Replace Clang's createRuntimeFunction with the definitions in OMPKinds.def

[jhuber6]

Summary

This patch changes Clang to generate runtime functions from the information inside OMPKinds.def and to use the enumeration values in OMPConstants. Testing currently fails because of incomplete type support for 32 and 64 bit targets.

[jhuber6]

Changed CGOpenMPRuntime to use the getOrCreateRuntimeFunction method in OpenMPIRBuilder. Also changed the type signature for kmpc_critical_with_hint to Int32Ty.

[jdoerfert]

We pass all clang tests, right?

[jdoerfert]

llvm/include/llvm/Frontend/OpenMP/OMPIRBuilder.h

203

Nit: M is the commonly used name for a module ;)

llvm/include/llvm/Frontend/OpenMP/OMPKinds.def

370

What about the comments with a ??

llvm/lib/Frontend/OpenMP/OMPIRBuilder.cpp

83

M

103

Please use FnID for the comparison.

[jhuber6]

Here's the tests it fails, there might be a few that are wrong for reasons beyond the size_t stuff but it's hard to tell until that issue is resolved. The cuda test is just because I have CUDA set up incorrectly on my machine, libomp used to pass until I changed it to us getOrCreateRuntimeFunction I think.

Clang :: Driver/cuda-simple.cu
Clang :: OpenMP/barrier_codegen.cpp
Clang :: OpenMP/distribute_parallel_for_reduction_codegen.cpp
Clang :: OpenMP/distribute_parallel_for_reduction_task_codegen.cpp
Clang :: OpenMP/for_reduction_task_codegen.cpp
Clang :: OpenMP/nvptx_target_parallel_reduction_codegen.cpp
Clang :: OpenMP/nvptx_teams_reduction_codegen.cpp
Clang :: OpenMP/openmp_win_codegen.cpp
Clang :: OpenMP/parallel_firstprivate_codegen.cpp
Clang :: OpenMP/parallel_for_reduction_task_codegen.cpp
Clang :: OpenMP/parallel_master_reduction_task_codegen.cpp
Clang :: OpenMP/parallel_reduction_task_codegen.cpp
Clang :: OpenMP/parallel_sections_reduction_task_codegen.cpp
Clang :: OpenMP/sections_reduction_task_codegen.cpp
Clang :: OpenMP/target_depend_codegen.cpp
Clang :: OpenMP/target_enter_data_depend_codegen.cpp
Clang :: OpenMP/target_exit_data_depend_codegen.cpp
Clang :: OpenMP/target_parallel_depend_codegen.cpp
Clang :: OpenMP/target_parallel_for_depend_codegen.cpp
Clang :: OpenMP/target_parallel_for_reduction_task_codegen.cpp
Clang :: OpenMP/target_parallel_for_simd_depend_codegen.cpp
Clang :: OpenMP/target_parallel_reduction_task_codegen.cpp
Clang :: OpenMP/target_simd_depend_codegen.cpp
Clang :: OpenMP/target_teams_depend_codegen.cpp
Clang :: OpenMP/target_teams_distribute_depend_codegen.cpp
Clang :: OpenMP/target_teams_distribute_parallel_for_depend_codegen.cpp
Clang :: OpenMP/target_teams_distribute_parallel_for_reduction_codegen.cpp
Clang :: OpenMP/target_teams_distribute_parallel_for_reduction_task_codegen.cpp
Clang :: OpenMP/target_teams_distribute_parallel_for_simd_depend_codegen.cpp
Clang :: OpenMP/target_teams_distribute_parallel_for_simd_reduction_codegen.cpp
Clang :: OpenMP/target_teams_distribute_reduction_codegen.cpp
Clang :: OpenMP/target_teams_distribute_simd_depend_codegen.cpp
Clang :: OpenMP/target_teams_distribute_simd_reduction_codegen.cpp
Clang :: OpenMP/target_teams_map_codegen.cpp
Clang :: OpenMP/target_update_depend_codegen.cpp
Clang :: OpenMP/teams_distribute_parallel_for_reduction_codegen.cpp
Clang :: OpenMP/teams_distribute_parallel_for_reduction_task_codegen.cpp
Clang :: OpenMP/teams_distribute_parallel_for_simd_reduction_codegen.cpp
Clang :: OpenMP/teams_distribute_reduction_codegen.cpp
Clang :: OpenMP/teams_distribute_simd_reduction_codegen.cpp
libomp :: tasking/kmp_taskloop.c
libomp :: worksharing/for/kmp_doacross_check.c

llvm/include/llvm/Frontend/OpenMP/OMPIRBuilder.h

203

The OpenMPIRBuilder struct has a Module named M so I wasn't sure if I should be shadowing it.

llvm/include/llvm/Frontend/OpenMP/OMPKinds.def

370

That's just where I was unsure because I saw some conflicting information, like this one with Int32 in the runtime and uintptr_t in Clang.

llvm/lib/Frontend/OpenMP/OMPIRBuilder.cpp

103

Don't know why I didn't think of doing that, will do.

[jhuber6]

llvm/include/llvm/Frontend/OpenMP/OMPKinds.def

483

For this one there's conflicting information on the return value. Clang expects this in its checks
// CK1: {{%.+}} = call i32 @__kmpc_push_num_teams({{.+}}, {{.+}}, i32 [[TE_VAL]], i32 [[TH_VAL]])

But when I look at the runtime I find this
void __kmpc_push_num_teams(ident_t *loc, kmp_int32 global_tid,

kmp_int32 num_teams, kmp_int32 num_threads) {

[jdoerfert]

libomp used to pass until I changed it to us getOrCreateRuntimeFunction I think.

Those tests are flaky, I think. Just run them a few times. I don't expect this to influence them at all.

llvm/include/llvm/Frontend/OpenMP/OMPIRBuilder.h

203

Go with M, it's static :)

llvm/include/llvm/Frontend/OpenMP/OMPKinds.def

483

Make it void here please, will need to update the tests:
sed -i -e 's:i32 @__kmpc_push_num_teams:void @__kmpc_push_num_teams:' clang/test/OpenMP/*.*

llvm/lib/Frontend/OpenMP/OMPIRBuilder.cpp

103

And if the callback is the same for both, merge the conditionals.

[jhuber6]

Making suggested changes. I added a quick implementation of size_t by getting the target triple data from the Module we have access to in the initializeTypes function, I'm not sure if this works in all situations. After changing the implementation some of the tests are invalidated and will need to be changed. push_num_teams was changed to void so now the tests that assign a register to the return value will fail. Additionally this applies the function attributes so some of the tests will check for not having function attributes. There are still tests that cause it to segfault, I think it might be related to OpenMP 5.0 because it's listed in the arguments. I just wanted to get this up for some feedback before rewriting the tests. Here's the tests that failed, the ones involving teams were mostly related to changing push_num_teams to void, the others are seg faults.

Clang :: Driver/cuda-simple.cu
Clang :: OpenMP/barrier_codegen.cpp
Clang :: OpenMP/distribute_parallel_for_reduction_task_codegen.cpp                                                                                
Clang :: OpenMP/for_reduction_task_codegen.cpp
Clang :: OpenMP/openmp_win_codegen.cpp
Clang :: OpenMP/parallel_for_reduction_task_codegen.cpp
Clang :: OpenMP/parallel_master_reduction_task_codegen.cpp
Clang :: OpenMP/parallel_reduction_task_codegen.cpp
Clang :: OpenMP/parallel_sections_reduction_task_codegen.cpp
Clang :: OpenMP/sections_reduction_task_codegen.cpp
Clang :: OpenMP/target_parallel_for_reduction_task_codegen.cpp
Clang :: OpenMP/target_parallel_reduction_task_codegen.cpp
Clang :: OpenMP/target_teams_distribute_parallel_for_reduction_task_codegen.cpp
Clang :: OpenMP/teams_codegen.cpp
Clang :: OpenMP/teams_distribute_parallel_for_reduction_task_codegen.cpp

[jhuber6]

llvm/include/llvm/Frontend/OpenMP/OMPIRBuilder.h

203

Forgot to write the file when I changed it here, not work updating the diff for a single character. I'll put it in the next one.

llvm/include/llvm/Frontend/OpenMP/OMPKinds.def

245

I'm just directly getting the SizeTy from the Module, I'm not sure if this is a permanent solution.

[jdoerfert]

llvm/include/llvm/Frontend/OpenMP/OMPKinds.def

245

I actually think this is fine. Less burden on the frontends also. @fghanim looks good?

[fghanim]

llvm/include/llvm/Frontend/OpenMP/OMPKinds.def

245

Agreed. But to avoid conflict with patch D79675, let this change go in a separate patch after that. This way, we can also do some clean ups related to this, and add Int8PtrPtr, Int8PtrPtrPtr, and any other typing issues, without delaying patch D79675 any further - I am done with that patch, and made my final changes a couple of days ago.

Nit: Also, I saw below that some formatting was done. Wouldn't it be better if formatting/comments related to changes in D79739, are handled as part of that patch?

[jhuber6]

llvm/include/llvm/Frontend/OpenMP/OMPKinds.def

245

Doesn't this patch already support Int8PtrPtrPtr? It's the same thing as a VoidPtrPtrPtr which I have defined above here.

[fghanim]

llvm/include/llvm/Frontend/OpenMP/OMPKinds.def

245

You are correct. But think of it in the same way we have voidptr and int8ptr, even though they are the same.

[jhuber6]

This passes all checks except the libomp ones. Those still seg-fault, I'm not entirely sure why. Cuda failure is again because I don't have it configured on my machine.

Failing Tests (3):

Clang :: Driver/cuda-simple.cu
libomp :: tasking/kmp_taskloop.c
libomp :: worksharing/for/kmp_doacross_check.c

[jdoerfert]

In D80222#2055149, @jhuber6 wrote:

This passes all checks except the libomp ones. Those still seg-fault, I'm not entirely sure why. Cuda failure is again because I don't have it configured on my machine.

Failing Tests (3):

Clang :: Driver/cuda-simple.cu
libomp :: tasking/kmp_taskloop.c
libomp :: worksharing/for/kmp_doacross_check.c

I believe libomp tests are not related to this but you should double check. run them locally without this patch maybe

[jhuber6]

Fixed the issue with the libomp tests and now pass every test. I combined this patch with D79739 so that patch is now obsolete.

[jhuber6]

llvm/include/llvm/Frontend/OpenMP/OMPKinds.def

245

I added these types as @fghanim suggested.

openmp/runtime/test/tasking/kmp_taskloop.c

100 ↗

(On Diff #266325)

I am not entirely sure why, but commenting this out causes the problem to go away. I tried adding proper names to the forward-declared functions but since clang already knew I had something called ident_t, I couldn't declare a new struct with the same name.

[jdoerfert]

llvm/include/llvm/Frontend/OpenMP/OMPKinds.def

250

Why the indirection via __OMP_SIZE_TYPE? Wouldn't OMP_TYPE(SizeTy, M.getDataLayout().getIntPtrType(Ctx)) suffice?

llvm/lib/Frontend/OpenMP/OMPIRBuilder.cpp

136

I now think we either put the stuff in #ifndef NDEBUG or move the FnTy = into the first switch that does the lookup.

openmp/runtime/test/tasking/kmp_taskloop.c

100 ↗

(On Diff #266325)

This is not good. The difference should only be that the kmpc_fork_call has a different argument, right? Does the segfault happen at compile or runtime?

You can just use the ident_t clang created, right? Did you print the function names requested by clang as we discussed?

[jhuber6]

llvm/include/llvm/Frontend/OpenMP/OMPKinds.def

250

I think I did it just to make it more explicit while it was temporary, but now that we're defining it here there's no point, I'll change it.

llvm/lib/Frontend/OpenMP/OMPIRBuilder.cpp

136

Yeah, I was wondering about the cost of the extra lookup. But we could definitely just get the type when we look up the function and just do a comparison in this branch.

openmp/runtime/test/tasking/kmp_taskloop.c

100 ↗

(On Diff #266325)

I added an assertion and debug statements. If I try to declare a struct named "Ident_t" I get the following error message in the seg-fault. I think the seg-fault is compile-time.

Found OpenMP runtime function __kmpc_global_thread_num with type i32 (%struct.ident_t.0*). Expected type is i32 (%struct.ident_t*)
clang: /home/jhuber/Documents/llvm-project/llvm/lib/Frontend/OpenMP/OMPIRBuilder.cpp:124: static llvm::Function* llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(llvm::Module&, llvm::omp::RuntimeFunction): Assertion `FnTy == Fn->getFunctionType() && "Found OpenMP runtime function has mismatched types"' failed.

[jhuber6]

openmp/runtime/test/tasking/kmp_taskloop.c

100 ↗

(On Diff #266325)

I'm not sure if there's a way around this without changing the getOrCreateRuntimeFunction method to return a FunctionCallee and removing the assertion. Clang doesn't know about the ident_t struct when it's compiling the file, but when its doing the codegen it sees two structs with the same name and creates a new name. So when it gets the types it says that ident_t and ident_t.0 don't match. As you said the old version got around this by adding a bitcasting instruction so it knew how to turn it into an ident_t pointer.

[AndreyChurbanov]

openmp/runtime/test/tasking/kmp_taskloop.c

100 ↗

(On Diff #266325)

Note that this change breaks the test on any system with more that 4 procs. Because array th_counter[4] is indexed by thread number which can easily be greater than 3 if number of threads is not limited.

[jhuber6]

Making suggested changes.

[jhuber6]

openmp/runtime/test/tasking/kmp_taskloop.c

100 ↗

(On Diff #266325)

The problem was that the num_threads clause required an implicit call to kmpc_global_thread_num so it could be passed to kmpc_push_num_threads. The types of the implicit function and the forward declaration then wouldn't match up. I added another forward declaration to explicitly call kmpc_push_num_threads. Is this a sufficient solution?

[jdoerfert]

openmp/runtime/test/tasking/kmp_taskloop.c

100 ↗

(On Diff #266325)

We need this to work with num_threads(8).

Clang doesn't know about the ident_t struct when it's compiling the file, but when its doing the codegen it sees two structs with the same name and creates a new name.

Where are the two structs coming from? We should have one. If clang introduces one it needs to use the one from OMPKindes.def instead. Is that a fix?

[jhuber6]

openmp/runtime/test/tasking/kmp_taskloop.c

100 ↗

(On Diff #266325)

The first struct is the one that I'm assuming comes from the OpenMP CodeGen that places the Ident_t struct in the IR file. if I declare a struct also named ident_t in the C source file it most likely will see that there's two structs with the same name and call the second one "ident_t.0" internally. The other ident_t struct is only known once clang generates the LLVM IR so I can't just use "ident_t" nor can I declare a struct with the same name.

[jdoerfert]

openmp/runtime/test/tasking/kmp_taskloop.c

100 ↗

(On Diff #266325)

1. Either Clang needs to use the llvm::omp::types::Ident *or* Clang needs to define llvm::omp::types::Ident and we do not do it via __OMP_STRUCT_TYPE(Ident, ident_t, Int32, Int32, Int32, Int32, Int8Ptr). I would prefer the first solution.

2. OMPConstants.cpp does pick up an existing struct type with the same name if present. That is, probably not what we want because it clashes with user types.

3. We can still return a FunctionCallee with the Function* and the expected type (as defined by OMPKinds.def) to mimic the old behavior for now.

[jhuber6]

openmp/runtime/test/tasking/kmp_taskloop.c

100 ↗

(On Diff #266325)

Either Clang needs to use the llvm::omp::types::Ident *or* Clang needs to define llvm::omp::types::Ident and we do not do it via __OMP_STRUCT_TYPE(Ident, ident_t, Int32, Int32, Int32, Int32, Int8Ptr). I would prefer the first solution.

I'm probably not understanding something correctly here. There's already an ident_t type declared in CGOpenMPRuntime:1065 which it uses for generating the types for the runtime functions, but this isn't used until code generation so it can't be used while compiling the file. If we declare it up front then wouldn't that make ident_t a reserved keyword?

We can still return a FunctionCallee with the Function* and the expected type (as defined by OMPKinds.def) to mimic the old behavior for now.

A FunctionCallee can be generated from a Function * but not vice-versa, right? This would require changing the code in OpenMPIRBuilder.

[jdoerfert]

openmp/runtime/test/tasking/kmp_taskloop.c

100 ↗

(On Diff #266325)

In CGOpenMPRuntime.cpp there is:
IdentTy = CGM.getTypes().ConvertRecordDeclType(RD);
what happens if you replace that with:
IdentTy = llvm::omp::types::Ident;
?

---

FunctionCallee is just a Function* and the "expected type". You can go from FunctionCallee to Function by just picking the proper member. You'd need to change the Builder in two ways: 1) return FunctionCallee(Fn, ExpectedType); 2) where it is used extract the Function * from the FunctionCallee. You might want to use a helper if you go this route.

[jhuber6]

openmp/runtime/test/tasking/kmp_taskloop.c

100 ↗

(On Diff #266325)

It causes all the OpenMP tests to fail, not sure why exactly. I'll look into it more tomorrow.

[jhuber6]

openmp/runtime/test/tasking/kmp_taskloop.c

100 ↗

(On Diff #266325)

I'm not exactly sure why, but if you move the llvm::omp::types:InitializeTypes(CGM.getModule()) statement before where IdenTy is assigned it causes some of the tests to fail without changing anything else. Any clue why?

Instead I changed llvm::Type *CGOpenMPRuntime::getIdentTyPointerTy() in CGOpenMPRuntime:1594 to return llvm::omp::types::Ident->getPointerTo() which is the only place IdentTy is used. But it has the same problem.

[jdoerfert]

Take a look at D80735, it works fine for me locally. Is that what you did? What problems do you observe now?

[jhuber6]

In D80222#2060420, @jdoerfert wrote:

Take a look at D80735, it works fine for me locally. Is that what you did? What problems do you observe now?

That's pretty much what I did. If I kept in the old Ident declaration, but didn't use it and returned llvm::omp::types::IdentPtr from that function it would work. But the compiler would segfault if I removed it. It still failed the two problematic tests. I have a patch that changes the function to return a FunctionCallee and does a bitcast on the function if the types don't match. That passed the tests but it would be beneficial if we could get it to work without needing to change it.

[jhuber6]

Changed getOrCreateRuntimeFunction to return a FunctionCallee instead of a Function*. Added a helper function inside of OpenMPIRBuilder to get the Function*. Added conditional bitcast to change the types if they do not match. This was done inside of CGM.createRuntimeFunction if it found an existing declaration and the types didn't match. Maybe be able to change this back to just return a Function*

[jdoerfert]

This passes all the tests? I think we should go with it and investigate the cast thing later.

llvm/lib/Frontend/OpenMP/OMPIRBuilder.cpp

125

here we probably want to cast the result if we don't do it above. Just do the getBitCast, it should be a no-op if there is nothing to cast.

136

Are you sure we need to do the cast here? I thought clang will create it when it creates calls based of FunctionCallee objects. That was, to my understanding, the idea of the FunctionCallee thing. We might be able to just return {FnTy, Fn} here. If not we could always cast and return, as you noted a function *. Internally, only the expected type makes any sense, everything else will not be handled gracefully.

[jhuber6]

Yes, this passed all the tests on my machine at least.

llvm/lib/Frontend/OpenMP/OMPIRBuilder.cpp

136

I tried it without and the same problem occured. The old method used CGM.createRuntimeFunction which internally has a cast like this one.

[jdoerfert]

I figured out what the problem is and we can relatively easily fix it *once* we always have an OpenMPIRBuilder available in Clangs CG. As noted in D80735, we should remove IdentQTy as it is used to create a new ident_t in some situations. The entire getOrCreateDefaultLocation in clangs CG, should be replaced by OpenMPIRBuilder::getOrCreateIdent. We do this later.

[jhuber6]

This passed all tests on my machine. I changed the bitcast to always try to cast the types. Let me know if you think there's anything else I should change because I think this might be good to go.

[jdoerfert]

LGTM. Thanks for taking this one, it was more complex than I thought but it is a really nice step in the right direction. I'll commit it for you soon if you don't have access yet. Feel free to get access though.

[jhuber6]

In D80222#2061810, @jdoerfert wrote:

LGTM. Thanks for taking this one, it was more complex than I thought but it is a really nice step in the right direction. I'll commit it for you soon if you don't have access yet. Feel free to get access though.

Thanks, I don't have access yet. I'll get on that soon.

[jhuber6]

Merging patch with origin and D80735. This now fails a lot of the tests after the merge. I had a previous branch that was merging it with origin after some new functions were added to CGOpenMPRuntime.cpp that only failed libomp/declare_mapper_api.cpp with a segfault. I'll keep looking into it.

Failing Tests (238):

