {{.*}} comdat {
-; ELF-NOT: define linkonce_odr hidden i32 @__llvm_profile_runtime_user() {{.*}} {
-; ELF-NOT:   %[[REG:.*]] = load i32, i32* @__llvm_profile_runtime

llvm/test/Instrumentation/InstrProfiling/profiling.ll

define void @baz() {
   call void @llvm.instrprof.increment(i8* getelementptr inbounds ([3 x i8], [3 x i8]* @__profn_baz, i32 0, i32 0), i64 0, i32 3, i32 0)
   call void @llvm.instrprof.increment(i8* getelementptr inbounds ([3 x i8], [3 x i8]* @__profn_baz, i32 0, i32 0), i64 0, i32 3, i32 1)
   call void @llvm.instrprof.increment(i8* getelementptr inbounds ([3 x i8], [3 x i8]* @__profn_baz, i32 0, i32 0), i64 0, i32 3, i32 2)
   ret void
 }
 
 declare void @llvm.instrprof.increment(i8*, i64, i32, i32)
 
-; ELF:   @llvm.compiler.used = appending global {{.*}} @__llvm_profile_runtime_user {{.*}} @__profd_foo {{.*}} @__profd_bar {{.*}} @__profd_baz
-; MACHO: @llvm.used = appending global {{.*}} @__llvm_profile_runtime_user {{.*}} @__profd_foo {{.*}} @__profd_bar {{.*}} @__profd_baz
-; WIN:   @llvm.used = appending global {{.*}} @__llvm_profile_runtime_user {{.*}} @__profd_foo {{.*}} @__profd_bar {{.*}} @__profd_baz
+; ELF:   @llvm.compiler.used = appending global {{.*}} @__llvm_profile_runtime {{.*}} @__profd_foo {{.*}} @__profd_bar {{.*}} @__profd_baz
+; MACHO: @llvm.used = appending global {{.*}} @__llvm_profile_runtime {{.*}} @__profd_foo {{.*}} @__profd_bar {{.*}} @__profd_baz
+; WIN:   @llvm.used = appending global {{.*}} @__llvm_profile_runtime {{.*}} @__profd_foo {{.*}} @__profd_bar {{.*}} @__profd_baz
 
 ; ELF_GENERIC:      define internal void @__llvm_profile_register_functions() unnamed_addr {
+; ELF_GENERIC-NEXT:   call void @__llvm_profile_register_function(i8* bitcast (i32* @__llvm_profile_runtime to i8*))
 ; ELF_GENERIC-NEXT:   call void @__llvm_profile_register_function(i8* bitcast ({ i64, i64, i64*, i8*, i8*, i32, [2 x i16] }* @__profd_foo to i8*))
 ; ELF_GENERIC-NEXT:   call void @__llvm_profile_register_function(i8* bitcast ({ i64, i64, i64*, i8*, i8*, i32, [2 x i16] }* @__profd_bar to i8*))
 ; ELF_GENERIC-NEXT:   call void @__llvm_profile_register_function(i8* bitcast ({ i64, i64, i64*, i8*, i8*, i32, [2 x i16] }* @__profd_baz to i8*))
 ; ELF_GENERIC-NEXT:   call void @__llvm_profile_register_names_function(i8* getelementptr inbounds {{.*}} @__llvm_prf_nm
 ; ELF_GENERIC-NEXT:   ret void
 ; ELF_GENERIC-NEXT: }
 
 ; ELF_LINUX-NOT: @__llvm_profile_register_functions()