Clang :: Driver/cuda-simple.cu
Clang :: Headers/nvptx_device_cmath_functions.c
Clang :: Headers/nvptx_device_cmath_functions.cpp
Clang :: Headers/nvptx_device_cmath_functions_cxx17.cpp
Clang :: Headers/nvptx_device_math_complex.c
Clang :: Headers/nvptx_device_math_functions.c
Clang :: Headers/nvptx_device_math_functions.cpp
Clang :: Headers/nvptx_device_math_functions_cxx17.cpp
Clang :: Headers/nvptx_device_math_modf.cpp
Clang :: Headers/nvptx_device_math_sin.c
Clang :: Headers/nvptx_device_math_sin.cpp
Clang :: Headers/nvptx_device_math_sin_cos.cpp
Clang :: Headers/nvptx_device_math_sincos.cpp
Clang :: OpenMP/amdgcn_device_function_call.cpp
Clang :: OpenMP/capturing_in_templates.cpp
Clang :: OpenMP/declare_mapper_codegen.cpp
Clang :: OpenMP/declare_target_codegen.cpp
Clang :: OpenMP/declare_target_codegen_globalization.cpp
Clang :: OpenMP/declare_target_link_codegen.cpp
Clang :: OpenMP/declare_variant_mixed_codegen.c
Clang :: OpenMP/distribute_codegen.cpp
Clang :: OpenMP/distribute_firstprivate_codegen.cpp
Clang :: OpenMP/distribute_lastprivate_codegen.cpp
Clang :: OpenMP/distribute_parallel_for_codegen.cpp
Clang :: OpenMP/distribute_parallel_for_firstprivate_codegen.cpp
Clang :: OpenMP/distribute_parallel_for_if_codegen.cpp
Clang :: OpenMP/distribute_parallel_for_lastprivate_codegen.cpp
Clang :: OpenMP/distribute_parallel_for_num_threads_codegen.cpp
Clang :: OpenMP/distribute_parallel_for_private_codegen.cpp
Clang :: OpenMP/distribute_parallel_for_proc_bind_codegen.cpp
Clang :: OpenMP/distribute_parallel_for_reduction_codegen.cpp
Clang :: OpenMP/distribute_parallel_for_simd_codegen.cpp
Clang :: OpenMP/distribute_parallel_for_simd_firstprivate_codegen.cpp
Clang :: OpenMP/distribute_parallel_for_simd_if_codegen.cpp
Clang :: OpenMP/distribute_parallel_for_simd_lastprivate_codegen.cpp
Clang :: OpenMP/distribute_parallel_for_simd_num_threads_codegen.cpp
Clang :: OpenMP/distribute_parallel_for_simd_private_codegen.cpp
Clang :: OpenMP/distribute_parallel_for_simd_proc_bind_codegen.cpp
Clang :: OpenMP/distribute_private_codegen.cpp
Clang :: OpenMP/distribute_simd_codegen.cpp
Clang :: OpenMP/distribute_simd_firstprivate_codegen.cpp
Clang :: OpenMP/distribute_simd_lastprivate_codegen.cpp
Clang :: OpenMP/distribute_simd_private_codegen.cpp
Clang :: OpenMP/distribute_simd_reduction_codegen.cpp
Clang :: OpenMP/is_initial_device.c
Clang :: OpenMP/nvptx_SPMD_codegen.cpp
Clang :: OpenMP/nvptx_data_sharing.cpp
Clang :: OpenMP/nvptx_declare_target_var_ctor_dtor_codegen.cpp
Clang :: OpenMP/nvptx_declare_variant_device_kind_codegen.cpp
Clang :: OpenMP/nvptx_declare_variant_implementation_vendor_codegen.cpp
Clang :: OpenMP/nvptx_distribute_parallel_generic_mode_codegen.cpp
Clang :: OpenMP/nvptx_force_full_runtime_SPMD_codegen.cpp
Clang :: OpenMP/nvptx_lambda_capturing.cpp
Clang :: OpenMP/nvptx_lambda_pointer_capturing.cpp
Clang :: OpenMP/nvptx_parallel_codegen.cpp
Clang :: OpenMP/nvptx_parallel_for_codegen.cpp
Clang :: OpenMP/nvptx_param_translate.c
Clang :: OpenMP/nvptx_target_codegen.cpp
Clang :: OpenMP/nvptx_target_cuda_mode_messages.cpp
Clang :: OpenMP/nvptx_target_exceptions_messages.cpp
Clang :: OpenMP/nvptx_target_firstprivate_codegen.cpp
Clang :: OpenMP/nvptx_target_parallel_codegen.cpp
Clang :: OpenMP/nvptx_target_parallel_num_threads_codegen.cpp
Clang :: OpenMP/nvptx_target_parallel_proc_bind_codegen.cpp
Clang :: OpenMP/nvptx_target_parallel_reduction_codegen.cpp
Clang :: OpenMP/nvptx_target_printf_codegen.c
Clang :: OpenMP/nvptx_target_pure_deleted_codegen.cpp
Clang :: OpenMP/nvptx_target_requires_unified_shared_memory.cpp
Clang :: OpenMP/nvptx_target_simd_codegen.cpp
Clang :: OpenMP/nvptx_target_teams_codegen.cpp
Clang :: OpenMP/nvptx_target_teams_distribute_codegen.cpp
Clang :: OpenMP/nvptx_target_teams_distribute_parallel_for_codegen.cpp
Clang :: OpenMP/nvptx_target_teams_distribute_parallel_for_generic_mode_codegen.cpp
Clang :: OpenMP/nvptx_target_teams_distribute_parallel_for_simd_codegen.cpp
Clang :: OpenMP/nvptx_target_teams_distribute_simd_codegen.cpp
Clang :: OpenMP/nvptx_teams_codegen.cpp
Clang :: OpenMP/nvptx_teams_reduction_codegen.cpp
Clang :: OpenMP/nvptx_unsupported_type_messages.cpp
Clang :: OpenMP/openmp_offload_codegen.cpp
Clang :: OpenMP/openmp_offload_registration.cpp
Clang :: OpenMP/parallel_codegen.cpp
Clang :: OpenMP/parallel_master_codegen.cpp
Clang :: OpenMP/parallel_master_taskloop_simd_reduction_codegen.cpp
Clang :: OpenMP/target_codegen.cpp
Clang :: OpenMP/target_codegen_global_capture.cpp
Clang :: OpenMP/target_codegen_registration.cpp
Clang :: OpenMP/target_codegen_registration_naming.cpp
Clang :: OpenMP/target_constant_device_codegen.cpp
Clang :: OpenMP/target_defaultmap_codegen.cpp
Clang :: OpenMP/target_depend_codegen.cpp
Clang :: OpenMP/target_device_codegen.cpp
Clang :: OpenMP/target_firstprivate_codegen.cpp
Clang :: OpenMP/target_is_device_ptr_codegen.cpp
Clang :: OpenMP/target_map_codegen.cpp
Clang :: OpenMP/target_messages.cpp
Clang :: OpenMP/target_parallel_codegen.cpp
Clang :: OpenMP/target_parallel_codegen_registration.cpp
Clang :: OpenMP/target_parallel_codegen_registration_naming.cpp
Clang :: OpenMP/target_parallel_debug_codegen.cpp
Clang :: OpenMP/target_parallel_depend_codegen.cpp
Clang :: OpenMP/target_parallel_for_codegen.cpp
Clang :: OpenMP/target_parallel_for_codegen_registration.cpp
Clang :: OpenMP/target_parallel_for_codegen_registration_naming.cpp
Clang :: OpenMP/target_parallel_for_debug_codegen.cpp
Clang :: OpenMP/target_parallel_for_depend_codegen.cpp
Clang :: OpenMP/target_parallel_for_simd_codegen.cpp
Clang :: OpenMP/target_parallel_for_simd_codegen_registration.cpp
Clang :: OpenMP/target_parallel_for_simd_codegen_registration_naming.cpp
Clang :: OpenMP/target_parallel_for_simd_depend_codegen.cpp
Clang :: OpenMP/target_parallel_for_simd_uses_allocators_codegen.cpp
Clang :: OpenMP/target_parallel_for_uses_allocators_codegen.cpp
Clang :: OpenMP/target_parallel_if_codegen.cpp
Clang :: OpenMP/target_parallel_no_exceptions.cpp
Clang :: OpenMP/target_parallel_num_threads_codegen.cpp
Clang :: OpenMP/target_parallel_uses_allocators_codegen.cpp
Clang :: OpenMP/target_private_codegen.cpp
Clang :: OpenMP/target_reduction_codegen.cpp
Clang :: OpenMP/target_simd_codegen.cpp
Clang :: OpenMP/target_simd_codegen_registration.cpp
Clang :: OpenMP/target_simd_codegen_registration_naming.cpp
Clang :: OpenMP/target_simd_depend_codegen.cpp
Clang :: OpenMP/target_simd_uses_allocators_codegen.cpp
Clang :: OpenMP/target_teams_codegen.cpp
Clang :: OpenMP/target_teams_codegen_registration.cpp
Clang :: OpenMP/target_teams_codegen_registration_naming.cpp
Clang :: OpenMP/target_teams_depend_codegen.cpp
Clang :: OpenMP/target_teams_distribute_codegen.cpp
Clang :: OpenMP/target_teams_distribute_codegen_registration.cpp
Clang :: OpenMP/target_teams_distribute_codegen_registration_naming.cpp
Clang :: OpenMP/target_teams_distribute_collapse_codegen.cpp
Clang :: OpenMP/target_teams_distribute_depend_codegen.cpp
Clang :: OpenMP/target_teams_distribute_dist_schedule_codegen.cpp
Clang :: OpenMP/target_teams_distribute_firstprivate_codegen.cpp
Clang :: OpenMP/target_teams_distribute_lastprivate_codegen.cpp
Clang :: OpenMP/target_teams_distribute_parallel_for_codegen.cpp
Clang :: OpenMP/target_teams_distribute_parallel_for_collapse_codegen.cpp
Clang :: OpenMP/target_teams_distribute_parallel_for_depend_codegen.cpp
Clang :: OpenMP/target_teams_distribute_parallel_for_dist_schedule_codegen.cpp
Clang :: OpenMP/target_teams_distribute_parallel_for_firstprivate_codegen.cpp
Clang :: OpenMP/target_teams_distribute_parallel_for_if_codegen.cpp
Clang :: OpenMP/target_teams_distribute_parallel_for_lastprivate_codegen.cpp
Clang :: OpenMP/target_teams_distribute_parallel_for_order_codegen.cpp
Clang :: OpenMP/target_teams_distribute_parallel_for_private_codegen.cpp
Clang :: OpenMP/target_teams_distribute_parallel_for_proc_bind_codegen.cpp
Clang :: OpenMP/target_teams_distribute_parallel_for_reduction_codegen.cpp
Clang :: OpenMP/target_teams_distribute_parallel_for_schedule_codegen.cpp
Clang :: OpenMP/target_teams_distribute_parallel_for_simd_codegen.cpp
Clang :: OpenMP/target_teams_distribute_parallel_for_simd_codegen_registration.cpp
Clang :: OpenMP/target_teams_distribute_parallel_for_simd_codegen_registration_naming.cpp
Clang :: OpenMP/target_teams_distribute_parallel_for_simd_collapse_codegen.cpp
Clang :: OpenMP/target_teams_distribute_parallel_for_simd_depend_codegen.cpp
Clang :: OpenMP/target_teams_distribute_parallel_for_simd_dist_schedule_codegen.cpp
Clang :: OpenMP/target_teams_distribute_parallel_for_simd_firstprivate_codegen.cpp
Clang :: OpenMP/target_teams_distribute_parallel_for_simd_if_codegen.cpp
Clang :: OpenMP/target_teams_distribute_parallel_for_simd_lastprivate_codegen.cpp
Clang :: OpenMP/target_teams_distribute_parallel_for_simd_private_codegen.cpp
Clang :: OpenMP/target_teams_distribute_parallel_for_simd_proc_bind_codegen.cpp
Clang :: OpenMP/target_teams_distribute_parallel_for_simd_reduction_codegen.cpp
Clang :: OpenMP/target_teams_distribute_parallel_for_simd_schedule_codegen.cpp
Clang :: OpenMP/target_teams_distribute_parallel_for_simd_uses_allocators_codegen.cpp
Clang :: OpenMP/target_teams_distribute_parallel_for_uses_allocators_codegen.cpp
Clang :: OpenMP/target_teams_distribute_private_codegen.cpp
Clang :: OpenMP/target_teams_distribute_reduction_codegen.cpp
Clang :: OpenMP/target_teams_distribute_simd_codegen.cpp
Clang :: OpenMP/target_teams_distribute_simd_codegen_registration.cpp
Clang :: OpenMP/target_teams_distribute_simd_codegen_registration_naming.cpp
Clang :: OpenMP/target_teams_distribute_simd_collapse_codegen.cpp
Clang :: OpenMP/target_teams_distribute_simd_depend_codegen.cpp
Clang :: OpenMP/target_teams_distribute_simd_dist_schedule_codegen.cpp
Clang :: OpenMP/target_teams_distribute_simd_firstprivate_codegen.cpp
Clang :: OpenMP/target_teams_distribute_simd_lastprivate_codegen.cpp
Clang :: OpenMP/target_teams_distribute_simd_private_codegen.cpp
Clang :: OpenMP/target_teams_distribute_simd_reduction_codegen.cpp
Clang :: OpenMP/target_teams_distribute_simd_uses_allocators_codegen.cpp
Clang :: OpenMP/target_teams_distribute_uses_allocators_codegen.cpp
Clang :: OpenMP/target_teams_map_codegen.cpp
Clang :: OpenMP/target_teams_num_teams_codegen.cpp
Clang :: OpenMP/target_teams_thread_limit_codegen.cpp
Clang :: OpenMP/target_teams_uses_allocators_codegen.cpp
Clang :: OpenMP/target_uses_allocators_codegen.cpp
Clang :: OpenMP/target_vla_messages.cpp
Clang :: OpenMP/teams_codegen.cpp
Clang :: OpenMP/teams_distribute_codegen.cpp
Clang :: OpenMP/teams_distribute_collapse_codegen.cpp
Clang :: OpenMP/teams_distribute_dist_schedule_codegen.cpp
Clang :: OpenMP/teams_distribute_firstprivate_codegen.cpp
Clang :: OpenMP/teams_distribute_lastprivate_codegen.cpp
Clang :: OpenMP/teams_distribute_parallel_for_codegen.cpp
Clang :: OpenMP/teams_distribute_parallel_for_collapse_codegen.cpp
Clang :: OpenMP/teams_distribute_parallel_for_copyin_codegen.cpp
Clang :: OpenMP/teams_distribute_parallel_for_dist_schedule_codegen.cpp
Clang :: OpenMP/teams_distribute_parallel_for_firstprivate_codegen.cpp
Clang :: OpenMP/teams_distribute_parallel_for_if_codegen.cpp
Clang :: OpenMP/teams_distribute_parallel_for_lastprivate_codegen.cpp
Clang :: OpenMP/teams_distribute_parallel_for_num_threads_codegen.cpp
Clang :: OpenMP/teams_distribute_parallel_for_private_codegen.cpp
Clang :: OpenMP/teams_distribute_parallel_for_proc_bind_codegen.cpp
Clang :: OpenMP/teams_distribute_parallel_for_reduction_codegen.cpp
Clang :: OpenMP/teams_distribute_parallel_for_schedule_codegen.cpp
Clang :: OpenMP/teams_distribute_parallel_for_simd_codegen.cpp
Clang :: OpenMP/teams_distribute_parallel_for_simd_collapse_codegen.cpp
Clang :: OpenMP/teams_distribute_parallel_for_simd_dist_schedule_codegen.cpp
Clang :: OpenMP/teams_distribute_parallel_for_simd_firstprivate_codegen.cpp
Clang :: OpenMP/teams_distribute_parallel_for_simd_if_codegen.cpp
Clang :: OpenMP/teams_distribute_parallel_for_simd_lastprivate_codegen.cpp
Clang :: OpenMP/teams_distribute_parallel_for_simd_num_threads_codegen.cpp
Clang :: OpenMP/teams_distribute_parallel_for_simd_private_codegen.cpp
Clang :: OpenMP/teams_distribute_parallel_for_simd_proc_bind_codegen.cpp
Clang :: OpenMP/teams_distribute_parallel_for_simd_reduction_codegen.cpp
Clang :: OpenMP/teams_distribute_parallel_for_simd_schedule_codegen.cpp
Clang :: OpenMP/teams_distribute_private_codegen.cpp
Clang :: OpenMP/teams_distribute_reduction_codegen.cpp
Clang :: OpenMP/teams_distribute_simd_codegen.cpp
Clang :: OpenMP/teams_distribute_simd_collapse_codegen.cpp
Clang :: OpenMP/teams_distribute_simd_dist_schedule_codegen.cpp
Clang :: OpenMP/teams_distribute_simd_firstprivate_codegen.cpp
Clang :: OpenMP/teams_distribute_simd_lastprivate_codegen.cpp
Clang :: OpenMP/teams_distribute_simd_private_codegen.cpp
Clang :: OpenMP/teams_distribute_simd_reduction_codegen.cpp
Clang :: OpenMP/teams_firstprivate_codegen.cpp
Clang :: OpenMP/teams_private_codegen.cpp
Clang :: OpenMP/threadprivate_codegen.cpp
LLVM :: Transforms/OpenMP/deduplication.ll
libomptarget :: api/omp_get_num_devices.c
libomptarget :: mapping/declare_mapper_api.cpp
libomptarget :: mapping/delete_inf_refcount.c
libomptarget :: mapping/pr38704.c
libomptarget :: offloading/dynamic_module.c
libomptarget :: offloading/dynamic_module_load.c
libomptarget :: offloading/offloading_success.c
libomptarget :: offloading/offloading_success.cpp
libomptarget :: offloading/parallel_offloading_map.c
libomptarget :: offloading/target_depend_nowait.cpp
libomptarget :: unified_shared_memory/api.c
libomptarget :: unified_shared_memory/close_enter_exit.c
libomptarget :: unified_shared_memory/close_manual.c
libomptarget :: unified_shared_memory/close_modifier.c
libomptarget :: unified_shared_memory/shared_update.c

[jdoerfert]

Initialize the internal OpenMPIRbuilder (see my patch).

[jhuber6]

In D80222#2065141, @jdoerfert wrote:

Initialize the internal OpenMPIRbuilder (see my patch).

Isn't it initialized on line 1063 in CGOpenMPRuntime.cpp? There weren't any conflicts listed there when I merged your patch so it should be the same.

[jdoerfert]

Oh, sorry. You merged all of my patch, right? It was not ready and even now it is unclear if the representation change in my patch is good or not. I was trying to suggest you only take the InternalOMPIRBuilder stuff so you can avoid the static function, not all of my patch. Sorry.

[jhuber6]

Reverting changed from D80735 and merging with origin/master. Adding an OpenMPIRBuilder object and removing the static methods caused a lot of the tests to fail. I think I'm going to try to get this version done and we can modify it later.

clang/lib/CodeGen/CGOpenMPRuntime.h

private:
   void scanForTargetRegionsFunctions(const Stmt *S, StringRef ParentName);
 
   /// Build type kmp_routine_entry_t (if not built yet).
   void emitKmpRoutineEntryT(QualType KmpInt32Ty);
 
   /// Returns pointer to kmpc_micro type.
   llvm::Type *getKmpc_MicroPointerTy();
 
-  /// Returns specified OpenMP runtime function.
-  /// \param Function OpenMP runtime function.
-  /// \return Specified function.
-  llvm::FunctionCallee createRuntimeFunction(unsigned Function);
-
   /// Returns __kmpc_for_static_init_* runtime function for the specified
   /// size \a IVSize and sign \a IVSigned.
   llvm::FunctionCallee createForStaticInitFunction(unsigned IVSize,
                                                    bool IVSigned);
 
   /// Returns __kmpc_dispatch_init_* runtime function for the specified
   /// size \a IVSize and sign \a IVSigned.
   llvm::FunctionCallee createDispatchInitFunction(unsigned IVSize,

clang/lib/CodeGen/CGOpenMPRuntime.cpp

enum OpenMPSchedType {
   OMP_dist_sch_static = 92,
   /// Support for OpenMP 4.5 monotonic and nonmonotonic schedule modifiers.
   /// Set if the monotonic schedule modifier was present.
   OMP_sch_modifier_monotonic = (1 << 29),
   /// Set if the nonmonotonic schedule modifier was present.
   OMP_sch_modifier_nonmonotonic = (1 << 30),
 };
 
-enum OpenMPRTLFunction {
-  /// Call to void __kmpc_fork_call(ident_t *loc, kmp_int32 argc,
-  /// kmpc_micro microtask, ...);
-  OMPRTL__kmpc_fork_call,
-  /// Call to void *__kmpc_threadprivate_cached(ident_t *loc,
-  /// kmp_int32 global_tid, void *data, size_t size, void ***cache);
-  OMPRTL__kmpc_threadprivate_cached,
-  /// Call to void __kmpc_threadprivate_register( ident_t *,
-  /// void *data, kmpc_ctor ctor, kmpc_cctor cctor, kmpc_dtor dtor);
-  OMPRTL__kmpc_threadprivate_register,
-  // Call to __kmpc_int32 kmpc_global_thread_num(ident_t *loc);
-  OMPRTL__kmpc_global_thread_num,
-  // Call to void __kmpc_critical(ident_t *loc, kmp_int32 global_tid,
-  // kmp_critical_name *crit);
-  OMPRTL__kmpc_critical,
-  // Call to void __kmpc_critical_with_hint(ident_t *loc, kmp_int32
-  // global_tid, kmp_critical_name *crit, uintptr_t hint);
-  OMPRTL__kmpc_critical_with_hint,
-  // Call to void __kmpc_end_critical(ident_t *loc, kmp_int32 global_tid,
-  // kmp_critical_name *crit);
-  OMPRTL__kmpc_end_critical,
-  // Call to kmp_int32 __kmpc_cancel_barrier(ident_t *loc, kmp_int32
-  // global_tid);
-  OMPRTL__kmpc_cancel_barrier,
-  // Call to void __kmpc_barrier(ident_t *loc, kmp_int32 global_tid);
-  OMPRTL__kmpc_barrier,
-  // Call to void __kmpc_for_static_fini(ident_t *loc, kmp_int32 global_tid);
-  OMPRTL__kmpc_for_static_fini,
-  // Call to void __kmpc_serialized_parallel(ident_t *loc, kmp_int32
-  // global_tid);
-  OMPRTL__kmpc_serialized_parallel,
-  // Call to void __kmpc_end_serialized_parallel(ident_t *loc, kmp_int32
-  // global_tid);
-  OMPRTL__kmpc_end_serialized_parallel,
-  // Call to void __kmpc_push_num_threads(ident_t *loc, kmp_int32 global_tid,
-  // kmp_int32 num_threads);
-  OMPRTL__kmpc_push_num_threads,
-  // Call to void __kmpc_flush(ident_t *loc);
-  OMPRTL__kmpc_flush,
-  // Call to kmp_int32 __kmpc_master(ident_t *, kmp_int32 global_tid);
-  OMPRTL__kmpc_master,
-  // Call to void __kmpc_end_master(ident_t *, kmp_int32 global_tid);
-  OMPRTL__kmpc_end_master,
-  // Call to kmp_int32 __kmpc_omp_taskyield(ident_t *, kmp_int32 global_tid,
-  // int end_part);
-  OMPRTL__kmpc_omp_taskyield,
-  // Call to kmp_int32 __kmpc_single(ident_t *, kmp_int32 global_tid);
-  OMPRTL__kmpc_single,
-  // Call to void __kmpc_end_single(ident_t *, kmp_int32 global_tid);
-  OMPRTL__kmpc_end_single,
-  // Call to kmp_task_t * __kmpc_omp_task_alloc(ident_t *, kmp_int32 gtid,
-  // kmp_int32 flags, size_t sizeof_kmp_task_t, size_t sizeof_shareds,
-  // kmp_routine_entry_t *task_entry);
-  OMPRTL__kmpc_omp_task_alloc,
-  // Call to kmp_task_t * __kmpc_omp_target_task_alloc(ident_t *,
-  // kmp_int32 gtid, kmp_int32 flags, size_t sizeof_kmp_task_t,
-  // size_t sizeof_shareds, kmp_routine_entry_t *task_entry,
-  // kmp_int64 device_id);
-  OMPRTL__kmpc_omp_target_task_alloc,
-  // Call to kmp_int32 __kmpc_omp_task(ident_t *, kmp_int32 gtid, kmp_task_t *
-  // new_task);
-  OMPRTL__kmpc_omp_task,
-  // Call to void __kmpc_copyprivate(ident_t *loc, kmp_int32 global_tid,
-  // size_t cpy_size, void *cpy_data, void(*cpy_func)(void *, void *),
-  // kmp_int32 didit);
-  OMPRTL__kmpc_copyprivate,
-  // Call to kmp_int32 __kmpc_reduce(ident_t *loc, kmp_int32 global_tid,
-  // kmp_int32 num_vars, size_t reduce_size, void *reduce_data, void
-  // (*reduce_func)(void *lhs_data, void *rhs_data), kmp_critical_name *lck);
-  OMPRTL__kmpc_reduce,
-  // Call to kmp_int32 __kmpc_reduce_nowait(ident_t *loc, kmp_int32
-  // global_tid, kmp_int32 num_vars, size_t reduce_size, void *reduce_data,
-  // void (*reduce_func)(void *lhs_data, void *rhs_data), kmp_critical_name
-  // *lck);
-  OMPRTL__kmpc_reduce_nowait,
-  // Call to void __kmpc_end_reduce(ident_t *loc, kmp_int32 global_tid,
-  // kmp_critical_name *lck);
-  OMPRTL__kmpc_end_reduce,
-  // Call to void __kmpc_end_reduce_nowait(ident_t *loc, kmp_int32 global_tid,
-  // kmp_critical_name *lck);
-  OMPRTL__kmpc_end_reduce_nowait,
-  // Call to void __kmpc_omp_task_begin_if0(ident_t *, kmp_int32 gtid,
-  // kmp_task_t * new_task);
-  OMPRTL__kmpc_omp_task_begin_if0,
-  // Call to void __kmpc_omp_task_complete_if0(ident_t *, kmp_int32 gtid,
-  // kmp_task_t * new_task);
-  OMPRTL__kmpc_omp_task_complete_if0,
-  // Call to void __kmpc_ordered(ident_t *loc, kmp_int32 global_tid);
-  OMPRTL__kmpc_ordered,
-  // Call to void __kmpc_end_ordered(ident_t *loc, kmp_int32 global_tid);
-  OMPRTL__kmpc_end_ordered,
-  // Call to kmp_int32 __kmpc_omp_taskwait(ident_t *loc, kmp_int32
-  // global_tid);
-  OMPRTL__kmpc_omp_taskwait,
-  // Call to void __kmpc_taskgroup(ident_t *loc, kmp_int32 global_tid);
-  OMPRTL__kmpc_taskgroup,
-  // Call to void __kmpc_end_taskgroup(ident_t *loc, kmp_int32 global_tid);
-  OMPRTL__kmpc_end_taskgroup,
-  // Call to void __kmpc_push_proc_bind(ident_t *loc, kmp_int32 global_tid,
-  // int proc_bind);
-  OMPRTL__kmpc_push_proc_bind,
-  // Call to kmp_int32 __kmpc_omp_task_with_deps(ident_t *loc_ref, kmp_int32
-  // gtid, kmp_task_t * new_task, kmp_int32 ndeps, kmp_depend_info_t
-  // *dep_list, kmp_int32 ndeps_noalias, kmp_depend_info_t *noalias_dep_list);
-  OMPRTL__kmpc_omp_task_with_deps,
-  // Call to void __kmpc_omp_wait_deps(ident_t *loc_ref, kmp_int32
-  // gtid, kmp_int32 ndeps, kmp_depend_info_t *dep_list, kmp_int32
-  // ndeps_noalias, kmp_depend_info_t *noalias_dep_list);
-  OMPRTL__kmpc_omp_wait_deps,
-  // Call to kmp_int32 __kmpc_cancellationpoint(ident_t *loc, kmp_int32
-  // global_tid, kmp_int32 cncl_kind);
-  OMPRTL__kmpc_cancellationpoint,
-  // Call to kmp_int32 __kmpc_cancel(ident_t *loc, kmp_int32 global_tid,
-  // kmp_int32 cncl_kind);
-  OMPRTL__kmpc_cancel,
-  // Call to void __kmpc_push_num_teams(ident_t *loc, kmp_int32 global_tid,
-  // kmp_int32 num_teams, kmp_int32 thread_limit);
-  OMPRTL__kmpc_push_num_teams,
-  // Call to void __kmpc_fork_teams(ident_t *loc, kmp_int32 argc, kmpc_micro
-  // microtask, ...);
-  OMPRTL__kmpc_fork_teams,
-  // Call to void __kmpc_taskloop(ident_t *loc, int gtid, kmp_task_t *task, int
-  // if_val, kmp_uint64 *lb, kmp_uint64 *ub, kmp_int64 st, int nogroup, int
-  // sched, kmp_uint64 grainsize, void *task_dup);
-  OMPRTL__kmpc_taskloop,
-  // Call to void __kmpc_doacross_init(ident_t *loc, kmp_int32 gtid, kmp_int32
-  // num_dims, struct kmp_dim *dims);
-  OMPRTL__kmpc_doacross_init,
-  // Call to void __kmpc_doacross_fini(ident_t *loc, kmp_int32 gtid);
-  OMPRTL__kmpc_doacross_fini,
-  // Call to void __kmpc_doacross_post(ident_t *loc, kmp_int32 gtid, kmp_int64
-  // *vec);
-  OMPRTL__kmpc_doacross_post,
-  // Call to void __kmpc_doacross_wait(ident_t *loc, kmp_int32 gtid, kmp_int64
-  // *vec);
-  OMPRTL__kmpc_doacross_wait,
-  // Call to void *__kmpc_taskred_init(int gtid, int num_data, void *data);
-  OMPRTL__kmpc_taskred_init,
-  // Call to void *__kmpc_task_reduction_get_th_data(int gtid, void *tg, void
-  // *d);
-  OMPRTL__kmpc_task_reduction_get_th_data,
-  // Call to void *__kmpc_taskred_modifier_init(ident_t *loc, int gtid, int
-  // is_ws, int num, void *data);
-  OMPRTL__kmpc_taskred_modifier_init,
-  // Call to void __kmpc_task_reduction_modifier_fini(ident_t *loc, int gtid,
-  // int is_ws);
-  OMPRTL__kmpc_task_reduction_modifier_fini,
-  // Call to void *__kmpc_alloc(int gtid, size_t sz, omp_allocator_handle_t al);
-  OMPRTL__kmpc_alloc,
-  // Call to void __kmpc_free(int gtid, void *ptr, omp_allocator_handle_t al);
-  OMPRTL__kmpc_free,
-
-  //
-  // Offloading related calls
-  //
-  // Call to void __kmpc_push_target_tripcount(int64_t device_id, kmp_uint64
-  // size);
-  OMPRTL__kmpc_push_target_tripcount,
-  // Call to int32_t __tgt_target(int64_t device_id, void *host_ptr, int32_t
-  // arg_num, void** args_base, void **args, int64_t *arg_sizes, int64_t
-  // *arg_types);
-  OMPRTL__tgt_target,
-  // Call to int32_t __tgt_target_nowait(int64_t device_id, void *host_ptr,
-  // int32_t arg_num, void** args_base, void **args, int64_t *arg_sizes, int64_t
-  // *arg_types);
-  OMPRTL__tgt_target_nowait,
-  // Call to int32_t __tgt_target_teams(int64_t device_id, void *host_ptr,
-  // int32_t arg_num, void** args_base, void **args, int64_t *arg_sizes, int64_t
-  // *arg_types, int32_t num_teams, int32_t thread_limit);
-  OMPRTL__tgt_target_teams,
-  // Call to int32_t __tgt_target_teams_nowait(int64_t device_id, void
-  // *host_ptr, int32_t arg_num, void** args_base, void **args, int64_t
-  // *arg_sizes, int64_t *arg_types, int32_t num_teams, int32_t thread_limit);
-  OMPRTL__tgt_target_teams_nowait,
-  // Call to void __tgt_register_requires(int64_t flags);
-  OMPRTL__tgt_register_requires,
-  // Call to void __tgt_target_data_begin(int64_t device_id, int32_t arg_num,
-  // void** args_base, void **args, int64_t *arg_sizes, int64_t *arg_types);
-  OMPRTL__tgt_target_data_begin,
-  // Call to void __tgt_target_data_begin_nowait(int64_t device_id, int32_t
-  // arg_num, void** args_base, void **args, int64_t *arg_sizes, int64_t
-  // *arg_types);
-  OMPRTL__tgt_target_data_begin_nowait,
-  // Call to void __tgt_target_data_end(int64_t device_id, int32_t arg_num,
-  // void** args_base, void **args, size_t *arg_sizes, int64_t *arg_types);
-  OMPRTL__tgt_target_data_end,
-  // Call to void __tgt_target_data_end_nowait(int64_t device_id, int32_t
-  // arg_num, void** args_base, void **args, int64_t *arg_sizes, int64_t
-  // *arg_types);
-  OMPRTL__tgt_target_data_end_nowait,
-  // Call to void __tgt_target_data_update(int64_t device_id, int32_t arg_num,
-  // void** args_base, void **args, int64_t *arg_sizes, int64_t *arg_types);
-  OMPRTL__tgt_target_data_update,
-  // Call to void __tgt_target_data_update_nowait(int64_t device_id, int32_t
-  // arg_num, void** args_base, void **args, int64_t *arg_sizes, int64_t
-  // *arg_types);
-  OMPRTL__tgt_target_data_update_nowait,
-  // Call to int64_t __tgt_mapper_num_components(void *rt_mapper_handle);
-  OMPRTL__tgt_mapper_num_components,
-  // Call to void __tgt_push_mapper_component(void *rt_mapper_handle, void
-  // *base, void *begin, int64_t size, int64_t type);
-  OMPRTL__tgt_push_mapper_component,
-  // Call to kmp_event_t *__kmpc_task_allow_completion_event(ident_t *loc_ref,
-  // int gtid, kmp_task_t *task);
-  OMPRTL__kmpc_task_allow_completion_event,
-};
-
 /// A basic class for pre|post-action for advanced codegen sequence for OpenMP
 /// region.
 class CleanupTy final : public EHScopeStack::Cleanup {
   PrePostActionTy *Action;
 
 public:
   explicit CleanupTy(PrePostActionTy *Action) : Action(Action) {}
   void Emit(CodeGenFunction &CGF, Flags /*flags*/) override {
CGOpenMPRuntime::CGOpenMPRuntime(CodeGenModule &CGM, StringRef FirstSeparator,
   addFieldToRecordDecl(C, RD, KmpInt32Ty);
   // psource
   addFieldToRecordDecl(C, RD, C.VoidPtrTy);
   RD->completeDefinition();
   IdentQTy = C.getRecordType(RD);
   IdentTy = CGM.getTypes().ConvertRecordDeclType(RD);
   KmpCriticalNameTy = llvm::ArrayType::get(CGM.Int32Ty, /*NumElements*/ 8);
 
+  // Initialize Types used in OpenMPIRBuilder from OMPKinds.def
+  llvm::omp::types::initializeTypes(CGM.getModule());
   loadOffloadInfoMetadata();
 }
 
 void CGOpenMPRuntime::clear() {
   InternalVars.clear();
   // Clean non-target variable declarations possibly used only in debug info.
   for (const auto &Data : EmittedNonTargetVariables) {
     if (!Data.getValue().pointsToAliveValue())
auto &&UntiedCodeGen = [this, &D, TaskTVar](CodeGenFunction &CGF,
                                               PrePostActionTy &) {
     llvm::Value *ThreadID = getThreadID(CGF, D.getBeginLoc());
     llvm::Value *UpLoc = emitUpdateLocation(CGF, D.getBeginLoc());
     llvm::Value *TaskArgs[] = {
         UpLoc, ThreadID,
         CGF.EmitLoadOfPointerLValue(CGF.GetAddrOfLocalVar(TaskTVar),
                                     TaskTVar->getType()->castAs<PointerType>())
             .getPointer(CGF)};
-    CGF.EmitRuntimeCall(createRuntimeFunction(OMPRTL__kmpc_omp_task), TaskArgs);
+    CGF.EmitRuntimeCall(llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+                            CGM.getModule(), OMPRTL___kmpc_omp_task),
+                        TaskArgs);
   };
   CGOpenMPTaskOutlinedRegionInfo::UntiedTaskActionTy Action(Tied, PartIDVar,
                                                             UntiedCodeGen);
   CodeGen.setAction(Action);
   assert(!ThreadIDVar->getType()->isPointerType() &&
          "thread id variable must be of type kmp_int32 for tasks");
   const OpenMPDirectiveKind Region =
       isOpenMPTaskLoopDirective(D.getDirectiveKind()) ? OMPD_taskloop
llvm::Value *CGOpenMPRuntime::getThreadID(CodeGenFunction &CGF,
   // Generate thread id value and cache this value for use across the
   // function.
   auto &Elem = OpenMPLocThreadIDMap.FindAndConstruct(CGF.CurFn);
   if (!Elem.second.ServiceInsertPt)
     setLocThreadIdInsertPt(CGF);
   CGBuilderTy::InsertPointGuard IPG(CGF.Builder);
   CGF.Builder.SetInsertPoint(Elem.second.ServiceInsertPt);
   llvm::CallInst *Call = CGF.Builder.CreateCall(
-      createRuntimeFunction(OMPRTL__kmpc_global_thread_num),
+      llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+          CGM.getModule(), OMPRTL___kmpc_global_thread_num),
       emitUpdateLocation(CGF, Loc));
   Call->setCallingConv(CGF.getRuntimeCC());
   Elem.second.ThreadID = Call;
   return Call;
 }
 
 void CGOpenMPRuntime::functionFinished(CodeGenFunction &CGF) {
   assert(CGF.CurFn && "No function in current CodeGenFunction.");
if (!Kmpc_MicroTy) {
     // Build void (*kmpc_micro)(kmp_int32 *global_tid, kmp_int32 *bound_tid,...)
     llvm::Type *MicroParams[] = {llvm::PointerType::getUnqual(CGM.Int32Ty),
                                  llvm::PointerType::getUnqual(CGM.Int32Ty)};
     Kmpc_MicroTy = llvm::FunctionType::get(CGM.VoidTy, MicroParams, true);
   }
   return llvm::PointerType::getUnqual(Kmpc_MicroTy);
 }
 
-llvm::FunctionCallee CGOpenMPRuntime::createRuntimeFunction(unsigned Function) {
-  llvm::FunctionCallee RTLFn = nullptr;
-  switch (static_cast<OpenMPRTLFunction>(Function)) {
-  case OMPRTL__kmpc_fork_call: {
-    // Build void __kmpc_fork_call(ident_t *loc, kmp_int32 argc, kmpc_micro
-    // microtask, ...);
-    llvm::Type *TypeParams[] = {getIdentTyPointerTy(), CGM.Int32Ty,
-                                getKmpc_MicroPointerTy()};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.VoidTy, TypeParams, /*isVarArg*/ true);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, "__kmpc_fork_call");
-    if (auto *F = dyn_cast<llvm::Function>(RTLFn.getCallee())) {
-      if (!F->hasMetadata(llvm::LLVMContext::MD_callback)) {
-        llvm::LLVMContext &Ctx = F->getContext();
-        llvm::MDBuilder MDB(Ctx);
-        // Annotate the callback behavior of the __kmpc_fork_call:
-        //  - The callback callee is argument number 2 (microtask).
-        //  - The first two arguments of the callback callee are unknown (-1).
-        //  - All variadic arguments to the __kmpc_fork_call are passed to the
-        //    callback callee.
-        F->addMetadata(
-            llvm::LLVMContext::MD_callback,
-            *llvm::MDNode::get(Ctx, {MDB.createCallbackEncoding(
-                                        2, {-1, -1},
-                                        /* VarArgsArePassed */ true)}));
-      }
-    }
-    break;
-  }
-  case OMPRTL__kmpc_global_thread_num: {
-    // Build kmp_int32 __kmpc_global_thread_num(ident_t *loc);
-    llvm::Type *TypeParams[] = {getIdentTyPointerTy()};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.Int32Ty, TypeParams, /*isVarArg*/ false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, "__kmpc_global_thread_num");
-    break;
-  }
-  case OMPRTL__kmpc_threadprivate_cached: {
-    // Build void *__kmpc_threadprivate_cached(ident_t *loc,
-    // kmp_int32 global_tid, void *data, size_t size, void ***cache);
-    llvm::Type *TypeParams[] = {getIdentTyPointerTy(), CGM.Int32Ty,
-                                CGM.VoidPtrTy, CGM.SizeTy,
-                                CGM.VoidPtrTy->getPointerTo()->getPointerTo()};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.VoidPtrTy, TypeParams, /*isVarArg*/ false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, "__kmpc_threadprivate_cached");
-    break;
-  }
-  case OMPRTL__kmpc_critical: {
-    // Build void __kmpc_critical(ident_t *loc, kmp_int32 global_tid,
-    // kmp_critical_name *crit);
-    llvm::Type *TypeParams[] = {
-        getIdentTyPointerTy(), CGM.Int32Ty,
-        llvm::PointerType::getUnqual(KmpCriticalNameTy)};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.VoidTy, TypeParams, /*isVarArg*/ false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, "__kmpc_critical");
-    break;
-  }
-  case OMPRTL__kmpc_critical_with_hint: {
-    // Build void __kmpc_critical_with_hint(ident_t *loc, kmp_int32 global_tid,
-    // kmp_critical_name *crit, uintptr_t hint);
-    llvm::Type *TypeParams[] = {getIdentTyPointerTy(), CGM.Int32Ty,
-                                llvm::PointerType::getUnqual(KmpCriticalNameTy),
-                                CGM.IntPtrTy};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.VoidTy, TypeParams, /*isVarArg*/ false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, "__kmpc_critical_with_hint");
-    break;
-  }
-  case OMPRTL__kmpc_threadprivate_register: {
-    // Build void __kmpc_threadprivate_register(ident_t *, void *data,
-    // kmpc_ctor ctor, kmpc_cctor cctor, kmpc_dtor dtor);
-    // typedef void *(*kmpc_ctor)(void *);
-    auto *KmpcCtorTy =
-        llvm::FunctionType::get(CGM.VoidPtrTy, CGM.VoidPtrTy,
-                                /*isVarArg*/ false)->getPointerTo();
-    // typedef void *(*kmpc_cctor)(void *, void *);
-    llvm::Type *KmpcCopyCtorTyArgs[] = {CGM.VoidPtrTy, CGM.VoidPtrTy};
-    auto *KmpcCopyCtorTy =
-        llvm::FunctionType::get(CGM.VoidPtrTy, KmpcCopyCtorTyArgs,
-                                /*isVarArg*/ false)
-            ->getPointerTo();
-    // typedef void (*kmpc_dtor)(void *);
-    auto *KmpcDtorTy =
-        llvm::FunctionType::get(CGM.VoidTy, CGM.VoidPtrTy, /*isVarArg*/ false)
-            ->getPointerTo();
-    llvm::Type *FnTyArgs[] = {getIdentTyPointerTy(), CGM.VoidPtrTy, KmpcCtorTy,
-                              KmpcCopyCtorTy, KmpcDtorTy};
-    auto *FnTy = llvm::FunctionType::get(CGM.VoidTy, FnTyArgs,
-                                        /*isVarArg*/ false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, "__kmpc_threadprivate_register");
-    break;
-  }
-  case OMPRTL__kmpc_end_critical: {
-    // Build void __kmpc_end_critical(ident_t *loc, kmp_int32 global_tid,
-    // kmp_critical_name *crit);
-    llvm::Type *TypeParams[] = {
-        getIdentTyPointerTy(), CGM.Int32Ty,
-        llvm::PointerType::getUnqual(KmpCriticalNameTy)};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.VoidTy, TypeParams, /*isVarArg*/ false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, "__kmpc_end_critical");
-    break;
-  }
-  case OMPRTL__kmpc_cancel_barrier: {
-    // Build kmp_int32 __kmpc_cancel_barrier(ident_t *loc, kmp_int32
-    // global_tid);
-    llvm::Type *TypeParams[] = {getIdentTyPointerTy(), CGM.Int32Ty};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.Int32Ty, TypeParams, /*isVarArg*/ false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, /*Name*/ "__kmpc_cancel_barrier");
-    break;
-  }
-  case OMPRTL__kmpc_barrier: {
-    // Build void __kmpc_barrier(ident_t *loc, kmp_int32 global_tid);
-    llvm::Type *TypeParams[] = {getIdentTyPointerTy(), CGM.Int32Ty};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.VoidTy, TypeParams, /*isVarArg*/ false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, /*Name*/ "__kmpc_barrier");
-    break;
-  }
-  case OMPRTL__kmpc_for_static_fini: {
-    // Build void __kmpc_for_static_fini(ident_t *loc, kmp_int32 global_tid);
-    llvm::Type *TypeParams[] = {getIdentTyPointerTy(), CGM.Int32Ty};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.VoidTy, TypeParams, /*isVarArg*/ false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, "__kmpc_for_static_fini");
-    break;
-  }
-  case OMPRTL__kmpc_push_num_threads: {
-    // Build void __kmpc_push_num_threads(ident_t *loc, kmp_int32 global_tid,
-    // kmp_int32 num_threads)
-    llvm::Type *TypeParams[] = {getIdentTyPointerTy(), CGM.Int32Ty,
-                                CGM.Int32Ty};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.VoidTy, TypeParams, /*isVarArg*/ false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, "__kmpc_push_num_threads");
-    break;
-  }
-  case OMPRTL__kmpc_serialized_parallel: {
-    // Build void __kmpc_serialized_parallel(ident_t *loc, kmp_int32
-    // global_tid);
-    llvm::Type *TypeParams[] = {getIdentTyPointerTy(), CGM.Int32Ty};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.VoidTy, TypeParams, /*isVarArg*/ false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, "__kmpc_serialized_parallel");
-    break;
-  }
-  case OMPRTL__kmpc_end_serialized_parallel: {
-    // Build void __kmpc_end_serialized_parallel(ident_t *loc, kmp_int32
-    // global_tid);
-    llvm::Type *TypeParams[] = {getIdentTyPointerTy(), CGM.Int32Ty};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.VoidTy, TypeParams, /*isVarArg*/ false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, "__kmpc_end_serialized_parallel");
-    break;
-  }
-  case OMPRTL__kmpc_flush: {
-    // Build void __kmpc_flush(ident_t *loc);
-    llvm::Type *TypeParams[] = {getIdentTyPointerTy()};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.VoidTy, TypeParams, /*isVarArg*/ false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, "__kmpc_flush");
-    break;
-  }
-  case OMPRTL__kmpc_master: {
-    // Build kmp_int32 __kmpc_master(ident_t *loc, kmp_int32 global_tid);
-    llvm::Type *TypeParams[] = {getIdentTyPointerTy(), CGM.Int32Ty};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.Int32Ty, TypeParams, /*isVarArg=*/false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, /*Name=*/"__kmpc_master");
-    break;
-  }
-  case OMPRTL__kmpc_end_master: {
-    // Build void __kmpc_end_master(ident_t *loc, kmp_int32 global_tid);
-    llvm::Type *TypeParams[] = {getIdentTyPointerTy(), CGM.Int32Ty};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.VoidTy, TypeParams, /*isVarArg=*/false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, /*Name=*/"__kmpc_end_master");
-    break;
-  }
-  case OMPRTL__kmpc_omp_taskyield: {
-    // Build kmp_int32 __kmpc_omp_taskyield(ident_t *, kmp_int32 global_tid,
-    // int end_part);
-    llvm::Type *TypeParams[] = {getIdentTyPointerTy(), CGM.Int32Ty, CGM.IntTy};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.Int32Ty, TypeParams, /*isVarArg=*/false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, /*Name=*/"__kmpc_omp_taskyield");
-    break;
-  }
-  case OMPRTL__kmpc_single: {
-    // Build kmp_int32 __kmpc_single(ident_t *loc, kmp_int32 global_tid);
-    llvm::Type *TypeParams[] = {getIdentTyPointerTy(), CGM.Int32Ty};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.Int32Ty, TypeParams, /*isVarArg=*/false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, /*Name=*/"__kmpc_single");
-    break;
-  }
-  case OMPRTL__kmpc_end_single: {
-    // Build void __kmpc_end_single(ident_t *loc, kmp_int32 global_tid);
-    llvm::Type *TypeParams[] = {getIdentTyPointerTy(), CGM.Int32Ty};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.VoidTy, TypeParams, /*isVarArg=*/false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, /*Name=*/"__kmpc_end_single");
-    break;
-  }
-  case OMPRTL__kmpc_omp_task_alloc: {
-    // Build kmp_task_t *__kmpc_omp_task_alloc(ident_t *, kmp_int32 gtid,
-    // kmp_int32 flags, size_t sizeof_kmp_task_t, size_t sizeof_shareds,
-    // kmp_routine_entry_t *task_entry);
-    assert(KmpRoutineEntryPtrTy != nullptr &&
-           "Type kmp_routine_entry_t must be created.");
-    llvm::Type *TypeParams[] = {getIdentTyPointerTy(), CGM.Int32Ty, CGM.Int32Ty,
-                                CGM.SizeTy, CGM.SizeTy, KmpRoutineEntryPtrTy};
-    // Return void * and then cast to particular kmp_task_t type.
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.VoidPtrTy, TypeParams, /*isVarArg=*/false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, /*Name=*/"__kmpc_omp_task_alloc");
-    break;
-  }
-  case OMPRTL__kmpc_omp_target_task_alloc: {
-    // Build kmp_task_t *__kmpc_omp_target_task_alloc(ident_t *, kmp_int32 gtid,
-    // kmp_int32 flags, size_t sizeof_kmp_task_t, size_t sizeof_shareds,
-    // kmp_routine_entry_t *task_entry, kmp_int64 device_id);
-    assert(KmpRoutineEntryPtrTy != nullptr &&
-           "Type kmp_routine_entry_t must be created.");
-    llvm::Type *TypeParams[] = {getIdentTyPointerTy(), CGM.Int32Ty, CGM.Int32Ty,
-                                CGM.SizeTy, CGM.SizeTy, KmpRoutineEntryPtrTy,
-                                CGM.Int64Ty};
-    // Return void * and then cast to particular kmp_task_t type.
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.VoidPtrTy, TypeParams, /*isVarArg=*/false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, /*Name=*/"__kmpc_omp_target_task_alloc");
-    break;
-  }
-  case OMPRTL__kmpc_omp_task: {
-    // Build kmp_int32 __kmpc_omp_task(ident_t *, kmp_int32 gtid, kmp_task_t
-    // *new_task);
-    llvm::Type *TypeParams[] = {getIdentTyPointerTy(), CGM.Int32Ty,
-                                CGM.VoidPtrTy};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.Int32Ty, TypeParams, /*isVarArg=*/false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, /*Name=*/"__kmpc_omp_task");
-    break;
-  }
-  case OMPRTL__kmpc_copyprivate: {
-    // Build void __kmpc_copyprivate(ident_t *loc, kmp_int32 global_tid,
-    // size_t cpy_size, void *cpy_data, void(*cpy_func)(void *, void *),
-    // kmp_int32 didit);
-    llvm::Type *CpyTypeParams[] = {CGM.VoidPtrTy, CGM.VoidPtrTy};
-    auto *CpyFnTy =
-        llvm::FunctionType::get(CGM.VoidTy, CpyTypeParams, /*isVarArg=*/false);
-    llvm::Type *TypeParams[] = {getIdentTyPointerTy(), CGM.Int32Ty, CGM.SizeTy,
-                                CGM.VoidPtrTy, CpyFnTy->getPointerTo(),
-                                CGM.Int32Ty};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.VoidTy, TypeParams, /*isVarArg=*/false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, /*Name=*/"__kmpc_copyprivate");
-    break;
-  }
-  case OMPRTL__kmpc_reduce: {
-    // Build kmp_int32 __kmpc_reduce(ident_t *loc, kmp_int32 global_tid,
-    // kmp_int32 num_vars, size_t reduce_size, void *reduce_data, void
-    // (*reduce_func)(void *lhs_data, void *rhs_data), kmp_critical_name *lck);
-    llvm::Type *ReduceTypeParams[] = {CGM.VoidPtrTy, CGM.VoidPtrTy};
-    auto *ReduceFnTy = llvm::FunctionType::get(CGM.VoidTy, ReduceTypeParams,
-                                               /*isVarArg=*/false);
-    llvm::Type *TypeParams[] = {
-        getIdentTyPointerTy(), CGM.Int32Ty, CGM.Int32Ty, CGM.SizeTy,
-        CGM.VoidPtrTy, ReduceFnTy->getPointerTo(),
-        llvm::PointerType::getUnqual(KmpCriticalNameTy)};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.Int32Ty, TypeParams, /*isVarArg=*/false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, /*Name=*/"__kmpc_reduce");
-    break;
-  }
-  case OMPRTL__kmpc_reduce_nowait: {
-    // Build kmp_int32 __kmpc_reduce_nowait(ident_t *loc, kmp_int32
-    // global_tid, kmp_int32 num_vars, size_t reduce_size, void *reduce_data,
-    // void (*reduce_func)(void *lhs_data, void *rhs_data), kmp_critical_name
-    // *lck);
-    llvm::Type *ReduceTypeParams[] = {CGM.VoidPtrTy, CGM.VoidPtrTy};
-    auto *ReduceFnTy = llvm::FunctionType::get(CGM.VoidTy, ReduceTypeParams,
-                                               /*isVarArg=*/false);
-    llvm::Type *TypeParams[] = {
-        getIdentTyPointerTy(), CGM.Int32Ty, CGM.Int32Ty, CGM.SizeTy,
-        CGM.VoidPtrTy, ReduceFnTy->getPointerTo(),
-        llvm::PointerType::getUnqual(KmpCriticalNameTy)};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.Int32Ty, TypeParams, /*isVarArg=*/false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, /*Name=*/"__kmpc_reduce_nowait");
-    break;
-  }
-  case OMPRTL__kmpc_end_reduce: {
-    // Build void __kmpc_end_reduce(ident_t *loc, kmp_int32 global_tid,
-    // kmp_critical_name *lck);
-    llvm::Type *TypeParams[] = {
-        getIdentTyPointerTy(), CGM.Int32Ty,
-        llvm::PointerType::getUnqual(KmpCriticalNameTy)};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.VoidTy, TypeParams, /*isVarArg=*/false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, /*Name=*/"__kmpc_end_reduce");
-    break;
-  }
-  case OMPRTL__kmpc_end_reduce_nowait: {
-    // Build __kmpc_end_reduce_nowait(ident_t *loc, kmp_int32 global_tid,
-    // kmp_critical_name *lck);
-    llvm::Type *TypeParams[] = {
-        getIdentTyPointerTy(), CGM.Int32Ty,
-        llvm::PointerType::getUnqual(KmpCriticalNameTy)};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.VoidTy, TypeParams, /*isVarArg=*/false);
-    RTLFn =
-        CGM.CreateRuntimeFunction(FnTy, /*Name=*/"__kmpc_end_reduce_nowait");
-    break;
-  }
-  case OMPRTL__kmpc_omp_task_begin_if0: {
-    // Build void __kmpc_omp_task(ident_t *, kmp_int32 gtid, kmp_task_t
-    // *new_task);
-    llvm::Type *TypeParams[] = {getIdentTyPointerTy(), CGM.Int32Ty,
-                                CGM.VoidPtrTy};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.VoidTy, TypeParams, /*isVarArg=*/false);
-    RTLFn =
-        CGM.CreateRuntimeFunction(FnTy, /*Name=*/"__kmpc_omp_task_begin_if0");
-    break;
-  }
-  case OMPRTL__kmpc_omp_task_complete_if0: {
-    // Build void __kmpc_omp_task(ident_t *, kmp_int32 gtid, kmp_task_t
-    // *new_task);
-    llvm::Type *TypeParams[] = {getIdentTyPointerTy(), CGM.Int32Ty,
-                                CGM.VoidPtrTy};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.VoidTy, TypeParams, /*isVarArg=*/false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy,
-                                      /*Name=*/"__kmpc_omp_task_complete_if0");
-    break;
-  }
-  case OMPRTL__kmpc_ordered: {
-    // Build void __kmpc_ordered(ident_t *loc, kmp_int32 global_tid);
-    llvm::Type *TypeParams[] = {getIdentTyPointerTy(), CGM.Int32Ty};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.VoidTy, TypeParams, /*isVarArg=*/false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, "__kmpc_ordered");
-    break;
-  }
-  case OMPRTL__kmpc_end_ordered: {
-    // Build void __kmpc_end_ordered(ident_t *loc, kmp_int32 global_tid);
-    llvm::Type *TypeParams[] = {getIdentTyPointerTy(), CGM.Int32Ty};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.VoidTy, TypeParams, /*isVarArg=*/false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, "__kmpc_end_ordered");
-    break;
-  }
-  case OMPRTL__kmpc_omp_taskwait: {
-    // Build kmp_int32 __kmpc_omp_taskwait(ident_t *loc, kmp_int32 global_tid);
-    llvm::Type *TypeParams[] = {getIdentTyPointerTy(), CGM.Int32Ty};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.Int32Ty, TypeParams, /*isVarArg=*/false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, "__kmpc_omp_taskwait");
-    break;
-  }
-  case OMPRTL__kmpc_taskgroup: {
-    // Build void __kmpc_taskgroup(ident_t *loc, kmp_int32 global_tid);
-    llvm::Type *TypeParams[] = {getIdentTyPointerTy(), CGM.Int32Ty};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.VoidTy, TypeParams, /*isVarArg=*/false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, "__kmpc_taskgroup");
-    break;
-  }
-  case OMPRTL__kmpc_end_taskgroup: {
-    // Build void __kmpc_end_taskgroup(ident_t *loc, kmp_int32 global_tid);
-    llvm::Type *TypeParams[] = {getIdentTyPointerTy(), CGM.Int32Ty};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.VoidTy, TypeParams, /*isVarArg=*/false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, "__kmpc_end_taskgroup");
-    break;
-  }
-  case OMPRTL__kmpc_push_proc_bind: {
-    // Build void __kmpc_push_proc_bind(ident_t *loc, kmp_int32 global_tid,
-    // int proc_bind)
-    llvm::Type *TypeParams[] = {getIdentTyPointerTy(), CGM.Int32Ty, CGM.IntTy};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.VoidTy, TypeParams, /*isVarArg*/ false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, "__kmpc_push_proc_bind");
-    break;
-  }
-  case OMPRTL__kmpc_omp_task_with_deps: {
-    // Build kmp_int32 __kmpc_omp_task_with_deps(ident_t *, kmp_int32 gtid,
-    // kmp_task_t *new_task, kmp_int32 ndeps, kmp_depend_info_t *dep_list,
-    // kmp_int32 ndeps_noalias, kmp_depend_info_t *noalias_dep_list);
-    llvm::Type *TypeParams[] = {
-        getIdentTyPointerTy(), CGM.Int32Ty, CGM.VoidPtrTy, CGM.Int32Ty,
-        CGM.VoidPtrTy,         CGM.Int32Ty, CGM.VoidPtrTy};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.Int32Ty, TypeParams, /*isVarArg=*/false);
-    RTLFn =
-        CGM.CreateRuntimeFunction(FnTy, /*Name=*/"__kmpc_omp_task_with_deps");
-    break;
-  }
-  case OMPRTL__kmpc_omp_wait_deps: {
-    // Build void __kmpc_omp_wait_deps(ident_t *, kmp_int32 gtid,
-    // kmp_int32 ndeps, kmp_depend_info_t *dep_list, kmp_int32 ndeps_noalias,
-    // kmp_depend_info_t *noalias_dep_list);
-    llvm::Type *TypeParams[] = {getIdentTyPointerTy(), CGM.Int32Ty,
-                                CGM.Int32Ty,           CGM.VoidPtrTy,
-                                CGM.Int32Ty,           CGM.VoidPtrTy};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.VoidTy, TypeParams, /*isVarArg=*/false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, /*Name=*/"__kmpc_omp_wait_deps");
-    break;
-  }
-  case OMPRTL__kmpc_cancellationpoint: {
-    // Build kmp_int32 __kmpc_cancellationpoint(ident_t *loc, kmp_int32
-    // global_tid, kmp_int32 cncl_kind)
-    llvm::Type *TypeParams[] = {getIdentTyPointerTy(), CGM.Int32Ty, CGM.IntTy};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.Int32Ty, TypeParams, /*isVarArg*/ false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, "__kmpc_cancellationpoint");
-    break;
-  }
-  case OMPRTL__kmpc_cancel: {
-    // Build kmp_int32 __kmpc_cancel(ident_t *loc, kmp_int32 global_tid,
-    // kmp_int32 cncl_kind)
-    llvm::Type *TypeParams[] = {getIdentTyPointerTy(), CGM.Int32Ty, CGM.IntTy};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.Int32Ty, TypeParams, /*isVarArg*/ false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, "__kmpc_cancel");
-    break;
-  }
-  case OMPRTL__kmpc_push_num_teams: {
-    // Build void kmpc_push_num_teams (ident_t loc, kmp_int32 global_tid,
-    // kmp_int32 num_teams, kmp_int32 num_threads)
-    llvm::Type *TypeParams[] = {getIdentTyPointerTy(), CGM.Int32Ty, CGM.Int32Ty,
-        CGM.Int32Ty};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.Int32Ty, TypeParams, /*isVarArg*/ false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, "__kmpc_push_num_teams");
-    break;
-  }
-  case OMPRTL__kmpc_fork_teams: {
-    // Build void __kmpc_fork_teams(ident_t *loc, kmp_int32 argc, kmpc_micro
-    // microtask, ...);
-    llvm::Type *TypeParams[] = {getIdentTyPointerTy(), CGM.Int32Ty,
-                                getKmpc_MicroPointerTy()};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.VoidTy, TypeParams, /*isVarArg*/ true);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, "__kmpc_fork_teams");
-    if (auto *F = dyn_cast<llvm::Function>(RTLFn.getCallee())) {
-      if (!F->hasMetadata(llvm::LLVMContext::MD_callback)) {
-        llvm::LLVMContext &Ctx = F->getContext();
-        llvm::MDBuilder MDB(Ctx);
-        // Annotate the callback behavior of the __kmpc_fork_teams:
-        //  - The callback callee is argument number 2 (microtask).
-        //  - The first two arguments of the callback callee are unknown (-1).
-        //  - All variadic arguments to the __kmpc_fork_teams are passed to the
-        //    callback callee.
-        F->addMetadata(
-            llvm::LLVMContext::MD_callback,
-            *llvm::MDNode::get(Ctx, {MDB.createCallbackEncoding(
-                                        2, {-1, -1},
-                                        /* VarArgsArePassed */ true)}));
-      }
-    }
-    break;
-  }
-  case OMPRTL__kmpc_taskloop: {
-    // Build void __kmpc_taskloop(ident_t *loc, int gtid, kmp_task_t *task, int
-    // if_val, kmp_uint64 *lb, kmp_uint64 *ub, kmp_int64 st, int nogroup, int
-    // sched, kmp_uint64 grainsize, void *task_dup);
-    llvm::Type *TypeParams[] = {getIdentTyPointerTy(),
-                                CGM.IntTy,
-                                CGM.VoidPtrTy,
-                                CGM.IntTy,
-                                CGM.Int64Ty->getPointerTo(),
-                                CGM.Int64Ty->getPointerTo(),
-                                CGM.Int64Ty,
-                                CGM.IntTy,
-                                CGM.IntTy,
-                                CGM.Int64Ty,
-                                CGM.VoidPtrTy};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.VoidTy, TypeParams, /*isVarArg=*/false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, /*Name=*/"__kmpc_taskloop");
-    break;
-  }
-  case OMPRTL__kmpc_doacross_init: {
-    // Build void __kmpc_doacross_init(ident_t *loc, kmp_int32 gtid, kmp_int32
-    // num_dims, struct kmp_dim *dims);
-    llvm::Type *TypeParams[] = {getIdentTyPointerTy(),
-                                CGM.Int32Ty,
-                                CGM.Int32Ty,
-                                CGM.VoidPtrTy};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.VoidTy, TypeParams, /*isVarArg=*/false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, /*Name=*/"__kmpc_doacross_init");
-    break;
-  }
-  case OMPRTL__kmpc_doacross_fini: {
-    // Build void __kmpc_doacross_fini(ident_t *loc, kmp_int32 gtid);
-    llvm::Type *TypeParams[] = {getIdentTyPointerTy(), CGM.Int32Ty};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.VoidTy, TypeParams, /*isVarArg=*/false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, /*Name=*/"__kmpc_doacross_fini");
-    break;
-  }
-  case OMPRTL__kmpc_doacross_post: {
-    // Build void __kmpc_doacross_post(ident_t *loc, kmp_int32 gtid, kmp_int64
-    // *vec);
-    llvm::Type *TypeParams[] = {getIdentTyPointerTy(), CGM.Int32Ty,
-                                CGM.Int64Ty->getPointerTo()};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.VoidTy, TypeParams, /*isVarArg=*/false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, /*Name=*/"__kmpc_doacross_post");
-    break;
-  }
-  case OMPRTL__kmpc_doacross_wait: {
-    // Build void __kmpc_doacross_wait(ident_t *loc, kmp_int32 gtid, kmp_int64
-    // *vec);
-    llvm::Type *TypeParams[] = {getIdentTyPointerTy(), CGM.Int32Ty,
-                                CGM.Int64Ty->getPointerTo()};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.VoidTy, TypeParams, /*isVarArg=*/false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, /*Name=*/"__kmpc_doacross_wait");
-    break;
-  }
-  case OMPRTL__kmpc_taskred_init: {
-    // Build void *__kmpc_taskred_init(int gtid, int num_data, void *data);
-    llvm::Type *TypeParams[] = {CGM.IntTy, CGM.IntTy, CGM.VoidPtrTy};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.VoidPtrTy, TypeParams, /*isVarArg=*/false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, /*Name=*/"__kmpc_taskred_init");
-    break;
-  }
-  case OMPRTL__kmpc_task_reduction_get_th_data: {
-    // Build void *__kmpc_task_reduction_get_th_data(int gtid, void *tg, void
-    // *d);
-    llvm::Type *TypeParams[] = {CGM.IntTy, CGM.VoidPtrTy, CGM.VoidPtrTy};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.VoidPtrTy, TypeParams, /*isVarArg=*/false);
-    RTLFn = CGM.CreateRuntimeFunction(
-        FnTy, /*Name=*/"__kmpc_task_reduction_get_th_data");
-    break;
-  }
-  case OMPRTL__kmpc_taskred_modifier_init: {
-    // Build void *__kmpc_taskred_modifier_init(ident_t *loc, int gtid, int
-    // is_ws, int num_data, void *data);
-    llvm::Type *TypeParams[] = {getIdentTyPointerTy(), CGM.IntTy, CGM.IntTy,
-                                CGM.IntTy, CGM.VoidPtrTy};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.VoidPtrTy, TypeParams, /*isVarArg=*/false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy,
-                                      /*Name=*/"__kmpc_taskred_modifier_init");
-    break;
-  }
-  case OMPRTL__kmpc_task_reduction_modifier_fini: {
-    // Build void __kmpc_task_reduction_modifier_fini(ident_t *loc, int gtid,
-    // int is_ws);
-    llvm::Type *TypeParams[] = {getIdentTyPointerTy(), CGM.IntTy, CGM.IntTy};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.VoidTy, TypeParams, /*isVarArg=*/false);
-    RTLFn = CGM.CreateRuntimeFunction(
-        FnTy,
-        /*Name=*/"__kmpc_task_reduction_modifier_fini");
-    break;
-  }
-  case OMPRTL__kmpc_alloc: {
-    // Build to void *__kmpc_alloc(int gtid, size_t sz, omp_allocator_handle_t
-    // al); omp_allocator_handle_t type is void *.
-    llvm::Type *TypeParams[] = {CGM.IntTy, CGM.SizeTy, CGM.VoidPtrTy};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.VoidPtrTy, TypeParams, /*isVarArg=*/false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, /*Name=*/"__kmpc_alloc");
-    break;
-  }
-  case OMPRTL__kmpc_free: {
-    // Build to void __kmpc_free(int gtid, void *ptr, omp_allocator_handle_t
-    // al); omp_allocator_handle_t type is void *.
-    llvm::Type *TypeParams[] = {CGM.IntTy, CGM.VoidPtrTy, CGM.VoidPtrTy};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.VoidTy, TypeParams, /*isVarArg=*/false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, /*Name=*/"__kmpc_free");
-    break;
-  }
-  case OMPRTL__kmpc_push_target_tripcount: {
-    // Build void __kmpc_push_target_tripcount(int64_t device_id, kmp_uint64
-    // size);
-    llvm::Type *TypeParams[] = {CGM.Int64Ty, CGM.Int64Ty};
-    llvm::FunctionType *FnTy =
-        llvm::FunctionType::get(CGM.VoidTy, TypeParams, /*isVarArg=*/false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, "__kmpc_push_target_tripcount");
-    break;
-  }
-  case OMPRTL__tgt_target: {
-    // Build int32_t __tgt_target(int64_t device_id, void *host_ptr, int32_t
-    // arg_num, void** args_base, void **args, int64_t *arg_sizes, int64_t
-    // *arg_types);
-    llvm::Type *TypeParams[] = {CGM.Int64Ty,
-                                CGM.VoidPtrTy,
-                                CGM.Int32Ty,
-                                CGM.VoidPtrPtrTy,
-                                CGM.VoidPtrPtrTy,
-                                CGM.Int64Ty->getPointerTo(),
-                                CGM.Int64Ty->getPointerTo()};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.Int32Ty, TypeParams, /*isVarArg*/ false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, "__tgt_target");
-    break;
-  }
-  case OMPRTL__tgt_target_nowait: {
-    // Build int32_t __tgt_target_nowait(int64_t device_id, void *host_ptr,
-    // int32_t arg_num, void** args_base, void **args, int64_t *arg_sizes,
-    // int64_t *arg_types);
-    llvm::Type *TypeParams[] = {CGM.Int64Ty,
-                                CGM.VoidPtrTy,
-                                CGM.Int32Ty,
-                                CGM.VoidPtrPtrTy,
-                                CGM.VoidPtrPtrTy,
-                                CGM.Int64Ty->getPointerTo(),
-                                CGM.Int64Ty->getPointerTo()};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.Int32Ty, TypeParams, /*isVarArg*/ false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, "__tgt_target_nowait");
-    break;
-  }
-  case OMPRTL__tgt_target_teams: {
-    // Build int32_t __tgt_target_teams(int64_t device_id, void *host_ptr,
-    // int32_t arg_num, void** args_base, void **args, int64_t *arg_sizes,
-    // int64_t *arg_types, int32_t num_teams, int32_t thread_limit);
-    llvm::Type *TypeParams[] = {CGM.Int64Ty,
-                                CGM.VoidPtrTy,
-                                CGM.Int32Ty,
-                                CGM.VoidPtrPtrTy,
-                                CGM.VoidPtrPtrTy,
-                                CGM.Int64Ty->getPointerTo(),
-                                CGM.Int64Ty->getPointerTo(),
-                                CGM.Int32Ty,
-                                CGM.Int32Ty};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.Int32Ty, TypeParams, /*isVarArg*/ false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, "__tgt_target_teams");
-    break;
-  }
-  case OMPRTL__tgt_target_teams_nowait: {
-    // Build int32_t __tgt_target_teams_nowait(int64_t device_id, void
-    // *host_ptr, int32_t arg_num, void** args_base, void **args, int64_t
-    // *arg_sizes, int64_t *arg_types, int32_t num_teams, int32_t thread_limit);
-    llvm::Type *TypeParams[] = {CGM.Int64Ty,
-                                CGM.VoidPtrTy,
-                                CGM.Int32Ty,
-                                CGM.VoidPtrPtrTy,
-                                CGM.VoidPtrPtrTy,
-                                CGM.Int64Ty->getPointerTo(),
-                                CGM.Int64Ty->getPointerTo(),
-                                CGM.Int32Ty,
-                                CGM.Int32Ty};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.Int32Ty, TypeParams, /*isVarArg*/ false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, "__tgt_target_teams_nowait");
-    break;
-  }
-  case OMPRTL__tgt_register_requires: {
-    // Build void __tgt_register_requires(int64_t flags);
-    llvm::Type *TypeParams[] = {CGM.Int64Ty};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.VoidTy, TypeParams, /*isVarArg*/ false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, "__tgt_register_requires");
-    break;
-  }
-  case OMPRTL__tgt_target_data_begin: {
-    // Build void __tgt_target_data_begin(int64_t device_id, int32_t arg_num,
-    // void** args_base, void **args, int64_t *arg_sizes, int64_t *arg_types);
-    llvm::Type *TypeParams[] = {CGM.Int64Ty,
-                                CGM.Int32Ty,
-                                CGM.VoidPtrPtrTy,
-                                CGM.VoidPtrPtrTy,
-                                CGM.Int64Ty->getPointerTo(),
-                                CGM.Int64Ty->getPointerTo()};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.VoidTy, TypeParams, /*isVarArg*/ false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, "__tgt_target_data_begin");
-    break;
-  }
-  case OMPRTL__tgt_target_data_begin_nowait: {
-    // Build void __tgt_target_data_begin_nowait(int64_t device_id, int32_t
-    // arg_num, void** args_base, void **args, int64_t *arg_sizes, int64_t
-    // *arg_types);
-    llvm::Type *TypeParams[] = {CGM.Int64Ty,
-                                CGM.Int32Ty,
-                                CGM.VoidPtrPtrTy,
-                                CGM.VoidPtrPtrTy,
-                                CGM.Int64Ty->getPointerTo(),
-                                CGM.Int64Ty->getPointerTo()};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.VoidTy, TypeParams, /*isVarArg=*/false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, "__tgt_target_data_begin_nowait");
-    break;
-  }
-  case OMPRTL__tgt_target_data_end: {
-    // Build void __tgt_target_data_end(int64_t device_id, int32_t arg_num,
-    // void** args_base, void **args, int64_t *arg_sizes, int64_t *arg_types);
-    llvm::Type *TypeParams[] = {CGM.Int64Ty,
-                                CGM.Int32Ty,
-                                CGM.VoidPtrPtrTy,
-                                CGM.VoidPtrPtrTy,
-                                CGM.Int64Ty->getPointerTo(),
-                                CGM.Int64Ty->getPointerTo()};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.VoidTy, TypeParams, /*isVarArg*/ false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, "__tgt_target_data_end");
-    break;
-  }
-  case OMPRTL__tgt_target_data_end_nowait: {
-    // Build void __tgt_target_data_end_nowait(int64_t device_id, int32_t
-    // arg_num, void** args_base, void **args, int64_t *arg_sizes, int64_t
-    // *arg_types);
-    llvm::Type *TypeParams[] = {CGM.Int64Ty,
-                                CGM.Int32Ty,
-                                CGM.VoidPtrPtrTy,
-                                CGM.VoidPtrPtrTy,
-                                CGM.Int64Ty->getPointerTo(),
-                                CGM.Int64Ty->getPointerTo()};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.VoidTy, TypeParams, /*isVarArg=*/false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, "__tgt_target_data_end_nowait");
-    break;
-  }
-  case OMPRTL__tgt_target_data_update: {
-    // Build void __tgt_target_data_update(int64_t device_id, int32_t arg_num,
-    // void** args_base, void **args, int64_t *arg_sizes, int64_t *arg_types);
-    llvm::Type *TypeParams[] = {CGM.Int64Ty,
-                                CGM.Int32Ty,
-                                CGM.VoidPtrPtrTy,
-                                CGM.VoidPtrPtrTy,
-                                CGM.Int64Ty->getPointerTo(),
-                                CGM.Int64Ty->getPointerTo()};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.VoidTy, TypeParams, /*isVarArg*/ false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, "__tgt_target_data_update");
-    break;
-  }
-  case OMPRTL__tgt_target_data_update_nowait: {
-    // Build void __tgt_target_data_update_nowait(int64_t device_id, int32_t
-    // arg_num, void** args_base, void **args, int64_t *arg_sizes, int64_t
-    // *arg_types);
-    llvm::Type *TypeParams[] = {CGM.Int64Ty,
-                                CGM.Int32Ty,
-                                CGM.VoidPtrPtrTy,
-                                CGM.VoidPtrPtrTy,
-                                CGM.Int64Ty->getPointerTo(),
-                                CGM.Int64Ty->getPointerTo()};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.VoidTy, TypeParams, /*isVarArg=*/false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, "__tgt_target_data_update_nowait");
-    break;
-  }
-  case OMPRTL__tgt_mapper_num_components: {
-    // Build int64_t __tgt_mapper_num_components(void *rt_mapper_handle);
-    llvm::Type *TypeParams[] = {CGM.VoidPtrTy};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.Int64Ty, TypeParams, /*isVarArg*/ false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, "__tgt_mapper_num_components");
-    break;
-  }
-  case OMPRTL__tgt_push_mapper_component: {
-    // Build void __tgt_push_mapper_component(void *rt_mapper_handle, void
-    // *base, void *begin, int64_t size, int64_t type);
-    llvm::Type *TypeParams[] = {CGM.VoidPtrTy, CGM.VoidPtrTy, CGM.VoidPtrTy,
-                                CGM.Int64Ty, CGM.Int64Ty};
-    auto *FnTy =
-        llvm::FunctionType::get(CGM.VoidTy, TypeParams, /*isVarArg*/ false);
-    RTLFn = CGM.CreateRuntimeFunction(FnTy, "__tgt_push_mapper_component");
-    break;
-  }
-  case OMPRTL__kmpc_task_allow_completion_event: {
-    // Build kmp_event_t *__kmpc_task_allow_completion_event(ident_t *loc_ref,
-    // int gtid, kmp_task_t *task);
-    auto *FnTy = llvm::FunctionType::get(
-        CGM.VoidPtrTy, {getIdentTyPointerTy(), CGM.IntTy, CGM.VoidPtrTy},
-        /*isVarArg=*/false);
-    RTLFn =
-        CGM.CreateRuntimeFunction(FnTy, "__kmpc_task_allow_completion_event");
-    break;
-  }
-  }
-  assert(RTLFn && "Unable to find OpenMP runtime function");
-  return RTLFn;
-}
-
 llvm::FunctionCallee
 CGOpenMPRuntime::createForStaticInitFunction(unsigned IVSize, bool IVSigned) {
   assert((IVSize == 32 || IVSize == 64) &&
          "IV size is not compatible with the omp runtime");
   StringRef Name = IVSize == 32 ? (IVSigned ? "__kmpc_for_static_init_4"
                                             : "__kmpc_for_static_init_4u")
                                 : (IVSigned ? "__kmpc_for_static_init_8"
                                             : "__kmpc_for_static_init_8u");
Address CGOpenMPRuntime::getAddrOfThreadPrivate(CodeGenFunction &CGF,
 
   llvm::Type *VarTy = VDAddr.getElementType();
   llvm::Value *Args[] = {emitUpdateLocation(CGF, Loc), getThreadID(CGF, Loc),
                          CGF.Builder.CreatePointerCast(VDAddr.getPointer(),
                                                        CGM.Int8PtrTy),
                          CGM.getSize(CGM.GetTargetTypeStoreSize(VarTy)),
                          getOrCreateThreadPrivateCache(VD)};
   return Address(CGF.EmitRuntimeCall(
-      createRuntimeFunction(OMPRTL__kmpc_threadprivate_cached), Args),
+                     llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+                         CGM.getModule(), OMPRTL___kmpc_threadprivate_cached),
+                     Args),
                  VDAddr.getAlignment());
 }
 
 void CGOpenMPRuntime::emitThreadPrivateVarInit(
     CodeGenFunction &CGF, Address VDAddr, llvm::Value *Ctor,
     llvm::Value *CopyCtor, llvm::Value *Dtor, SourceLocation Loc) {
   // Call kmp_int32 __kmpc_global_thread_num(&loc) to init OpenMP runtime
   // library.
   llvm::Value *OMPLoc = emitUpdateLocation(CGF, Loc);
-  CGF.EmitRuntimeCall(createRuntimeFunction(OMPRTL__kmpc_global_thread_num),
+  CGF.EmitRuntimeCall(llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+                          CGM.getModule(), OMPRTL___kmpc_global_thread_num),
                       OMPLoc);
   // Call __kmpc_threadprivate_register(&loc, &var, ctor, cctor/*NULL*/, dtor)
   // to register constructor/destructor for variable.
   llvm::Value *Args[] = {
       OMPLoc, CGF.Builder.CreatePointerCast(VDAddr.getPointer(), CGM.VoidPtrTy),
       Ctor, CopyCtor, Dtor};
   CGF.EmitRuntimeCall(
-      createRuntimeFunction(OMPRTL__kmpc_threadprivate_register), Args);
+      llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+          CGM.getModule(), OMPRTL___kmpc_threadprivate_register),
+      Args);
 }
 
 llvm::Function *CGOpenMPRuntime::emitThreadPrivateVarDefinition(
     const VarDecl *VD, Address VDAddr, SourceLocation Loc,
     bool PerformInit, CodeGenFunction *CGF) {
   if (CGM.getLangOpts().OpenMPUseTLS &&
       CGM.getContext().getTargetInfo().isTLSSupported())
     return nullptr;
llvm::Value *Args[] = {
       CGF.Builder.CreatePointerBitCastOrAddrSpaceCast(GAddr, CGM.VoidPtrTy),
       CGF.Builder.CreateIntCast(CGF.getTypeSize(VarType), CGM.SizeTy,
                                 /*isSigned=*/false),
       getOrCreateInternalVariable(
           CGM.VoidPtrPtrTy, Twine(Name).concat(Suffix).concat(CacheSuffix))};
   return Address(
       CGF.Builder.CreatePointerBitCastOrAddrSpaceCast(
           CGF.EmitRuntimeCall(
-              createRuntimeFunction(OMPRTL__kmpc_threadprivate_cached), Args),
+              llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+                  CGM.getModule(), OMPRTL___kmpc_threadprivate_cached),
+              Args),
           VarLVType->getPointerTo(/*AddrSpace=*/0)),
       CGM.getContext().getTypeAlignInChars(VarType));
 }
 
 void CGOpenMPRuntime::emitIfClause(CodeGenFunction &CGF, const Expr *Cond,
                                    const RegionCodeGenTy &ThenGen,
                                    const RegionCodeGenTy &ElseGen) {
   CodeGenFunction::LexicalScope ConditionScope(CGF, Cond->getSourceRange());

 
 void CGOpenMPRuntime::emitParallelCall(CodeGenFunction &CGF, SourceLocation Loc,
                                        llvm::Function *OutlinedFn,
                                        ArrayRef<llvm::Value *> CapturedVars,
                                        const Expr *IfCond) {
   if (!CGF.HaveInsertPoint())
     return;
   llvm::Value *RTLoc = emitUpdateLocation(CGF, Loc);
-  auto &&ThenGen = [OutlinedFn, CapturedVars, RTLoc](CodeGenFunction &CGF,
+  auto &M = CGM.getModule();
+  auto &&ThenGen = [&M, OutlinedFn, CapturedVars, RTLoc](CodeGenFunction &CGF,
                                                          PrePostActionTy &) {
     // Build call __kmpc_fork_call(loc, n, microtask, var1, .., varn);
     CGOpenMPRuntime &RT = CGF.CGM.getOpenMPRuntime();
     llvm::Value *Args[] = {
         RTLoc,
         CGF.Builder.getInt32(CapturedVars.size()), // Number of captured vars
         CGF.Builder.CreateBitCast(OutlinedFn, RT.getKmpc_MicroPointerTy())};
     llvm::SmallVector<llvm::Value *, 16> RealArgs;
     RealArgs.append(std::begin(Args), std::end(Args));
     RealArgs.append(CapturedVars.begin(), CapturedVars.end());
 
     llvm::FunctionCallee RTLFn =
-        RT.createRuntimeFunction(OMPRTL__kmpc_fork_call);
+        llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+            M, OMPRTL___kmpc_fork_call);
     CGF.EmitRuntimeCall(RTLFn, RealArgs);
   };
-  auto &&ElseGen = [OutlinedFn, CapturedVars, RTLoc, Loc](CodeGenFunction &CGF,
-                                                          PrePostActionTy &) {
+  auto &&ElseGen = [&M, OutlinedFn, CapturedVars, RTLoc,
+                    Loc](CodeGenFunction &CGF, PrePostActionTy &) {
     CGOpenMPRuntime &RT = CGF.CGM.getOpenMPRuntime();
     llvm::Value *ThreadID = RT.getThreadID(CGF, Loc);
     // Build calls:
     // __kmpc_serialized_parallel(&Loc, GTid);
     llvm::Value *Args[] = {RTLoc, ThreadID};
-    CGF.EmitRuntimeCall(
-        RT.createRuntimeFunction(OMPRTL__kmpc_serialized_parallel), Args);
+    CGF.EmitRuntimeCall(llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+                            M, OMPRTL___kmpc_serialized_parallel),
+                        Args);
 
     // OutlinedFn(&GTid, &zero_bound, CapturedStruct);
     Address ThreadIDAddr = RT.emitThreadIDAddress(CGF, Loc);
     Address ZeroAddrBound =
         CGF.CreateDefaultAlignTempAlloca(CGF.Int32Ty,
                                          /*Name=*/".bound.zero.addr");
     CGF.InitTempAlloca(ZeroAddrBound, CGF.Builder.getInt32(/*C*/ 0));
     llvm::SmallVector<llvm::Value *, 16> OutlinedFnArgs;
     // ThreadId for serialized parallels is 0.
     OutlinedFnArgs.push_back(ThreadIDAddr.getPointer());
     OutlinedFnArgs.push_back(ZeroAddrBound.getPointer());
     OutlinedFnArgs.append(CapturedVars.begin(), CapturedVars.end());
     RT.emitOutlinedFunctionCall(CGF, Loc, OutlinedFn, OutlinedFnArgs);
 
     // __kmpc_end_serialized_parallel(&Loc, GTid);
     llvm::Value *EndArgs[] = {RT.emitUpdateLocation(CGF, Loc), ThreadID};
-    CGF.EmitRuntimeCall(
-        RT.createRuntimeFunction(OMPRTL__kmpc_end_serialized_parallel),
+    CGF.EmitRuntimeCall(llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+                            M, OMPRTL___kmpc_end_serialized_parallel),
                         EndArgs);
   };
   if (IfCond) {
     emitIfClause(CGF, IfCond, ThenGen, ElseGen);
   } else {
     RegionCodeGenTy ThenRCG(ThenGen);
     ThenRCG(CGF);
   }
 }
void CGOpenMPRuntime::emitCriticalRegion(CodeGenFunction &CGF,
   if (!CGF.HaveInsertPoint())
     return;
   llvm::Value *Args[] = {emitUpdateLocation(CGF, Loc), getThreadID(CGF, Loc),
                          getCriticalRegionLock(CriticalName)};
   llvm::SmallVector<llvm::Value *, 4> EnterArgs(std::begin(Args),
                                                 std::end(Args));
   if (Hint) {
     EnterArgs.push_back(CGF.Builder.CreateIntCast(
-        CGF.EmitScalarExpr(Hint), CGM.IntPtrTy, /*isSigned=*/false));
+        CGF.EmitScalarExpr(Hint), CGM.Int32Ty, /*isSigned=*/false));
   }
   CommonActionTy Action(
-      createRuntimeFunction(Hint ? OMPRTL__kmpc_critical_with_hint
-                                 : OMPRTL__kmpc_critical),
-      EnterArgs, createRuntimeFunction(OMPRTL__kmpc_end_critical), Args);
+      llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+          CGM.getModule(),
+          Hint ? OMPRTL___kmpc_critical_with_hint : OMPRTL___kmpc_critical),
+      EnterArgs,
+      llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+          CGM.getModule(), OMPRTL___kmpc_end_critical),
+      Args);
   CriticalOpGen.setAction(Action);
   emitInlinedDirective(CGF, OMPD_critical, CriticalOpGen);
 }
 
 void CGOpenMPRuntime::emitMasterRegion(CodeGenFunction &CGF,
                                        const RegionCodeGenTy &MasterOpGen,
                                        SourceLocation Loc) {
   if (!CGF.HaveInsertPoint())
     return;
   // if(__kmpc_master(ident_t *, gtid)) {
   //   MasterOpGen();
   //   __kmpc_end_master(ident_t *, gtid);
   // }
   // Prepare arguments and build a call to __kmpc_master
   llvm::Value *Args[] = {emitUpdateLocation(CGF, Loc), getThreadID(CGF, Loc)};
-  CommonActionTy Action(createRuntimeFunction(OMPRTL__kmpc_master), Args,
-                        createRuntimeFunction(OMPRTL__kmpc_end_master), Args,
+  CommonActionTy Action(llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+                            CGM.getModule(), OMPRTL___kmpc_master),
+                        Args,
+                        llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+                            CGM.getModule(), OMPRTL___kmpc_end_master),
+                        Args,
                         /*Conditional=*/true);
   MasterOpGen.setAction(Action);
   emitInlinedDirective(CGF, OMPD_master, MasterOpGen);
   Action.Done(CGF);
 }
 
 void CGOpenMPRuntime::emitTaskyieldCall(CodeGenFunction &CGF,
                                         SourceLocation Loc) {
   if (!CGF.HaveInsertPoint())
     return;
   llvm::OpenMPIRBuilder *OMPBuilder = CGF.CGM.getOpenMPIRBuilder();
   if (OMPBuilder) {
     OMPBuilder->CreateTaskyield(CGF.Builder);
   } else {
     // Build call __kmpc_omp_taskyield(loc, thread_id, 0);
     llvm::Value *Args[] = {
         emitUpdateLocation(CGF, Loc), getThreadID(CGF, Loc),
         llvm::ConstantInt::get(CGM.IntTy, /*V=*/0, /*isSigned=*/true)};
-    CGF.EmitRuntimeCall(createRuntimeFunction(OMPRTL__kmpc_omp_taskyield),
+    CGF.EmitRuntimeCall(llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+                            CGM.getModule(), OMPRTL___kmpc_omp_taskyield),
                         Args);
   }
 
   if (auto *Region = dyn_cast_or_null<CGOpenMPRegionInfo>(CGF.CapturedStmtInfo))
     Region->emitUntiedSwitch(CGF);
 }
 
 void CGOpenMPRuntime::emitTaskgroupRegion(CodeGenFunction &CGF,
                                           const RegionCodeGenTy &TaskgroupOpGen,
                                           SourceLocation Loc) {
   if (!CGF.HaveInsertPoint())
     return;
   // __kmpc_taskgroup(ident_t *, gtid);
   // TaskgroupOpGen();
   // __kmpc_end_taskgroup(ident_t *, gtid);
   // Prepare arguments and build a call to __kmpc_taskgroup
   llvm::Value *Args[] = {emitUpdateLocation(CGF, Loc), getThreadID(CGF, Loc)};
-  CommonActionTy Action(createRuntimeFunction(OMPRTL__kmpc_taskgroup), Args,
-                        createRuntimeFunction(OMPRTL__kmpc_end_taskgroup),
+  CommonActionTy Action(llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+                            CGM.getModule(), OMPRTL___kmpc_taskgroup),
+                        Args,
+                        llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+                            CGM.getModule(), OMPRTL___kmpc_end_taskgroup),
                         Args);
   TaskgroupOpGen.setAction(Action);
   emitInlinedDirective(CGF, OMPD_taskgroup, TaskgroupOpGen);
 }
 
 /// Given an array of pointers to variables, project the address of a
 /// given variable.
 static Address emitAddrOfVarFromArray(CodeGenFunction &CGF, Address Array,
if (!CopyprivateVars.empty()) {
     // int32 did_it = 0;
     QualType KmpInt32Ty =
         C.getIntTypeForBitwidth(/*DestWidth=*/32, /*Signed=*/1);
     DidIt = CGF.CreateMemTemp(KmpInt32Ty, ".omp.copyprivate.did_it");
     CGF.Builder.CreateStore(CGF.Builder.getInt32(0), DidIt);
   }
   // Prepare arguments and build a call to __kmpc_single
   llvm::Value *Args[] = {emitUpdateLocation(CGF, Loc), getThreadID(CGF, Loc)};
-  CommonActionTy Action(createRuntimeFunction(OMPRTL__kmpc_single), Args,
-                        createRuntimeFunction(OMPRTL__kmpc_end_single), Args,
+  CommonActionTy Action(llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+                            CGM.getModule(), OMPRTL___kmpc_single),
+                        Args,
+                        llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+                            CGM.getModule(), OMPRTL___kmpc_end_single),
+                        Args,
                         /*Conditional=*/true);
   SingleOpGen.setAction(Action);
   emitInlinedDirective(CGF, OMPD_single, SingleOpGen);
   if (DidIt.isValid()) {
     // did_it = 1;
     CGF.Builder.CreateStore(CGF.Builder.getInt32(1), DidIt);
   }
   Action.Done(CGF);
if (DidIt.isValid()) {
     llvm::Value *Args[] = {
         emitUpdateLocation(CGF, Loc), // ident_t *<loc>
         getThreadID(CGF, Loc),        // i32 <gtid>
         BufSize,                      // size_t <buf_size>
         CL.getPointer(),              // void *<copyprivate list>
         CpyFn,                        // void (*) (void *, void *) <copy_func>
         DidItVal                      // i32 did_it
     };
-    CGF.EmitRuntimeCall(createRuntimeFunction(OMPRTL__kmpc_copyprivate), Args);
+    CGF.EmitRuntimeCall(llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+                            CGM.getModule(), OMPRTL___kmpc_copyprivate),
+                        Args);
   }
 }
 
 void CGOpenMPRuntime::emitOrderedRegion(CodeGenFunction &CGF,
                                         const RegionCodeGenTy &OrderedOpGen,
                                         SourceLocation Loc, bool IsThreads) {
   if (!CGF.HaveInsertPoint())
     return;
   // __kmpc_ordered(ident_t *, gtid);
   // OrderedOpGen();
   // __kmpc_end_ordered(ident_t *, gtid);
   // Prepare arguments and build a call to __kmpc_ordered
   if (IsThreads) {
     llvm::Value *Args[] = {emitUpdateLocation(CGF, Loc), getThreadID(CGF, Loc)};
-    CommonActionTy Action(createRuntimeFunction(OMPRTL__kmpc_ordered), Args,
-                          createRuntimeFunction(OMPRTL__kmpc_end_ordered),
+    CommonActionTy Action(llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+                              CGM.getModule(), OMPRTL___kmpc_ordered),
+                          Args,
+                          llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+                              CGM.getModule(), OMPRTL___kmpc_end_ordered),
                           Args);
     OrderedOpGen.setAction(Action);
     emitInlinedDirective(CGF, OMPD_ordered, OrderedOpGen);
     return;
   }
   emitInlinedDirective(CGF, OMPD_ordered, OrderedOpGen);
 }
 
void CGOpenMPRuntime::emitBarrierCall(CodeGenFunction &CGF, SourceLocation Loc,
   unsigned Flags = getDefaultFlagsForBarriers(Kind);
   // Build call __kmpc_cancel_barrier(loc, thread_id) or __kmpc_barrier(loc,
   // thread_id);
   llvm::Value *Args[] = {emitUpdateLocation(CGF, Loc, Flags),
                          getThreadID(CGF, Loc)};
   if (OMPRegionInfo) {
     if (!ForceSimpleCall && OMPRegionInfo->hasCancel()) {
       llvm::Value *Result = CGF.EmitRuntimeCall(
-          createRuntimeFunction(OMPRTL__kmpc_cancel_barrier), Args);
+          llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+              CGM.getModule(), OMPRTL___kmpc_cancel_barrier),
+          Args);
       if (EmitChecks) {
         // if (__kmpc_cancel_barrier()) {
         //   exit from construct;
         // }
         llvm::BasicBlock *ExitBB = CGF.createBasicBlock(".cancel.exit");
         llvm::BasicBlock *ContBB = CGF.createBasicBlock(".cancel.continue");
         llvm::Value *Cmp = CGF.Builder.CreateIsNotNull(Result);
         CGF.Builder.CreateCondBr(Cmp, ExitBB, ContBB);
         CGF.EmitBlock(ExitBB);
         //   exit from construct;
         CodeGenFunction::JumpDest CancelDestination =
             CGF.getOMPCancelDestination(OMPRegionInfo->getDirectiveKind());
         CGF.EmitBranchThroughCleanup(CancelDestination);
         CGF.EmitBlock(ContBB, /*IsFinished=*/true);
       }
       return;
     }
   }
-  CGF.EmitRuntimeCall(createRuntimeFunction(OMPRTL__kmpc_barrier), Args);
+  CGF.EmitRuntimeCall(llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+                          CGM.getModule(), OMPRTL___kmpc_barrier),
+                      Args);
 }
 
 /// Map the OpenMP loop schedule to the runtime enumeration.
 static OpenMPSchedType getRuntimeSchedule(OpenMPScheduleClauseKind ScheduleKind,
                                           bool Chunked, bool Ordered) {
   switch (ScheduleKind) {
   case OMPC_SCHEDULE_static:
     return Chunked ? (Ordered ? OMP_ord_static_chunked : OMP_sch_static_chunked)
llvm::Value *Args[] = {
       emitUpdateLocation(CGF, Loc,
                          isOpenMPDistributeDirective(DKind)
                              ? OMP_IDENT_WORK_DISTRIBUTE
                              : isOpenMPLoopDirective(DKind)
                                    ? OMP_IDENT_WORK_LOOP
                                    : OMP_IDENT_WORK_SECTIONS),
       getThreadID(CGF, Loc)};
   auto DL = ApplyDebugLocation::CreateDefaultArtificial(CGF, Loc);
-  CGF.EmitRuntimeCall(createRuntimeFunction(OMPRTL__kmpc_for_static_fini),
+  CGF.EmitRuntimeCall(llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+                          CGM.getModule(), OMPRTL___kmpc_for_static_fini),
                       Args);
 }
 
 void CGOpenMPRuntime::emitForOrderedIterationEnd(CodeGenFunction &CGF,
                                                  SourceLocation Loc,
                                                  unsigned IVSize,
                                                  bool IVSigned) {
   if (!CGF.HaveInsertPoint())
void CGOpenMPRuntime::emitNumThreadsClause(CodeGenFunction &CGF,
                                            llvm::Value *NumThreads,
                                            SourceLocation Loc) {
   if (!CGF.HaveInsertPoint())
     return;
   // Build call __kmpc_push_num_threads(&loc, global_tid, num_threads)
   llvm::Value *Args[] = {
       emitUpdateLocation(CGF, Loc), getThreadID(CGF, Loc),
       CGF.Builder.CreateIntCast(NumThreads, CGF.Int32Ty, /*isSigned*/ true)};
-  CGF.EmitRuntimeCall(createRuntimeFunction(OMPRTL__kmpc_push_num_threads),
+  CGF.EmitRuntimeCall(llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+                          CGM.getModule(), OMPRTL___kmpc_push_num_threads),
                       Args);
 }
 
 void CGOpenMPRuntime::emitProcBindClause(CodeGenFunction &CGF,
                                          ProcBindKind ProcBind,
                                          SourceLocation Loc) {
   if (!CGF.HaveInsertPoint())
     return;
   assert(ProcBind != OMP_PROC_BIND_unknown && "Unsupported proc_bind value.");
   // Build call __kmpc_push_proc_bind(&loc, global_tid, proc_bind)
   llvm::Value *Args[] = {
       emitUpdateLocation(CGF, Loc), getThreadID(CGF, Loc),
       llvm::ConstantInt::get(CGM.IntTy, unsigned(ProcBind), /*isSigned=*/true)};
-  CGF.EmitRuntimeCall(createRuntimeFunction(OMPRTL__kmpc_push_proc_bind), Args);
+  CGF.EmitRuntimeCall(llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+                          CGM.getModule(), OMPRTL___kmpc_push_proc_bind),
+                      Args);
 }
 
 void CGOpenMPRuntime::emitFlush(CodeGenFunction &CGF, ArrayRef<const Expr *>,
                                 SourceLocation Loc, llvm::AtomicOrdering AO) {
   llvm::OpenMPIRBuilder *OMPBuilder = CGF.CGM.getOpenMPIRBuilder();
   if (OMPBuilder) {
     OMPBuilder->CreateFlush(CGF.Builder);
   } else {
     if (!CGF.HaveInsertPoint())
       return;
     // Build call void __kmpc_flush(ident_t *loc)
-    CGF.EmitRuntimeCall(createRuntimeFunction(OMPRTL__kmpc_flush),
+    CGF.EmitRuntimeCall(llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+                            CGM.getModule(), OMPRTL___kmpc_flush),
                         emitUpdateLocation(CGF, Loc));
   }
 }
 
 namespace {
 /// Indexes of fields for type kmp_task_t.
 enum KmpTaskTFields {
   /// List of shared variables.
if (D.hasClausesOfKind<OMPNowaitClause>()) {
     llvm::Value *DeviceID;
     if (Device)
       DeviceID = CGF.Builder.CreateIntCast(CGF.EmitScalarExpr(Device),
                                            CGF.Int64Ty, /*isSigned=*/true);
     else
       DeviceID = CGF.Builder.getInt64(OMP_DEVICEID_UNDEF);
     AllocArgs.push_back(DeviceID);
     NewTask = CGF.EmitRuntimeCall(
-      createRuntimeFunction(OMPRTL__kmpc_omp_target_task_alloc), AllocArgs);
-  } else {
-    NewTask = CGF.EmitRuntimeCall(
-      createRuntimeFunction(OMPRTL__kmpc_omp_task_alloc), AllocArgs);
+        llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+            CGM.getModule(), OMPRTL___kmpc_omp_target_task_alloc),
+        AllocArgs);
+  } else {
+    NewTask =
+        CGF.EmitRuntimeCall(llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+                                CGM.getModule(), OMPRTL___kmpc_omp_task_alloc),
+                            AllocArgs);
   }
   // Emit detach clause initialization.
   // evt = (typeof(evt))__kmpc_task_allow_completion_event(loc, tid,
   // task_descriptor);
   if (const auto *DC = D.getSingleClause<OMPDetachClause>()) {
     const Expr *Evt = DC->getEventHandler()->IgnoreParenImpCasts();
     LValue EvtLVal = CGF.EmitLValue(Evt);
 
     // Build kmp_event_t *__kmpc_task_allow_completion_event(ident_t *loc_ref,
     // int gtid, kmp_task_t *task);
     llvm::Value *Loc = emitUpdateLocation(CGF, DC->getBeginLoc());
     llvm::Value *Tid = getThreadID(CGF, DC->getBeginLoc());
     Tid = CGF.Builder.CreateIntCast(Tid, CGF.IntTy, /*isSigned=*/false);
     llvm::Value *EvtVal = CGF.EmitRuntimeCall(
-        createRuntimeFunction(OMPRTL__kmpc_task_allow_completion_event),
+        llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+            CGM.getModule(), OMPRTL___kmpc_task_allow_completion_event),
         {Loc, Tid, NewTask});
     EvtVal = CGF.EmitScalarConversion(EvtVal, C.VoidPtrTy, Evt->getType(),
                                       Evt->getExprLoc());
     CGF.EmitStoreOfScalar(EvtVal, EvtLVal);
   }
   llvm::Value *NewTaskNewTaskTTy =
       CGF.Builder.CreatePointerBitCastOrAddrSpaceCast(
           NewTask, KmpTaskTWithPrivatesPtrTy);
if (const auto *IE =
     NumDepsVal = llvm::ConstantInt::get(CGF.IntPtrTy, NumDependencies);
   }
   // Need to allocate on the dynamic memory.
   llvm::Value *ThreadID = getThreadID(CGF, Loc);
   // Use default allocator.
   llvm::Value *Allocator = llvm::ConstantPointerNull::get(CGF.VoidPtrTy);
   llvm::Value *Args[] = {ThreadID, Size, Allocator};
 
-  llvm::Value *Addr = CGF.EmitRuntimeCall(
-      createRuntimeFunction(OMPRTL__kmpc_alloc), Args, ".dep.arr.addr");
+  llvm::Value *Addr =
+      CGF.EmitRuntimeCall(llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+                              CGM.getModule(), OMPRTL___kmpc_alloc),
+                          Args, ".dep.arr.addr");
   Addr = CGF.Builder.CreatePointerBitCastOrAddrSpaceCast(
       Addr, CGF.ConvertTypeForMem(KmpDependInfoTy)->getPointerTo());
   DependenciesArray = Address(Addr, Align);
   // Write number of elements in the first element of array for depobj.
   LValue Base = CGF.MakeAddrLValue(DependenciesArray, KmpDependInfoTy);
   // deps[i].base_addr = NumDependencies;
   LValue BaseAddrLVal = CGF.EmitLValueForField(
       Base, *std::next(KmpDependInfoRD->field_begin(), BaseAddr));
void CGOpenMPRuntime::emitDestroyClause(CodeGenFunction &CGF, LValue DepobjLVal,
   DepObjAddr = CGF.Builder.CreatePointerBitCastOrAddrSpaceCast(DepObjAddr,
                                                                CGF.VoidPtrTy);
   llvm::Value *ThreadID = getThreadID(CGF, Loc);
   // Use default allocator.
   llvm::Value *Allocator = llvm::ConstantPointerNull::get(CGF.VoidPtrTy);
   llvm::Value *Args[] = {ThreadID, DepObjAddr, Allocator};
 
   // _kmpc_free(gtid, addr, nullptr);
-  (void)CGF.EmitRuntimeCall(createRuntimeFunction(OMPRTL__kmpc_free), Args);
+  (void)CGF.EmitRuntimeCall(llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+                                CGM.getModule(), OMPRTL___kmpc_free),
+                            Args);
 }
 
 void CGOpenMPRuntime::emitUpdateClause(CodeGenFunction &CGF, LValue DepobjLVal,
                                        OpenMPDependClauseKind NewDepKind,
                                        SourceLocation Loc) {
   ASTContext &C = CGM.getContext();
   QualType FlagsTy;
   getDependTypes(C, KmpDependInfoTy, FlagsTy);
auto &&ThenCodeGen = [this, &Data, TDBase, KmpTaskTQTyRD, &TaskArgs,
                         &DepTaskArgs](CodeGenFunction &CGF, PrePostActionTy &) {
     if (!Data.Tied) {
       auto PartIdFI = std::next(KmpTaskTQTyRD->field_begin(), KmpTaskTPartId);
       LValue PartIdLVal = CGF.EmitLValueForField(TDBase, *PartIdFI);
       CGF.EmitStoreOfScalar(CGF.Builder.getInt32(0), PartIdLVal);
     }
     if (!Data.Dependences.empty()) {
       CGF.EmitRuntimeCall(
-          createRuntimeFunction(OMPRTL__kmpc_omp_task_with_deps), DepTaskArgs);
+          llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+              CGM.getModule(), OMPRTL___kmpc_omp_task_with_deps),
+          DepTaskArgs);
     } else {
-      CGF.EmitRuntimeCall(createRuntimeFunction(OMPRTL__kmpc_omp_task),
+      CGF.EmitRuntimeCall(llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+                              CGM.getModule(), OMPRTL___kmpc_omp_task),
                           TaskArgs);
     }
     // Check if parent region is untied and build return for untied task;
     if (auto *Region =
             dyn_cast_or_null<CGOpenMPRegionInfo>(CGF.CapturedStmtInfo))
       Region->emitUntiedSwitch(CGF);
   };
 
   llvm::Value *DepWaitTaskArgs[6];
   if (!Data.Dependences.empty()) {
     DepWaitTaskArgs[0] = UpLoc;
     DepWaitTaskArgs[1] = ThreadID;
     DepWaitTaskArgs[2] = NumOfElements;
     DepWaitTaskArgs[3] = DependenciesArray.getPointer();
     DepWaitTaskArgs[4] = CGF.Builder.getInt32(0);
     DepWaitTaskArgs[5] = llvm::ConstantPointerNull::get(CGF.VoidPtrTy);
   }
-  auto &&ElseCodeGen = [&TaskArgs, ThreadID, NewTaskNewTaskTTy, TaskEntry,
+  auto &M = CGM.getModule();
+  auto &&ElseCodeGen = [&M, &TaskArgs, ThreadID, NewTaskNewTaskTTy, TaskEntry,
                         &Data, &DepWaitTaskArgs,
                         Loc](CodeGenFunction &CGF, PrePostActionTy &) {
     CGOpenMPRuntime &RT = CGF.CGM.getOpenMPRuntime();
     CodeGenFunction::RunCleanupsScope LocalScope(CGF);
     // Build void __kmpc_omp_wait_deps(ident_t *, kmp_int32 gtid,
     // kmp_int32 ndeps, kmp_depend_info_t *dep_list, kmp_int32
     // ndeps_noalias, kmp_depend_info_t *noalias_dep_list); if dependence info
     // is specified.
     if (!Data.Dependences.empty())
-      CGF.EmitRuntimeCall(RT.createRuntimeFunction(OMPRTL__kmpc_omp_wait_deps),
+      CGF.EmitRuntimeCall(llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+                              M, OMPRTL___kmpc_omp_wait_deps),
                           DepWaitTaskArgs);
     // Call proxy_task_entry(gtid, new_task);
     auto &&CodeGen = [TaskEntry, ThreadID, NewTaskNewTaskTTy,
                       Loc](CodeGenFunction &CGF, PrePostActionTy &Action) {
       Action.Enter(CGF);
       llvm::Value *OutlinedFnArgs[] = {ThreadID, NewTaskNewTaskTTy};
       CGF.CGM.getOpenMPRuntime().emitOutlinedFunctionCall(CGF, Loc, TaskEntry,
                                                           OutlinedFnArgs);
     };
 
     // Build void __kmpc_omp_task_begin_if0(ident_t *, kmp_int32 gtid,
     // kmp_task_t *new_task);
     // Build void __kmpc_omp_task_complete_if0(ident_t *, kmp_int32 gtid,
     // kmp_task_t *new_task);
     RegionCodeGenTy RCG(CodeGen);
-    CommonActionTy Action(
-        RT.createRuntimeFunction(OMPRTL__kmpc_omp_task_begin_if0), TaskArgs,
-        RT.createRuntimeFunction(OMPRTL__kmpc_omp_task_complete_if0), TaskArgs);
+    CommonActionTy Action(llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+                              M, OMPRTL___kmpc_omp_task_begin_if0),
+                          TaskArgs,
+                          llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+                              M, OMPRTL___kmpc_omp_task_complete_if0),
+                          TaskArgs);
     RCG.setAction(Action);
     RCG(CGF);
   };
 
   if (IfCond) {
     emitIfClause(CGF, IfCond, ThenCodeGen, ElseCodeGen);
   } else {
     RegionCodeGenTy ThenRCG(ThenCodeGen);
llvm::Value *TaskArgs[] = {
                          : NoSchedule),
       Data.Schedule.getPointer()
           ? CGF.Builder.CreateIntCast(Data.Schedule.getPointer(), CGF.Int64Ty,
                                       /*isSigned=*/false)
           : llvm::ConstantInt::get(CGF.Int64Ty, /*V=*/0),
       Result.TaskDupFn ? CGF.Builder.CreatePointerBitCastOrAddrSpaceCast(
                              Result.TaskDupFn, CGF.VoidPtrTy)
                        : llvm::ConstantPointerNull::get(CGF.VoidPtrTy)};
-  CGF.EmitRuntimeCall(createRuntimeFunction(OMPRTL__kmpc_taskloop), TaskArgs);
+  CGF.EmitRuntimeCall(llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+                          CGM.getModule(), OMPRTL___kmpc_taskloop),
+                      TaskArgs);
 }
 
 /// Emit reduction operation for each element of array (required for
 /// array sections) LHS op = RHS.
 /// \param Type Type of array.
 /// \param LHSVar Variable on the left side of the reduction operation
 /// (references element of array in original variable).
 /// \param RHSVar Variable on the right side of the reduction operation
llvm::Value *Args[] = {
       ThreadId,                              // i32 <gtid>
       CGF.Builder.getInt32(RHSExprs.size()), // i32 <n>
       ReductionArrayTySize,                  // size_type sizeof(RedList)
       RL,                                    // void *RedList
       ReductionFn, // void (*) (void *, void *) <reduce_func>
       Lock         // kmp_critical_name *&<lock>
   };
   llvm::Value *Res = CGF.EmitRuntimeCall(
-      createRuntimeFunction(WithNowait ? OMPRTL__kmpc_reduce_nowait
-                                       : OMPRTL__kmpc_reduce),
+      llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+          CGM.getModule(),
+          WithNowait ? OMPRTL___kmpc_reduce_nowait : OMPRTL___kmpc_reduce),
       Args);
 
   // 5. Build switch(res)
   llvm::BasicBlock *DefaultBB = CGF.createBasicBlock(".omp.reduction.default");
   llvm::SwitchInst *SwInst =
       CGF.Builder.CreateSwitch(Res, DefaultBB, /*NumCases=*/2);
 
   // 6. Build case 1:
for (const Expr *E : ReductionOps) {
       ++IPriv;
       ++ILHS;
       ++IRHS;
     }
   };
   RegionCodeGenTy RCG(CodeGen);
   CommonActionTy Action(
       nullptr, llvm::None,
-      createRuntimeFunction(WithNowait ? OMPRTL__kmpc_end_reduce_nowait
-                                       : OMPRTL__kmpc_end_reduce),
+      llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+          CGM.getModule(), WithNowait ? OMPRTL___kmpc_end_reduce_nowait
+                                      : OMPRTL___kmpc_end_reduce),
       EndArgs);
   RCG.setAction(Action);
   RCG(CGF);
 
   CGF.EmitBranch(DefaultBB);
 
   // 7. Build case 2:
   //  ...
void CGOpenMPRuntime::emitReduction(CodeGenFunction &CGF, SourceLocation Loc,
   if (!WithNowait) {
     // Add emission of __kmpc_end_reduce(<loc>, <gtid>, &<lock>);
     llvm::Value *EndArgs[] = {
         IdentTLoc, // ident_t *<loc>
         ThreadId,  // i32 <gtid>
         Lock       // kmp_critical_name *&<lock>
     };
     CommonActionTy Action(nullptr, llvm::None,
-                          createRuntimeFunction(OMPRTL__kmpc_end_reduce),
+                          llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+                              CGM.getModule(), OMPRTL___kmpc_end_reduce),
                           EndArgs);
     AtomicRCG.setAction(Action);
     AtomicRCG(CGF);
   } else {
     AtomicRCG(CGF);
   }
 
   CGF.EmitBranch(DefaultBB);
if (Data.IsReductionWithTaskMod) {
     llvm::Value *Args[] = {
         IdentTLoc, GTid,
         llvm::ConstantInt::get(CGM.IntTy, Data.IsWorksharingReduction ? 1 : 0,
                                /*isSigned=*/true),
         llvm::ConstantInt::get(CGM.IntTy, Size, /*isSigned=*/true),
         CGF.Builder.CreatePointerBitCastOrAddrSpaceCast(
             TaskRedInput.getPointer(), CGM.VoidPtrTy)};
     return CGF.EmitRuntimeCall(
-        createRuntimeFunction(OMPRTL__kmpc_taskred_modifier_init), Args);
+        llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+            CGM.getModule(), OMPRTL___kmpc_taskred_modifier_init),
+        Args);
   }
   // Build call void *__kmpc_taskred_init(int gtid, int num_data, void *data);
   llvm::Value *Args[] = {
       CGF.Builder.CreateIntCast(getThreadID(CGF, Loc), CGM.IntTy,
                                 /*isSigned=*/true),
       llvm::ConstantInt::get(CGM.IntTy, Size, /*isSigned=*/true),
       CGF.Builder.CreatePointerBitCastOrAddrSpaceCast(TaskRedInput.getPointer(),
                                                       CGM.VoidPtrTy)};
-  return CGF.EmitRuntimeCall(createRuntimeFunction(OMPRTL__kmpc_taskred_init),
+  return CGF.EmitRuntimeCall(llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+                                 CGM.getModule(), OMPRTL___kmpc_taskred_init),
                              Args);
 }
 
 void CGOpenMPRuntime::emitTaskReductionFini(CodeGenFunction &CGF,
                                             SourceLocation Loc,
                                             bool IsWorksharingReduction) {
   // Build call void *__kmpc_taskred_modifier_init(ident_t *loc, int gtid, int
   // is_ws, int num, void *data);
   llvm::Value *IdentTLoc = emitUpdateLocation(CGF, Loc);
   llvm::Value *GTid = CGF.Builder.CreateIntCast(getThreadID(CGF, Loc),
                                                 CGM.IntTy, /*isSigned=*/true);
   llvm::Value *Args[] = {IdentTLoc, GTid,
                          llvm::ConstantInt::get(CGM.IntTy,
                                                 IsWorksharingReduction ? 1 : 0,
                                                 /*isSigned=*/true)};
   (void)CGF.EmitRuntimeCall(
-      createRuntimeFunction(OMPRTL__kmpc_task_reduction_modifier_fini), Args);
+      llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+          CGM.getModule(), OMPRTL___kmpc_task_reduction_modifier_fini),
+      Args);
 }
 
 void CGOpenMPRuntime::emitTaskReductionFixups(CodeGenFunction &CGF,
                                               SourceLocation Loc,
                                               ReductionCodeGen &RCG,
                                               unsigned N) {
   auto Sizes = RCG.getSizes(N);
   // Emit threadprivate global variable if the type is non-constant
Address CGOpenMPRuntime::getTaskReductionItem(CodeGenFunction &CGF,
   llvm::Value *Args[] = {CGF.Builder.CreateIntCast(getThreadID(CGF, Loc),
                                                    CGM.IntTy,
                                                    /*isSigned=*/true),
                          ReductionsPtr,
                          CGF.Builder.CreatePointerBitCastOrAddrSpaceCast(
                              SharedLVal.getPointer(CGF), CGM.VoidPtrTy)};
   return Address(
       CGF.EmitRuntimeCall(
-          createRuntimeFunction(OMPRTL__kmpc_task_reduction_get_th_data), Args),
+          llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+              CGM.getModule(), OMPRTL___kmpc_task_reduction_get_th_data),
+          Args),
       SharedLVal.getAlignment());
 }
 
 void CGOpenMPRuntime::emitTaskwaitCall(CodeGenFunction &CGF,
                                        SourceLocation Loc) {
   if (!CGF.HaveInsertPoint())
     return;
 
   llvm::OpenMPIRBuilder *OMPBuilder = CGF.CGM.getOpenMPIRBuilder();
   if (OMPBuilder) {
     OMPBuilder->CreateTaskwait(CGF.Builder);
   } else {
     // Build call kmp_int32 __kmpc_omp_taskwait(ident_t *loc, kmp_int32
     // global_tid);
     llvm::Value *Args[] = {emitUpdateLocation(CGF, Loc), getThreadID(CGF, Loc)};
     // Ignore return result until untied tasks are supported.
-    CGF.EmitRuntimeCall(createRuntimeFunction(OMPRTL__kmpc_omp_taskwait), Args);
+    CGF.EmitRuntimeCall(llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+                            CGM.getModule(), OMPRTL___kmpc_omp_taskwait),
+                        Args);
   }
 
   if (auto *Region = dyn_cast_or_null<CGOpenMPRegionInfo>(CGF.CapturedStmtInfo))
     Region->emitUntiedSwitch(CGF);
 }
 
 void CGOpenMPRuntime::emitInlinedDirective(CodeGenFunction &CGF,
                                            OpenMPDirectiveKind InnerKind,
if (auto *OMPRegionInfo =
     // For 'cancellation point taskgroup', the task region info may not have a
     // cancel. This may instead happen in another adjacent task.
     if (CancelRegion == OMPD_taskgroup || OMPRegionInfo->hasCancel()) {
       llvm::Value *Args[] = {
           emitUpdateLocation(CGF, Loc), getThreadID(CGF, Loc),
           CGF.Builder.getInt32(getCancellationKind(CancelRegion))};
       // Ignore return result until untied tasks are supported.
       llvm::Value *Result = CGF.EmitRuntimeCall(
-          createRuntimeFunction(OMPRTL__kmpc_cancellationpoint), Args);
+          llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+              CGM.getModule(), OMPRTL___kmpc_cancellationpoint),
+          Args);
       // if (__kmpc_cancellationpoint()) {
       //   exit from construct;
       // }
       llvm::BasicBlock *ExitBB = CGF.createBasicBlock(".cancel.exit");
       llvm::BasicBlock *ContBB = CGF.createBasicBlock(".cancel.continue");
       llvm::Value *Cmp = CGF.Builder.CreateIsNotNull(Result);
       CGF.Builder.CreateCondBr(Cmp, ExitBB, ContBB);
       CGF.EmitBlock(ExitBB);
       // exit from construct;
       CodeGenFunction::JumpDest CancelDest =
           CGF.getOMPCancelDestination(OMPRegionInfo->getDirectiveKind());
       CGF.EmitBranchThroughCleanup(CancelDest);
       CGF.EmitBlock(ContBB, /*IsFinished=*/true);
     }
   }
 }
 
 void CGOpenMPRuntime::emitCancelCall(CodeGenFunction &CGF, SourceLocation Loc,
                                      const Expr *IfCond,
                                      OpenMPDirectiveKind CancelRegion) {
   if (!CGF.HaveInsertPoint())
     return;
   // Build call kmp_int32 __kmpc_cancel(ident_t *loc, kmp_int32 global_tid,
   // kmp_int32 cncl_kind);
+  auto &M = CGM.getModule();
   if (auto *OMPRegionInfo =
           dyn_cast_or_null<CGOpenMPRegionInfo>(CGF.CapturedStmtInfo)) {
-    auto &&ThenGen = [Loc, CancelRegion, OMPRegionInfo](CodeGenFunction &CGF,
-                                                        PrePostActionTy &) {
+    auto &&ThenGen = [&M, Loc, CancelRegion,
+                      OMPRegionInfo](CodeGenFunction &CGF, PrePostActionTy &) {
       CGOpenMPRuntime &RT = CGF.CGM.getOpenMPRuntime();
       llvm::Value *Args[] = {
           RT.emitUpdateLocation(CGF, Loc), RT.getThreadID(CGF, Loc),
           CGF.Builder.getInt32(getCancellationKind(CancelRegion))};
       // Ignore return result until untied tasks are supported.
-      llvm::Value *Result = CGF.EmitRuntimeCall(
-          RT.createRuntimeFunction(OMPRTL__kmpc_cancel), Args);
+      llvm::Value *Result =
+          CGF.EmitRuntimeCall(llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+                                  M, OMPRTL___kmpc_cancel),
+                              Args);
       // if (__kmpc_cancel()) {
       //   exit from construct;
       // }
       llvm::BasicBlock *ExitBB = CGF.createBasicBlock(".cancel.exit");
       llvm::BasicBlock *ContBB = CGF.createBasicBlock(".cancel.continue");
       llvm::Value *Cmp = CGF.Builder.CreateIsNotNull(Result);
       CGF.Builder.CreateCondBr(Cmp, ExitBB, ContBB);
       CGF.EmitBlock(ExitBB);
void CGOpenMPRuntime::emitUserDefinedMapper(const OMPDeclareMapperDecl *D,
   MappableExprsHandler::MapFlagsArrayTy MapTypes;
   MappableExprsHandler MEHandler(*D, MapperCGF);
   MEHandler.generateAllInfoForMapper(BasePointers, Pointers, Sizes, MapTypes);
 
   // Call the runtime API __tgt_mapper_num_components to get the number of
   // pre-existing components.
   llvm::Value *OffloadingArgs[] = {Handle};
   llvm::Value *PreviousSize = MapperCGF.EmitRuntimeCall(
-      createRuntimeFunction(OMPRTL__tgt_mapper_num_components), OffloadingArgs);
+      llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+          CGM.getModule(), OMPRTL___tgt_mapper_num_components),
+      OffloadingArgs);
   llvm::Value *ShiftedPreviousSize = MapperCGF.Builder.CreateShl(
       PreviousSize,
       MapperCGF.Builder.getInt64(MappableExprsHandler::getFlagMemberOffset()));
 
   // Fill up the runtime mapper handle for all components.
   for (unsigned I = 0; I < BasePointers.size(); ++I) {
     llvm::Value *CurBaseArg = MapperCGF.Builder.CreateBitCast(
         *BasePointers[I], CGM.getTypes().ConvertTypeForMem(C.VoidPtrTy));
for (unsigned I = 0; I < BasePointers.size(); ++I) {
 
     // TODO: call the corresponding mapper function if a user-defined mapper is
     // associated with this map clause.
     // Call the runtime API __tgt_push_mapper_component to fill up the runtime
     // data structure.
     llvm::Value *OffloadingArgs[] = {Handle, CurBaseArg, CurBeginArg,
                                      CurSizeArg, CurMapType};
     MapperCGF.EmitRuntimeCall(
-        createRuntimeFunction(OMPRTL__tgt_push_mapper_component),
+        llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+            CGM.getModule(), OMPRTL___tgt_push_mapper_component),
         OffloadingArgs);
   }
 
   // Update the pointer to point to the next element that needs to be mapped,
   // and check whether we have mapped all elements.
   llvm::Value *PtrNext = MapperCGF.Builder.CreateConstGEP1_32(
       PtrPHI, /*Idx0=*/1, "omp.arraymap.next");
   PtrPHI->addIncoming(PtrNext, BodyBB);
void CGOpenMPRuntime::emitUDMapperArrayInitOrDel(
   llvm::Value *MapTypeArg = MapperCGF.Builder.CreateAnd(
       MapType,
       MapperCGF.Builder.getInt64(~(MappableExprsHandler::OMP_MAP_TO |
                                    MappableExprsHandler::OMP_MAP_FROM)));
   // Call the runtime API __tgt_push_mapper_component to fill up the runtime
   // data structure.
   llvm::Value *OffloadingArgs[] = {Handle, Base, Begin, ArraySize, MapTypeArg};
   MapperCGF.EmitRuntimeCall(
-      createRuntimeFunction(OMPRTL__tgt_push_mapper_component), OffloadingArgs);
+      llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+          CGM.getModule(), OMPRTL___tgt_push_mapper_component),
+      OffloadingArgs);
 }
 
 void CGOpenMPRuntime::emitTargetNumIterationsCall(
     CodeGenFunction &CGF, const OMPExecutableDirective &D,
     llvm::Value *DeviceID,
     llvm::function_ref<llvm::Value *(CodeGenFunction &CGF,
                                      const OMPLoopDirective &D)>
         SizeEmitter) {
   OpenMPDirectiveKind Kind = D.getDirectiveKind();
   const OMPExecutableDirective *TD = &D;
   // Get nested teams distribute kind directive, if any.
   if (!isOpenMPDistributeDirective(Kind) || !isOpenMPTeamsDirective(Kind))
     TD = getNestedDistributeDirective(CGM.getContext(), D);
   if (!TD)
     return;
   const auto *LD = cast<OMPLoopDirective>(TD);
   auto &&CodeGen = [LD, DeviceID, SizeEmitter, this](CodeGenFunction &CGF,
                                                      PrePostActionTy &) {
     if (llvm::Value *NumIterations = SizeEmitter(CGF, *LD)) {
       llvm::Value *Args[] = {DeviceID, NumIterations};
       CGF.EmitRuntimeCall(
-          createRuntimeFunction(OMPRTL__kmpc_push_target_tripcount), Args);
+          llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+              CGM.getModule(), OMPRTL___kmpc_push_target_tripcount),
+          Args);
     }
   };
   emitInlinedDirective(CGF, OMPD_unknown, CodeGen);
 }
 
 void CGOpenMPRuntime::emitTargetCall(
     CodeGenFunction &CGF, const OMPExecutableDirective &D,
     llvm::Function *OutlinedFn, llvm::Value *OutlinedFnID, const Expr *IfCond,
if (NumTeams) {
                                        PointerNum,
                                        InputInfo.BasePointersArray.getPointer(),
                                        InputInfo.PointersArray.getPointer(),
                                        InputInfo.SizesArray.getPointer(),
                                        MapTypesArray,
                                        NumTeams,
                                        NumThreads};
       Return = CGF.EmitRuntimeCall(
-          createRuntimeFunction(HasNowait ? OMPRTL__tgt_target_teams_nowait
-                                          : OMPRTL__tgt_target_teams),
+          llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+              CGM.getModule(), HasNowait ? OMPRTL___tgt_target_teams_nowait
+                                         : OMPRTL___tgt_target_teams),
           OffloadingArgs);
     } else {
       llvm::Value *OffloadingArgs[] = {DeviceID,
                                        OutlinedFnID,
                                        PointerNum,
                                        InputInfo.BasePointersArray.getPointer(),
                                        InputInfo.PointersArray.getPointer(),
                                        InputInfo.SizesArray.getPointer(),
                                        MapTypesArray};
       Return = CGF.EmitRuntimeCall(
-          createRuntimeFunction(HasNowait ? OMPRTL__tgt_target_nowait
-                                          : OMPRTL__tgt_target),
+          llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+              CGM.getModule(),
+              HasNowait ? OMPRTL___tgt_target_nowait : OMPRTL___tgt_target),
           OffloadingArgs);
     }
 
     // Check the error code and execute the host version if required.
     llvm::BasicBlock *OffloadFailedBlock =
         CGF.createBasicBlock("omp_offload.failed");
     llvm::BasicBlock *OffloadContBlock =
         CGF.createBasicBlock("omp_offload.cont");
llvm::Function *RequiresRegFn;
     // for mismatching requires clauses across compilation units that don't
     // contain at least 1 target region.
     assert((HasEmittedTargetRegion ||
             HasEmittedDeclareTargetRegion ||
             !OffloadEntriesInfoManager.empty()) &&
            "Target or declare target region expected.");
     if (HasRequiresUnifiedSharedMemory)
       Flags = OMP_REQ_UNIFIED_SHARED_MEMORY;
-    CGF.EmitRuntimeCall(createRuntimeFunction(OMPRTL__tgt_register_requires),
+    CGF.EmitRuntimeCall(llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+                            CGM.getModule(), OMPRTL___tgt_register_requires),
                         llvm::ConstantInt::get(CGM.Int64Ty, Flags));
     CGF.FinishFunction();
   }
   return RequiresRegFn;
 }
 
 void CGOpenMPRuntime::emitTeamsCall(CodeGenFunction &CGF,
                                     const OMPExecutableDirective &D,
                                     SourceLocation Loc,
void CGOpenMPRuntime::emitTeamsCall(CodeGenFunction &CGF,
   llvm::Value *Args[] = {
       RTLoc,
       CGF.Builder.getInt32(CapturedVars.size()), // Number of captured vars
       CGF.Builder.CreateBitCast(OutlinedFn, getKmpc_MicroPointerTy())};
   llvm::SmallVector<llvm::Value *, 16> RealArgs;
   RealArgs.append(std::begin(Args), std::end(Args));
   RealArgs.append(CapturedVars.begin(), CapturedVars.end());
 
-  llvm::FunctionCallee RTLFn = createRuntimeFunction(OMPRTL__kmpc_fork_teams);
+  llvm::FunctionCallee RTLFn =
+      llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+          CGM.getModule(), OMPRTL___kmpc_fork_teams);
   CGF.EmitRuntimeCall(RTLFn, RealArgs);
 }
 
 void CGOpenMPRuntime::emitNumTeamsClause(CodeGenFunction &CGF,
                                          const Expr *NumTeams,
                                          const Expr *ThreadLimit,
                                          SourceLocation Loc) {
   if (!CGF.HaveInsertPoint())
llvm::Value *ThreadLimitVal =
       ThreadLimit
           ? CGF.Builder.CreateIntCast(CGF.EmitScalarExpr(ThreadLimit),
                                       CGF.CGM.Int32Ty, /* isSigned = */ true)
           : CGF.Builder.getInt32(0);
 
   // Build call __kmpc_push_num_teamss(&loc, global_tid, num_teams, thread_limit)
   llvm::Value *PushNumTeamsArgs[] = {RTLoc, getThreadID(CGF, Loc), NumTeamsVal,
                                      ThreadLimitVal};
-  CGF.EmitRuntimeCall(createRuntimeFunction(OMPRTL__kmpc_push_num_teams),
+  CGF.EmitRuntimeCall(llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+                          CGM.getModule(), OMPRTL___kmpc_push_num_teams),
                       PushNumTeamsArgs);
 }
 
 void CGOpenMPRuntime::emitTargetDataCalls(
     CodeGenFunction &CGF, const OMPExecutableDirective &D, const Expr *IfCond,
     const Expr *Device, const RegionCodeGenTy &CodeGen, TargetDataInfo &Info) {
   if (!CGF.HaveInsertPoint())
     return;
auto &&BeginThenGen = [this, &D, Device, &Info,
     }
 
     // Emit the number of elements in the offloading arrays.
     llvm::Value *PointerNum = CGF.Builder.getInt32(Info.NumberOfPtrs);
 
     llvm::Value *OffloadingArgs[] = {
         DeviceID,         PointerNum,    BasePointersArrayArg,
         PointersArrayArg, SizesArrayArg, MapTypesArrayArg};
-    CGF.EmitRuntimeCall(createRuntimeFunction(OMPRTL__tgt_target_data_begin),
+    CGF.EmitRuntimeCall(llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+                            CGM.getModule(), OMPRTL___tgt_target_data_begin),
                         OffloadingArgs);
 
     // If device pointer privatization is required, emit the body of the region
     // here. It will have to be duplicated: with and without privatization.
     if (!Info.CaptureDeviceAddrMap.empty())
       CodeGen(CGF);
   };
 
auto &&EndThenGen = [this, Device, &Info](CodeGenFunction &CGF,
     }
 
     // Emit the number of elements in the offloading arrays.
     llvm::Value *PointerNum = CGF.Builder.getInt32(Info.NumberOfPtrs);
 
     llvm::Value *OffloadingArgs[] = {
         DeviceID,         PointerNum,    BasePointersArrayArg,
         PointersArrayArg, SizesArrayArg, MapTypesArrayArg};
-    CGF.EmitRuntimeCall(createRuntimeFunction(OMPRTL__tgt_target_data_end),
+    CGF.EmitRuntimeCall(llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+                            CGM.getModule(), OMPRTL___tgt_target_data_end),
                         OffloadingArgs);
   };
 
   // If we need device pointer privatization, we need to emit the body of the
   // region with no privatization in the 'else' branch of the conditional.
   // Otherwise, we don't have to do anything.
   auto &&BeginElseGen = [&Info, &CodeGen, &NoPrivAction](CodeGenFunction &CGF,
                                                          PrePostActionTy &) {
llvm::Value *OffloadingArgs[] = {DeviceID,
                                      InputInfo.BasePointersArray.getPointer(),
                                      InputInfo.PointersArray.getPointer(),
                                      InputInfo.SizesArray.getPointer(),
                                      MapTypesArray};
 
     // Select the right runtime function call for each expected standalone
     // directive.
     const bool HasNowait = D.hasClausesOfKind<OMPNowaitClause>();
-    OpenMPRTLFunction RTLFn;
+    RuntimeFunction RTLFn;
     switch (D.getDirectiveKind()) {
     case OMPD_target_enter_data:
-      RTLFn = HasNowait ? OMPRTL__tgt_target_data_begin_nowait
-                        : OMPRTL__tgt_target_data_begin;
+      RTLFn = HasNowait ? OMPRTL___tgt_target_data_begin_nowait
+                        : OMPRTL___tgt_target_data_begin;
       break;
     case OMPD_target_exit_data:
-      RTLFn = HasNowait ? OMPRTL__tgt_target_data_end_nowait
-                        : OMPRTL__tgt_target_data_end;
+      RTLFn = HasNowait ? OMPRTL___tgt_target_data_end_nowait
+                        : OMPRTL___tgt_target_data_end;
       break;
     case OMPD_target_update:
-      RTLFn = HasNowait ? OMPRTL__tgt_target_data_update_nowait
-                        : OMPRTL__tgt_target_data_update;
+      RTLFn = HasNowait ? OMPRTL___tgt_target_data_update_nowait
+                        : OMPRTL___tgt_target_data_update;
       break;
     case OMPD_parallel:
     case OMPD_for:
     case OMPD_parallel_for:
     case OMPD_parallel_master:
     case OMPD_parallel_sections:
     case OMPD_for_simd:
     case OMPD_parallel_for_simd:
auto &&ThenGen = [this, &D, Device, &InputInfo,
     case OMPD_target_parallel:
     case OMPD_target_parallel_for:
     case OMPD_target_parallel_for_simd:
     case OMPD_requires:
     case OMPD_unknown:
       llvm_unreachable("Unexpected standalone target data directive.");
       break;
     }
-    CGF.EmitRuntimeCall(createRuntimeFunction(RTLFn), OffloadingArgs);
+    CGF.EmitRuntimeCall(llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+                            CGM.getModule(), RTLFn),
+                        OffloadingArgs);
   };
 
   auto &&TargetThenGen = [this, &ThenGen, &D, &InputInfo, &MapTypesArray](
                              CodeGenFunction &CGF, PrePostActionTy &) {
     // Fill up the arrays with all the mapped variables.
     MappableExprsHandler::MapBaseValuesArrayTy BasePointers;
     MappableExprsHandler::MapValuesArrayTy Pointers;
     MappableExprsHandler::MapValuesArrayTy Sizes;
llvm::Value *Args[] = {
       emitUpdateLocation(CGF, D.getBeginLoc()),
       getThreadID(CGF, D.getBeginLoc()),
       llvm::ConstantInt::getSigned(CGM.Int32Ty, NumIterations.size()),
       CGF.Builder.CreatePointerBitCastOrAddrSpaceCast(
           CGF.Builder.CreateConstArrayGEP(DimsAddr, 0).getPointer(),
           CGM.VoidPtrTy)};
 
   llvm::FunctionCallee RTLFn =
-      createRuntimeFunction(OMPRTL__kmpc_doacross_init);
+      llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+          CGM.getModule(), OMPRTL___kmpc_doacross_init);
   CGF.EmitRuntimeCall(RTLFn, Args);
   llvm::Value *FiniArgs[DoacrossCleanupTy::DoacrossFinArgs] = {
       emitUpdateLocation(CGF, D.getEndLoc()), getThreadID(CGF, D.getEndLoc())};
   llvm::FunctionCallee FiniRTLFn =
-      createRuntimeFunction(OMPRTL__kmpc_doacross_fini);
+      llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+          CGM.getModule(), OMPRTL___kmpc_doacross_fini);
   CGF.EHStack.pushCleanup<DoacrossCleanupTy>(NormalAndEHCleanup, FiniRTLFn,
                                              llvm::makeArrayRef(FiniArgs));
 }
 
 void CGOpenMPRuntime::emitDoacrossOrdered(CodeGenFunction &CGF,
                                           const OMPDependClause *C) {
   QualType Int64Ty =
       CGM.getContext().getIntTypeForBitwidth(/*DestWidth=*/64, /*Signed=*/1);
CGF.EmitStoreOfScalar(CntVal, CGF.Builder.CreateConstArrayGEP(CntAddr, I),
                           /*Volatile=*/false, Int64Ty);
   }
   llvm::Value *Args[] = {
       emitUpdateLocation(CGF, C->getBeginLoc()),
       getThreadID(CGF, C->getBeginLoc()),
       CGF.Builder.CreateConstArrayGEP(CntAddr, 0).getPointer()};
   llvm::FunctionCallee RTLFn;
   if (C->getDependencyKind() == OMPC_DEPEND_source) {
-    RTLFn = createRuntimeFunction(OMPRTL__kmpc_doacross_post);
+    RTLFn = llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+        CGM.getModule(), OMPRTL___kmpc_doacross_post);
   } else {
     assert(C->getDependencyKind() == OMPC_DEPEND_sink);
-    RTLFn = createRuntimeFunction(OMPRTL__kmpc_doacross_wait);
+    RTLFn = llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+        CGM.getModule(), OMPRTL___kmpc_doacross_wait);
   }
   CGF.EmitRuntimeCall(RTLFn, Args);
 }
 
 void CGOpenMPRuntime::emitCall(CodeGenFunction &CGF, SourceLocation Loc,
                                llvm::FunctionCallee Callee,
                                ArrayRef<llvm::Value *> Args) const {
   assert(Loc.isValid() && "Outlined function call location must be valid.");
Address CGOpenMPRuntime::getAddressOfLocalVariable(CodeGenFunction &CGF,
   if (Allocator->getType()->isIntegerTy())
     Allocator = CGF.Builder.CreateIntToPtr(Allocator, CGM.VoidPtrTy);
   else if (Allocator->getType()->isPointerTy())
     Allocator = CGF.Builder.CreatePointerBitCastOrAddrSpaceCast(Allocator,
                                                                 CGM.VoidPtrTy);
   llvm::Value *Args[] = {ThreadID, Size, Allocator};
 
   llvm::Value *Addr =
-      CGF.EmitRuntimeCall(createRuntimeFunction(OMPRTL__kmpc_alloc), Args,
-                          getName({CVD->getName(), ".void.addr"}));
+      CGF.EmitRuntimeCall(llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(
+                              CGM.getModule(), OMPRTL___kmpc_alloc),
+                          Args, getName({CVD->getName(), ".void.addr"}));
   llvm::Value *FiniArgs[OMPAllocateCleanupTy::CleanupArgs] = {ThreadID, Addr,
                                                               Allocator};
-  llvm::FunctionCallee FiniRTLFn = createRuntimeFunction(OMPRTL__kmpc_free);
+  llvm::FunctionCallee FiniRTLFn =
+      llvm::OpenMPIRBuilder::getOrCreateRuntimeFunction(CGM.getModule(),
+                                                        OMPRTL___kmpc_free);
 
   CGF.EHStack.pushCleanup<OMPAllocateCleanupTy>(NormalAndEHCleanup, FiniRTLFn,
                                                 llvm::makeArrayRef(FiniArgs));
   Addr = CGF.Builder.CreatePointerBitCastOrAddrSpaceCast(
       Addr,
       CGF.ConvertTypeForMem(CGM.getContext().getPointerType(CVD->getType())),
       getName({CVD->getName(), ".addr"}));
   return Address(Addr, Align);

llvm/include/llvm/Frontend/OpenMP/OMPIRBuilder.h

public:
   /// potentially other helpers into the underlying module. Must be called
   /// before any other method and only once!
   void initialize();
 
   /// Finalize the underlying module, e.g., by outlining regions.
   void finalize();
 
   /// Add attributes known for \p FnID to \p Fn.
-  void addAttributes(omp::RuntimeFunction FnID, Function &Fn);
+  static void addAttributes(omp::RuntimeFunction FnID, Function &Fn);
 
   /// Type used throughout for insertion points.
   using InsertPointTy = IRBuilder<>::InsertPoint;
 
   /// Callback type for variable finalization (think destructors).
   ///
   /// \param CodeGenIP is the insertion point at which the finalization code
   ///                  should be placed.
public:
   /// Update the internal location to \p Loc.
   bool updateToLocation(const LocationDescription &Loc) {
     Builder.restoreIP(Loc.IP);
     Builder.SetCurrentDebugLocation(Loc.DL);
     return Loc.IP.getBlock() != nullptr;
   }
 
   /// Return the function declaration for the runtime function with \p FnID.
-  Function *getOrCreateRuntimeFunction(omp::RuntimeFunction FnID);
+  static Function *getOrCreateRuntimeFunction(Module &Md,
+                                              omp::RuntimeFunction FnID);
 
   /// Return the (LLVM-IR) string describing the source location \p LocStr.
   Constant *getOrCreateSrcLocStr(StringRef LocStr);
 
   /// Return the (LLVM-IR) string describing the default source location.
   Constant *getOrCreateDefaultSrcLocStr();
 
   /// Return the (LLVM-IR) string describing the source location \p Loc.

llvm/include/llvm/Frontend/OpenMP/OMPKinds.def

 
 #define __OMP_PTR_TYPE(NAME, BASE) OMP_TYPE(NAME, BASE->getPointerTo())
 
 __OMP_PTR_TYPE(VoidPtr, Int8)
 __OMP_PTR_TYPE(VoidPtrPtr, VoidPtr)
 __OMP_PTR_TYPE(VoidPtrPtrPtr, VoidPtrPtr)
 
 // TODO: Replace this with the real size_t type
 #define __OMP_SIZE_TYPE(NAME) OMP_TYPE(NAME, Type::getInt64Ty(Ctx))
 __OMP_SIZE_TYPE(SizeTy)
 #undef __OMP_SIZE_TYPE
 
 #undef __OMP_PTR_TYPE
 
 #undef __OMP_TYPE
 #undef OMP_TYPE
 
 ///}
 
 /// array types
 ///
 ///{
#define __OMP_FUNCTION_TYPE(VarName, IsVarArg, ReturnType, ...)                \
   OMP_FUNCTION_TYPE(VarName, IsVarArg, ReturnType, __VA_ARGS__)
 
 __OMP_FUNCTION_TYPE(ParallelTask, true, Void, Int32Ptr, Int32Ptr)
 __OMP_FUNCTION_TYPE(ReduceFunction, false, Void, VoidPtr, VoidPtr)
 __OMP_FUNCTION_TYPE(CopyFunction, false, Void, VoidPtr, VoidPtr)
 __OMP_FUNCTION_TYPE(KmpcCtor, false, VoidPtr, VoidPtr)
 __OMP_FUNCTION_TYPE(KmpcDtor, false, Void, VoidPtr)
 __OMP_FUNCTION_TYPE(KmpcCopyCtor, false, VoidPtr, VoidPtr, VoidPtr)
-__OMP_FUNCTION_TYPE(TaskRoutineEntry, false, Int32, Int32, VoidPtr)
+__OMP_FUNCTION_TYPE(TaskRoutineEntry, false, Int32, Int32, 
+                    /* kmp_task_t */ VoidPtr)
 
 #undef __OMP_FUNCTION_TYPE
 #undef OMP_FUNCTION_TYPE
 
 ///}
 
 /// Runtime library function (and their attributes)
 ///
 ///{
 
 #ifndef OMP_RTL
 #define OMP_RTL(Enum, Str, IsVarArg, ReturnType, ...)
 #endif
 
 #define __OMP_RTL(Name, IsVarArg, ReturnType, ...)                             \
   OMP_RTL(OMPRTL_##Name, #Name, IsVarArg, ReturnType, __VA_ARGS__)
 
 __OMP_RTL(__kmpc_barrier, false, Void, IdentPtr, Int32)
 __OMP_RTL(__kmpc_cancel, false, Int32, IdentPtr, Int32, Int32)
 __OMP_RTL(__kmpc_cancel_barrier, false, Int32, IdentPtr, Int32)
 __OMP_RTL(__kmpc_flush, false, Void, IdentPtr)
 __OMP_RTL(__kmpc_global_thread_num, false, Int32, IdentPtr)
 __OMP_RTL(__kmpc_fork_call, true, Void, IdentPtr, Int32, ParallelTaskPtr)
 __OMP_RTL(__kmpc_omp_taskwait, false, Int32, IdentPtr, Int32)
-__OMP_RTL(__kmpc_omp_taskyield, false, Int32, IdentPtr, Int32, Int32)
+__OMP_RTL(__kmpc_omp_taskyield, false, Int32, IdentPtr, Int32, /* Int */ Int32)
 __OMP_RTL(__kmpc_push_num_threads, false, Void, IdentPtr, Int32,
           /* Int */ Int32)
 __OMP_RTL(__kmpc_push_proc_bind, false, Void, IdentPtr, Int32, /* Int */ Int32)
 __OMP_RTL(__kmpc_serialized_parallel, false, Void, IdentPtr, Int32)
 __OMP_RTL(__kmpc_end_serialized_parallel, false, Void, IdentPtr, Int32)
 
 __OMP_RTL(omp_get_thread_num, false, Int32, )
 __OMP_RTL(omp_get_num_threads, false, Int32, )

 __OMP_RTL(omp_set_nested, false, Void, Int32)
 __OMP_RTL(omp_set_schedule, false, Void, Int32, Int32)
 __OMP_RTL(omp_set_max_active_levels, false, Void, Int32)
 
 __OMP_RTL(__kmpc_master, false, Int32, IdentPtr, Int32)
 __OMP_RTL(__kmpc_end_master, false, Void, IdentPtr, Int32)
 __OMP_RTL(__kmpc_critical, false, Void, IdentPtr, Int32, KmpCriticalNamePtrTy)
 __OMP_RTL(__kmpc_critical_with_hint, false, Void, IdentPtr, Int32,
-          KmpCriticalNamePtrTy, Int32)
+          KmpCriticalNamePtrTy, /* int32? */ Int32)
 __OMP_RTL(__kmpc_end_critical, false, Void, IdentPtr, Int32,
           KmpCriticalNamePtrTy)
 
 __OMP_RTL(__kmpc_begin, false, Void, IdentPtr, Int32)
 __OMP_RTL(__kmpc_end, false, Void, IdentPtr)
 
 __OMP_RTL(__kmpc_reduce, false, Int32, IdentPtr, Int32, Int32, SizeTy, VoidPtr,
           ReduceFunctionPtr, KmpCriticalNamePtrTy)

 __OMP_RTL(__kmpc_dist_for_static_init_8, false, Void, IdentPtr, Int32, Int32,
           Int32Ptr, Int64Ptr, Int64Ptr, Int64Ptr, Int64Ptr, Int64, Int64)
 __OMP_RTL(__kmpc_dist_for_static_init_8u, false, Void, IdentPtr, Int32, Int32,
           Int32Ptr, Int64Ptr, Int64Ptr, Int64Ptr, Int64Ptr, Int64, Int64)
 
 __OMP_RTL(__kmpc_single, false, Int32, IdentPtr, Int32)
 __OMP_RTL(__kmpc_end_single, false, Void, IdentPtr, Int32)
 
-__OMP_RTL(__kmpc_omp_task_alloc, false, VoidPtr, IdentPtr, Int32, Int32, SizeTy,
-          SizeTy, TaskRoutineEntryPtr)
-__OMP_RTL(__kmpc_omp_task, false, Int32, IdentPtr, Int32, VoidPtr)
+__OMP_RTL(__kmpc_omp_task_alloc, false, /* kmp_task_t */ VoidPtr, IdentPtr,
+          Int32, Int32, SizeTy, SizeTy, TaskRoutineEntryPtr)
+__OMP_RTL(__kmpc_omp_task, false, Int32, IdentPtr, Int32, 
+          /* kmp_task_t */ VoidPtr)
 __OMP_RTL(__kmpc_end_taskgroup, false, Void, IdentPtr, Int32)
 __OMP_RTL(__kmpc_taskgroup, false, Void, IdentPtr, Int32)
-__OMP_RTL(__kmpc_omp_task_begin_if0, false, Void, IdentPtr, Int32, VoidPtr)
-__OMP_RTL(__kmpc_omp_task_complete_if0, false, Void, IdentPtr, Int32, VoidPtr)
-__OMP_RTL(__kmpc_omp_task_with_deps, false, Int32, IdentPtr, Int32, VoidPtr,
-          Int32, VoidPtr, Int32, VoidPtr)
-__OMP_RTL(__kmpc_taskloop, false, Void, IdentPtr, Int32, VoidPtr, Int32,
-          Int64Ptr, Int64Ptr, Int64, Int32, Int32, Int64, VoidPtr)
-__OMP_RTL(__kmpc_task_allow_completion_event, false, VoidPtr, IdentPtr, Int32,
-          VoidPtr)
-__OMP_RTL(__kmpc_omp_target_task_alloc, false, VoidPtr, IdentPtr, Int32, Int32,
-          SizeTy, SizeTy, VoidPtr, Int64)
-__OMP_RTL(__kmpc_taskred_modifier_init, false, VoidPtr, VoidPtr, Int32, Int32,
-          Int32, VoidPtr)
-__OMP_RTL(__kmpc_taskred_init, false, VoidPtr, Int32, Int32, VoidPtr)
-__OMP_RTL(__kmpc_task_reduction_modifier_fini, false, Void, VoidPtr, Int32,
-          Int32)
-__OMP_RTL(__kmpc_task_reduction_get_th_data, false, VoidPtr, Int32, VoidPtr,
+__OMP_RTL(__kmpc_omp_task_begin_if0, false, Void, IdentPtr, Int32, 
+          /* kmp_task_t */ VoidPtr)
+__OMP_RTL(__kmpc_omp_task_complete_if0, false, Void, IdentPtr, Int32, 
+          /* kmp_task_t */ VoidPtr)
+__OMP_RTL(__kmpc_omp_task_with_deps, false, Int32, IdentPtr, Int32, 
+          /* kmp_task_t */ VoidPtr, Int32,
+          /* kmp_depend_info_t */ VoidPtr, Int32,
+          /* kmp_depend_info_t */ VoidPtr)
+__OMP_RTL(__kmpc_taskloop, false, Void, IdentPtr, /* Int */ Int32, VoidPtr, 
+          /* Int */ Int32, Int64Ptr, Int64Ptr, Int64, /* Int */ Int32, 
+          /* Int */ Int32, Int64, VoidPtr)
+__OMP_RTL(__kmpc_omp_target_task_alloc, false, /* kmpc_task_t */ VoidPtr,
+          IdentPtr, Int32, Int32, SizeTy, SizeTy, TaskRoutineEntryPtr, Int64)
+__OMP_RTL(__kmpc_taskred_modifier_init, false, VoidPtr, VoidPtr, 
+          /* Int */ Int32, /* Int */ Int32, /* Int */ Int32, VoidPtr)
+__OMP_RTL(__kmpc_taskred_init, false, VoidPtr, /* Int */ Int32, /* Int */ Int32,
           VoidPtr)
+__OMP_RTL(__kmpc_task_reduction_modifier_fini, false, Void, IdentPtr,
+          /* Int */ Int32, /* Int */ Int32)
+__OMP_RTL(__kmpc_task_reduction_get_th_data, false, VoidPtr, /* Int */ Int32,
+          VoidPtr, VoidPtr)
 __OMP_RTL(__kmpc_task_reduction_init, false, VoidPtr, Int32, Int32, VoidPtr)
 __OMP_RTL(__kmpc_task_reduction_modifier_init, false, VoidPtr, VoidPtr, Int32,
           Int32, Int32, VoidPtr)
 __OMP_RTL(__kmpc_proxy_task_completed_ooo, false, Void, VoidPtr)
 
-__OMP_RTL(__kmpc_omp_wait_deps, false, Void, IdentPtr, Int32, Int32, VoidPtr,
-          Int32, VoidPtr)
+__OMP_RTL(__kmpc_omp_wait_deps, false, Void, IdentPtr, Int32, Int32, 
+          /* kmp_depend_info_t */ VoidPtr, Int32, 
+          /* kmp_depend_info_t */ VoidPtr)
 __OMP_RTL(__kmpc_cancellationpoint, false, Int32, IdentPtr, Int32, Int32)
 
 __OMP_RTL(__kmpc_fork_teams, true, Void, IdentPtr, Int32, ParallelTaskPtr)
-__OMP_RTL(__kmpc_push_num_teams, false, Void, IdentPtr, Int32, Int32, Int32)
+__OMP_RTL(__kmpc_push_num_teams, false, /* Void? */ Int32, IdentPtr, Int32, 
+          Int32, Int32)
 
 __OMP_RTL(__kmpc_copyprivate, false, Void, IdentPtr, Int32, SizeTy, VoidPtr,
           CopyFunctionPtr, Int32)
 __OMP_RTL(__kmpc_threadprivate_cached, false, VoidPtr, IdentPtr, Int32, VoidPtr,
           SizeTy, VoidPtrPtrPtr)
 __OMP_RTL(__kmpc_threadprivate_register, false, Void, IdentPtr, VoidPtr,
           KmpcCtorPtr, KmpcCopyCtorPtr, KmpcDtorPtr)
 
-__OMP_RTL(__kmpc_doacross_init, false, Void, IdentPtr, Int32, Int32, VoidPtr)
+__OMP_RTL(__kmpc_doacross_init, false, Void, IdentPtr, Int32, Int32, 
+          /* kmp_dim */ VoidPtr)
 __OMP_RTL(__kmpc_doacross_post, false, Void, IdentPtr, Int32, Int64Ptr)
 __OMP_RTL(__kmpc_doacross_wait, false, Void, IdentPtr, Int32, Int64Ptr)
 __OMP_RTL(__kmpc_doacross_fini, false, Void, IdentPtr, Int32)
 
-__OMP_RTL(__kmpc_alloc, false, VoidPtr, Int32, SizeTy, VoidPtr)
-__OMP_RTL(__kmpc_free, false, Void, Int32, VoidPtr, VoidPtr)
+__OMP_RTL(__kmpc_alloc, false, VoidPtr, /* Int */ Int32, SizeTy, VoidPtr)
+__OMP_RTL(__kmpc_free, false, Void, /* Int */ Int32, VoidPtr, VoidPtr)
+
+//
+// Offloading related calls
+//
+
+__OMP_RTL(__kmpc_push_target_tripcount, false, Void, Int64, Int64)
+__OMP_RTL(__tgt_target, false, Int32, Int64, VoidPtr, Int32, VoidPtrPtr,
+          VoidPtrPtr, Int64Ptr, Int64Ptr)
+__OMP_RTL(__tgt_target_nowait, false, Int32, Int64, VoidPtr, Int32, VoidPtrPtr,
+          VoidPtrPtr, Int64Ptr, Int64Ptr)
+__OMP_RTL(__tgt_target_teams, false, Int32, Int64, VoidPtr, Int32, VoidPtrPtr,
+          VoidPtrPtr, Int64Ptr, Int64Ptr, Int32, Int32)
+__OMP_RTL(__tgt_target_teams_nowait, false, Int32, Int64, VoidPtr, Int32,
+          VoidPtrPtr, VoidPtrPtr, Int64Ptr, Int64Ptr, Int32, Int32)
+__OMP_RTL(__tgt_register_requires, false, Void, Int64)
+__OMP_RTL(__tgt_target_data_begin, false, Void, Int64, Int32, VoidPtrPtr,
+          VoidPtrPtr, Int64Ptr, Int64Ptr)
+__OMP_RTL(__tgt_target_data_begin_nowait, false, Void, Int64, Int32, VoidPtrPtr,
+          VoidPtrPtr, Int64Ptr, Int64Ptr)
+__OMP_RTL(__tgt_target_data_end, false, Void, Int64, Int32, VoidPtrPtr,
+          VoidPtrPtr, Int64Ptr, Int64Ptr)
+__OMP_RTL(__tgt_target_data_end_nowait, false, Void, Int64, Int32, VoidPtrPtr,
+          VoidPtrPtr, Int64Ptr, Int64Ptr)
+__OMP_RTL(__tgt_target_data_update, false, Void, Int64, Int32, VoidPtrPtr,
+          VoidPtrPtr, Int64Ptr, Int64Ptr)
+__OMP_RTL(__tgt_target_data_update_nowait, false, Void, Int64, Int32,
+          VoidPtrPtr, VoidPtrPtr, Int64Ptr, Int64Ptr)
+__OMP_RTL(__tgt_mapper_num_components, false, Int64, VoidPtr)
+__OMP_RTL(__tgt_push_mapper_component, false, Void, VoidPtr, VoidPtr, VoidPtr,
+          Int64, Int64)
+__OMP_RTL(__kmpc_task_allow_completion_event, false, VoidPtr, IdentPtr, 
+          /* Int */ Int32, /* kmp_task_t */ VoidPtr)
 
 __OMP_RTL(__last, false, Void, )
 
 #undef __OMP_RTL
 #undef OMP_RTL
 
 #define EnumAttr(Kind) Attribute::get(Ctx, Attribute::AttrKind::Kind)
 #define AttributeSet(...)                                                      \

 __OMP_RTL_ATTRS(__kmpc_omp_task_begin_if0, AttributeSet(EnumAttr(NoUnwind)),
                 AttributeSet(), {})
 __OMP_RTL_ATTRS(__kmpc_omp_task_complete_if0, AttributeSet(EnumAttr(NoUnwind)),
                 AttributeSet(), {})
 __OMP_RTL_ATTRS(__kmpc_omp_task_with_deps, AttributeSet(EnumAttr(NoUnwind)),
                 AttributeSet(), {})
 __OMP_RTL_ATTRS(__kmpc_taskloop, AttributeSet(EnumAttr(NoUnwind)),
                 AttributeSet(), {})
-__OMP_RTL_ATTRS(__kmpc_task_allow_completion_event,
-                AttributeSet(EnumAttr(NoUnwind)),
-                AttributeSet(), {})
 __OMP_RTL_ATTRS(__kmpc_omp_target_task_alloc,
                 AttributeSet(EnumAttr(NoUnwind)),
                 AttributeSet(), {})
 __OMP_RTL_ATTRS(__kmpc_taskred_modifier_init,
                 AttributeSet(EnumAttr(NoUnwind)),
                 AttributeSet(), {})
 __OMP_RTL_ATTRS(__kmpc_taskred_init,
                 AttributeSet(EnumAttr(NoUnwind)),

 __OMP_RTL_ATTRS(__kmpc_doacross_fini, AttributeSet(EnumAttr(NoUnwind)),
                 AttributeSet(), {})
 
 __OMP_RTL_ATTRS(__kmpc_alloc, AttributeSet(EnumAttr(NoUnwind)),
                 AttributeSet(), {})
 __OMP_RTL_ATTRS(__kmpc_free, AttributeSet(EnumAttr(NoUnwind)),
                 AttributeSet(), {})
 
+__OMP_RTL_ATTRS(__kmpc_push_target_tripcount,
+                AttributeSet(EnumAttr(NoUnwind)),
+                AttributeSet(), {})
+__OMP_RTL_ATTRS(__tgt_target,
+                AttributeSet(EnumAttr(NoUnwind)),
+                AttributeSet(), {})
+__OMP_RTL_ATTRS(__tgt_target_nowait,
+                AttributeSet(EnumAttr(NoUnwind)),
+                AttributeSet(), {})
+__OMP_RTL_ATTRS(__tgt_target_teams,
+                AttributeSet(EnumAttr(NoUnwind)),
+                AttributeSet(), {})
+__OMP_RTL_ATTRS(__tgt_target_teams_nowait,
+                AttributeSet(EnumAttr(NoUnwind)),
+                AttributeSet(), {})
+__OMP_RTL_ATTRS(__tgt_register_requires,
+                AttributeSet(EnumAttr(NoUnwind)),
+                AttributeSet(), {})
+__OMP_RTL_ATTRS(__tgt_target_data_begin,
+                AttributeSet(EnumAttr(NoUnwind)),
+                AttributeSet(), {})
+__OMP_RTL_ATTRS(__tgt_target_data_begin_nowait,
+                AttributeSet(EnumAttr(NoUnwind)),
+                AttributeSet(), {})
+__OMP_RTL_ATTRS(__tgt_target_data_end,
+                AttributeSet(EnumAttr(NoUnwind)),
+                AttributeSet(), {})
+__OMP_RTL_ATTRS(__tgt_target_data_end_nowait,
+                AttributeSet(EnumAttr(NoUnwind)),
+                AttributeSet(), {})
+__OMP_RTL_ATTRS(__tgt_target_data_update,
+                AttributeSet(EnumAttr(NoUnwind)),
+                AttributeSet(), {})
+__OMP_RTL_ATTRS(__tgt_target_data_update_nowait,
+                AttributeSet(EnumAttr(NoUnwind)),
+                AttributeSet(), {})
+__OMP_RTL_ATTRS(__tgt_mapper_num_components,
+                AttributeSet(EnumAttr(NoUnwind)),
+                AttributeSet(), {})
+__OMP_RTL_ATTRS(__tgt_push_mapper_component,
+                AttributeSet(EnumAttr(NoUnwind)),
+                AttributeSet(), {})
+__OMP_RTL_ATTRS(__kmpc_task_allow_completion_event,
+                AttributeSet(EnumAttr(NoUnwind)),
+                AttributeSet(), {})
+
 #undef __OMP_RTL_ATTRS
 #undef OMP_RTL_ATTRS
 #undef AttributeSet
 #undef EnumAttr
 
 ///}
 
 /// KMP ident_t bit flags

llvm/lib/Frontend/OpenMP/OMPIRBuilder.cpp

case Enum:                                                                   \
     break;
 #include "llvm/Frontend/OpenMP/OMPKinds.def"
   default:
     // Attributes are optional.
     break;
   }
 }
 
-Function *OpenMPIRBuilder::getOrCreateRuntimeFunction(RuntimeFunction FnID) {
+Function *OpenMPIRBuilder::getOrCreateRuntimeFunction(Module &Md,
+                                                      RuntimeFunction FnID) {
   Function *Fn = nullptr;
 
   // Try to find the declation in the module first.
   switch (FnID) {
 #define OMP_RTL(Enum, Str, IsVarArg, ReturnType, ...)                          \
   case Enum:                                                                   \
-    Fn = M.getFunction(Str);                                                   \
+    Fn = Md.getFunction(Str);                                                  \
     break;
 #include "llvm/Frontend/OpenMP/OMPKinds.def"
   }
 
   if (!Fn) {
     // Create a new declaration if we need one.
     switch (FnID) {
 #define OMP_RTL(Enum, Str, IsVarArg, ReturnType, ...)                          \
   case Enum:                                                                   \
     Fn = Function::Create(FunctionType::get(ReturnType,                        \
                                             ArrayRef<Type *>{__VA_ARGS__},     \
                                             IsVarArg),                         \
-                          GlobalValue::ExternalLinkage, Str, M);               \
+                          GlobalValue::ExternalLinkage, Str, Md);              \
     break;
 #include "llvm/Frontend/OpenMP/OMPKinds.def"
     }
 
+    // Add information if the runtime function takes a callback function
+    if (Fn->getName() == "__kmpc_fork_call") {
+      if (!Fn->hasMetadata(LLVMContext::MD_callback)) {
+        LLVMContext &Ctx = Fn->getContext();
+        MDBuilder MDB(Ctx);
+        // Annotate the callback behavior of the __kmpc_fork_call:
+        //  - The callback callee is argument number 2 (microtask).
+        //  - The first two arguments of the callback callee are unknown (-1).
+        //  - All variadic arguments to the __kmpc_fork_call are passed to the
+        //    callback callee.
+        Fn->addMetadata(
+            LLVMContext::MD_callback,
+            *MDNode::get(Ctx, {MDB.createCallbackEncoding(
+                                  2, {-1, -1}, /* VarArgsArePassed */ true)}));
+      }
+    } else if (Fn->getName() == "__kmpc_fork_teams") {
+      if (!Fn->hasMetadata(LLVMContext::MD_callback)) {
+        LLVMContext &Ctx = Fn->getContext();
+        MDBuilder MDB(Ctx);
+        // Annotate the callback behavior of the __kmpc_fork_teams:
+        //  - The callback callee is argument number 2 (microtask).
+        //  - The first two arguments of the callback callee are unknown (-1).
+        //  - All variadic arguments to the __kmpc_fork_teams are passed to the
+        //    callback callee.
+        Fn->addMetadata(
+            LLVMContext::MD_callback,
+            *MDNode::get(Ctx, {MDB.createCallbackEncoding(
+                                  2, {-1, -1}, /* VarArgsArePassed */ true)}));
+      }
+    }
+
     addAttributes(FnID, *Fn);
   }
 
   assert(Fn && "Failed to create OpenMP runtime function");
   return Fn;
 }
 
 void OpenMPIRBuilder::initialize() { initializeTypes(M); }
 
 void OpenMPIRBuilder::finalize() {
   for (OutlineInfo &OI : OutlineInfos) {
     assert(!OI.Blocks.empty() &&
            "Outlined regions should have at least a single block!");
     BasicBlock *RegEntryBB = OI.Blocks.front();
     Function *OuterFn = RegEntryBB->getParent();
     CodeExtractorAnalysisCache CEAC(*OuterFn);
     CodeExtractor Extractor(OI.Blocks, /* DominatorTree */ nullptr,
                             /* AggregateArgs */ false,
                             /* BlockFrequencyInfo */ nullptr,
                             /* BranchProbabilityInfo */ nullptr,
                             /* AssumptionCache */ nullptr,
                             /* AllowVarArgs */ true,
                             /* AllowAlloca */ true,
                             /* Suffix */ ".omp_par");
 
     LLVM_DEBUG(dbgs() << "Before     outlining: " << *OuterFn << "\n");
OpenMPIRBuilder::getOrCreateSrcLocStr(const LocationDescription &Loc) {
   std::stringstream SrcLocStr;
   SrcLocStr << ";" << Filename.data() << ";" << Function.data() << ";"
             << LineStr << ";" << ColumnStr << ";;";
   return getOrCreateSrcLocStr(SrcLocStr.str());
 }
 
 Value *OpenMPIRBuilder::getOrCreateThreadID(Value *Ident) {
   return Builder.CreateCall(
-      getOrCreateRuntimeFunction(OMPRTL___kmpc_global_thread_num), Ident,
+      getOrCreateRuntimeFunction(M, OMPRTL___kmpc_global_thread_num), Ident,
       "omp_global_thread_num");
 }
 
 OpenMPIRBuilder::InsertPointTy
 OpenMPIRBuilder::CreateBarrier(const LocationDescription &Loc, Directive DK,
                                bool ForceSimpleCall, bool CheckCancelFlag) {
   if (!updateToLocation(Loc))
     return Loc.IP;
Value *Args[] = {getOrCreateIdent(SrcLocStr, BarrierLocFlags),
                    getOrCreateThreadID(getOrCreateIdent(SrcLocStr))};
 
   // If we are in a cancellable parallel region, barriers are cancellation
   // points.
   // TODO: Check why we would force simple calls or to ignore the cancel flag.
   bool UseCancelBarrier =
       !ForceSimpleCall && isLastFinalizationInfoCancellable(OMPD_parallel);
 
-  Value *Result = Builder.CreateCall(
-      getOrCreateRuntimeFunction(UseCancelBarrier ? OMPRTL___kmpc_cancel_barrier
-                                                  : OMPRTL___kmpc_barrier),
+  Value *Result =
+      Builder.CreateCall(getOrCreateRuntimeFunction(
+                             M, UseCancelBarrier ? OMPRTL___kmpc_cancel_barrier
                          Args);
 
   if (UseCancelBarrier && CheckCancelFlag)
     emitCancelationCheckImpl(Result, OMPD_parallel);
 
   return Builder.saveIP();
 }
 
 OpenMPIRBuilder::InsertPointTy
#include "llvm/Frontend/OpenMP/OMPKinds.def"
   default:
     llvm_unreachable("Unknown cancel kind!");
   }
 
   Constant *SrcLocStr = getOrCreateSrcLocStr(Loc);
   Value *Ident = getOrCreateIdent(SrcLocStr);
   Value *Args[] = {Ident, getOrCreateThreadID(Ident), CancelKind};
   Value *Result = Builder.CreateCall(
-      getOrCreateRuntimeFunction(OMPRTL___kmpc_cancel), Args);
+      getOrCreateRuntimeFunction(M, OMPRTL___kmpc_cancel), Args);
 
   // The actual cancel logic is shared with others, e.g., cancel_barriers.
   emitCancelationCheckImpl(Result, CanceledDirective);
 
   // Update the insertion point and remove the terminator we introduced.
   Builder.SetInsertPoint(UI->getParent());
   UI->eraseFromParent();
 
IRBuilder<>::InsertPoint OpenMPIRBuilder::CreateParallel(
   Value *ThreadID = getOrCreateThreadID(Ident);
 
   if (NumThreads) {
     // Build call __kmpc_push_num_threads(&Ident, global_tid, num_threads)
     Value *Args[] = {
         Ident, ThreadID,
         Builder.CreateIntCast(NumThreads, Int32, /*isSigned*/ false)};
     Builder.CreateCall(
-        getOrCreateRuntimeFunction(OMPRTL___kmpc_push_num_threads), Args);
+        getOrCreateRuntimeFunction(M, OMPRTL___kmpc_push_num_threads), Args);
   }
 
   if (ProcBind != OMP_PROC_BIND_default) {
     // Build call __kmpc_push_proc_bind(&Ident, global_tid, proc_bind)
     Value *Args[] = {
         Ident, ThreadID,
         ConstantInt::get(Int32, unsigned(ProcBind), /*isSigned=*/true)};
-    Builder.CreateCall(getOrCreateRuntimeFunction(OMPRTL___kmpc_push_proc_bind),
-                       Args);
+    Builder.CreateCall(
+        getOrCreateRuntimeFunction(M, OMPRTL___kmpc_push_proc_bind), Args);
   }
 
   BasicBlock *InsertBB = Builder.GetInsertBlock();
   Function *OuterFn = InsertBB->getParent();
 
   // Vector to remember instructions we used only during the modeling but which
   // we want to delete at the end.
   SmallVector<Instruction *, 4> ToBeDeleted;
IRBuilder<>::InsertPoint OpenMPIRBuilder::CreateParallel(
 
   // Let the caller create the body.
   assert(BodyGenCB && "Expected body generation callback!");
   InsertPointTy CodeGenIP(PRegBodyBB, PRegBodyBB->begin());
   BodyGenCB(AllocaIP, CodeGenIP, *PRegPreFiniBB);
 
   LLVM_DEBUG(dbgs() << "After  body codegen: " << *OuterFn << "\n");
 
-  FunctionCallee RTLFn = getOrCreateRuntimeFunction(OMPRTL___kmpc_fork_call);
+  FunctionCallee RTLFn = getOrCreateRuntimeFunction(M, OMPRTL___kmpc_fork_call);
   if (auto *F = dyn_cast<llvm::Function>(RTLFn.getCallee())) {
     if (!F->hasMetadata(llvm::LLVMContext::MD_callback)) {
       llvm::LLVMContext &Ctx = F->getContext();
       MDBuilder MDB(Ctx);
       // Annotate the callback behavior of the __kmpc_fork_call:
       //  - The callback callee is argument number 2 (microtask).
       //  - The first two arguments of the callback callee are unknown (-1).
       //  - All variadic arguments to the __kmpc_fork_call are passed to the
if (!ElseTI) {
 
       // If an "if" clause was present we are now generating the serialized
       // version into the "else" branch.
       Builder.SetInsertPoint(ElseTI);
 
       // Build calls __kmpc_serialized_parallel(&Ident, GTid);
       Value *SerializedParallelCallArgs[] = {Ident, ThreadID};
       Builder.CreateCall(
-          getOrCreateRuntimeFunction(OMPRTL___kmpc_serialized_parallel),
+          getOrCreateRuntimeFunction(M, OMPRTL___kmpc_serialized_parallel),
           SerializedParallelCallArgs);
 
       // OutlinedFn(&GTid, &zero, CapturedStruct);
       CI->removeFromParent();
       Builder.Insert(CI);
 
       // __kmpc_end_serialized_parallel(&Ident, GTid);
       Value *EndArgs[] = {Ident, ThreadID};
       Builder.CreateCall(
-          getOrCreateRuntimeFunction(OMPRTL___kmpc_end_serialized_parallel),
+          getOrCreateRuntimeFunction(M, OMPRTL___kmpc_end_serialized_parallel),
           EndArgs);
 
       LLVM_DEBUG(dbgs() << "With serialized parallel region: "
                         << *Builder.GetInsertBlock()->getParent() << "\n");
     }
 
     for (Instruction *I : ToBeDeleted)
       I->eraseFromParent();
IRBuilder<>::InsertPoint OpenMPIRBuilder::CreateParallel(
   BasicBlock *CommonExit = nullptr;
   SetVector<Value *> Inputs, Outputs, SinkingCands, HoistingCands;
   Extractor.findAllocas(CEAC, SinkingCands, HoistingCands, CommonExit);
   Extractor.findInputsOutputs(Inputs, Outputs, SinkingCands);
 
   LLVM_DEBUG(dbgs() << "Before privatization: " << *OuterFn << "\n");
 
   FunctionCallee TIDRTLFn =
-      getOrCreateRuntimeFunction(OMPRTL___kmpc_global_thread_num);
+      getOrCreateRuntimeFunction(M, OMPRTL___kmpc_global_thread_num);
 
   auto PrivHelper = [&](Value &V) {
     if (&V == TIDAddr || &V == ZeroAddr)
       return;
 
     SmallVector<Use *, 8> Uses;
     for (Use &U : V.uses())
       if (auto *UserI = dyn_cast<Instruction>(U.getUser()))
IRBuilder<>::InsertPoint OpenMPIRBuilder::CreateParallel(
   return AfterIP;
 }
 
 void OpenMPIRBuilder::emitFlush(const LocationDescription &Loc) {
   // Build call void __kmpc_flush(ident_t *loc)
   Constant *SrcLocStr = getOrCreateSrcLocStr(Loc);
   Value *Args[] = {getOrCreateIdent(SrcLocStr)};
 
-  Builder.CreateCall(getOrCreateRuntimeFunction(OMPRTL___kmpc_flush), Args);
+  Builder.CreateCall(getOrCreateRuntimeFunction(M, OMPRTL___kmpc_flush), Args);
 }
 
 void OpenMPIRBuilder::CreateFlush(const LocationDescription &Loc) {
   if (!updateToLocation(Loc))
     return;
   emitFlush(Loc);
 }
 
 void OpenMPIRBuilder::emitTaskwaitImpl(const LocationDescription &Loc) {
   // Build call kmp_int32 __kmpc_omp_taskwait(ident_t *loc, kmp_int32
   // global_tid);
   Constant *SrcLocStr = getOrCreateSrcLocStr(Loc);
   Value *Ident = getOrCreateIdent(SrcLocStr);
   Value *Args[] = {Ident, getOrCreateThreadID(Ident)};
 
   // Ignore return result until untied tasks are supported.
-  Builder.CreateCall(getOrCreateRuntimeFunction(OMPRTL___kmpc_omp_taskwait),
+  Builder.CreateCall(getOrCreateRuntimeFunction(M, OMPRTL___kmpc_omp_taskwait),
                      Args);
 }
 
 void OpenMPIRBuilder::CreateTaskwait(const LocationDescription &Loc) {
   if (!updateToLocation(Loc))
     return;
   emitTaskwaitImpl(Loc);
 }
 
 void OpenMPIRBuilder::emitTaskyieldImpl(const LocationDescription &Loc) {
   // Build call __kmpc_omp_taskyield(loc, thread_id, 0);
   Constant *SrcLocStr = getOrCreateSrcLocStr(Loc);
   Value *Ident = getOrCreateIdent(SrcLocStr);
   Constant *I32Null = ConstantInt::getNullValue(Int32);
   Value *Args[] = {Ident, getOrCreateThreadID(Ident), I32Null};
 
-  Builder.CreateCall(getOrCreateRuntimeFunction(OMPRTL___kmpc_omp_taskyield),
+  Builder.CreateCall(getOrCreateRuntimeFunction(M, OMPRTL___kmpc_omp_taskyield),
                      Args);
 }
 
 void OpenMPIRBuilder::CreateTaskyield(const LocationDescription &Loc) {
   if (!updateToLocation(Loc))
     return;
   emitTaskyieldImpl(Loc);
 }
 
 OpenMPIRBuilder::InsertPointTy
 OpenMPIRBuilder::CreateMaster(const LocationDescription &Loc,
                               BodyGenCallbackTy BodyGenCB,
                               FinalizeCallbackTy FiniCB) {
 
   if (!updateToLocation(Loc))
     return Loc.IP;
 
   Directive OMPD = Directive::OMPD_master;
   Constant *SrcLocStr = getOrCreateSrcLocStr(Loc);
   Value *Ident = getOrCreateIdent(SrcLocStr);
   Value *ThreadId = getOrCreateThreadID(Ident);
   Value *Args[] = {Ident, ThreadId};
 
-  Function *EntryRTLFn = getOrCreateRuntimeFunction(OMPRTL___kmpc_master);
+  Function *EntryRTLFn = getOrCreateRuntimeFunction(M, OMPRTL___kmpc_master);
   Instruction *EntryCall = Builder.CreateCall(EntryRTLFn, Args);
 
-  Function *ExitRTLFn = getOrCreateRuntimeFunction(OMPRTL___kmpc_end_master);
+  Function *ExitRTLFn = getOrCreateRuntimeFunction(M, OMPRTL___kmpc_end_master);
   Instruction *ExitCall = Builder.CreateCall(ExitRTLFn, Args);
 
   return EmitOMPInlinedRegion(OMPD, EntryCall, ExitCall, BodyGenCB, FiniCB,
                               /*Conditional*/ true, /*hasFinalize*/ true);
 }
 
 OpenMPIRBuilder::InsertPointTy OpenMPIRBuilder::CreateCritical(
     const LocationDescription &Loc, BodyGenCallbackTy BodyGenCB,
OpenMPIRBuilder::InsertPointTy OpenMPIRBuilder::CreateCritical(
   Value *LockVar = getOMPCriticalRegionLock(CriticalName);
   Value *Args[] = {Ident, ThreadId, LockVar};
 
   SmallVector<llvm::Value *, 4> EnterArgs(std::begin(Args), std::end(Args));
   Function *RTFn = nullptr;
   if (HintInst) {
     // Add Hint to entry Args and create call
     EnterArgs.push_back(HintInst);
-    RTFn = getOrCreateRuntimeFunction(OMPRTL___kmpc_critical_with_hint);
+    RTFn = getOrCreateRuntimeFunction(M, OMPRTL___kmpc_critical_with_hint);
   } else {
-    RTFn = getOrCreateRuntimeFunction(OMPRTL___kmpc_critical);
+    RTFn = getOrCreateRuntimeFunction(M, OMPRTL___kmpc_critical);
   }
   Instruction *EntryCall = Builder.CreateCall(RTFn, EnterArgs);
 
-  Function *ExitRTLFn = getOrCreateRuntimeFunction(OMPRTL___kmpc_end_critical);
+  Function *ExitRTLFn =
+      getOrCreateRuntimeFunction(M, OMPRTL___kmpc_end_critical);
   Instruction *ExitCall = Builder.CreateCall(ExitRTLFn, Args);
 
   return EmitOMPInlinedRegion(OMPD, EntryCall, ExitCall, BodyGenCB, FiniCB,
                               /*Conditional*/ false, /*hasFinalize*/ true);
 }
 
 OpenMPIRBuilder::InsertPointTy OpenMPIRBuilder::EmitOMPInlinedRegion(
     Directive OMPD, Instruction *EntryCall, Instruction *ExitCall,

llvm/test/Transforms/OpenMP/add_attributes.ll

 declare void @__kmpc_omp_task_complete_if0(%struct.ident_t*, i32, i8*)
 
 declare i32 @__kmpc_omp_task_with_deps(%struct.ident_t*, i32, i8*, i32, i8*, i32, i8*)
 
 declare void @__kmpc_omp_wait_deps(%struct.ident_t*, i32, i32, i8*, i32, i8*)
 
 declare i32 @__kmpc_cancellationpoint(%struct.ident_t*, i32, i32)
 
-declare void @__kmpc_push_num_teams(%struct.ident_t*, i32, i32, i32)
+declare i32 @__kmpc_push_num_teams(%struct.ident_t*, i32, i32, i32)
 
 declare void @__kmpc_fork_teams(%struct.ident_t*, i32, void (i32*, i32*, ...)*, ...)
 
 declare void @__kmpc_taskloop(%struct.ident_t*, i32, i8*, i32, i64*, i64*, i64, i32, i32, i64, i8*)
 
-declare i8* @__kmpc_omp_target_task_alloc(%struct.ident_t*, i32, i32, i64, i64, i8*, i64)
+declare i8* @__kmpc_omp_target_task_alloc(%struct.ident_t*, i32, i32, i64, i64, i32 (i32, i8*)*, i64)
 
 declare i8* @__kmpc_taskred_modifier_init(i8*, i32, i32, i32, i8*)
 
 declare i8* @__kmpc_taskred_init(i32, i32, i8*)
 
-declare void @__kmpc_task_reduction_modifier_fini(i8*, i32, i32)
+declare void @__kmpc_task_reduction_modifier_fini(%struct.ident_t*, i32, i32)
 
 declare void @__kmpc_copyprivate(%struct.ident_t*, i32, i64, i8*, void (i8*, i8*)*, i32)
 
 declare i8* @__kmpc_threadprivate_cached(%struct.ident_t*, i32, i8*, i64, i8***)
 
 declare void @__kmpc_threadprivate_register(%struct.ident_t*, i8*, i8* (i8*)*, i8* (i8*, i8*)*, void (i8*)*)
 
 declare void @__kmpc_doacross_init(%struct.ident_t*, i32, i32, i8*)
 
 declare void @__kmpc_doacross_wait(%struct.ident_t*, i32, i64*)
 
 declare void @__kmpc_doacross_post(%struct.ident_t*, i32, i64*)
 
 declare void @__kmpc_doacross_fini(%struct.ident_t*, i32)
 
 declare i8* @__kmpc_alloc(i32, i64, i8*)
 
 declare void @__kmpc_free(i32, i8*, i8*)
 
+declare void @__kmpc_push_target_tripcount(i64, i64)
+
+declare i32 @__tgt_target(i64, i8*, i32, i8**, i8**, i64*, i64*)
+
+declare i32 @__tgt_target_nowait(i64, i8*, i32, i8**, i8**, i64*, i64*)
+
+declare i32 @__tgt_target_teams(i64, i8*, i32, i8**, i8**, i64*, i64*, i32, i32)
+
+declare void @__tgt_register_requires(i64)
+
+declare void @__tgt_target_data_begin(i64, i32, i8**, i8**, i64*, i64*)
+
+declare void @__tgt_target_data_begin_nowait(i64, i32, i8**, i8**, i64*, i64*)
+
+declare void @__tgt_target_data_end(i64, i32, i8**, i8**, i64*, i64*)
+
+declare void @__tgt_target_data_end_nowait(i64, i32, i8**, i8**, i64*, i64*)
+
+declare void @__tgt_target_data_update(i64, i32, i8**, i8**, i64*, i64*)
+
+declare void @__tgt_target_data_update_nowait(i64, i32, i8**, i8**, i64*, i64*)
+
+declare i64 @__tgt_mapper_num_components(i8*)
+
+declare void @__tgt_push_mapper_component(i8*, i8*, i8*, i64, i64)
+
 declare i8* @__kmpc_task_allow_completion_event(%struct.ident_t*, i32, i8*)
 
 declare i8* @__kmpc_task_reduction_get_th_data(i32, i8*, i8*)
 
 declare i8* @__kmpc_task_reduction_init(i32, i32, i8*)
 
 declare i8* @__kmpc_task_reduction_modifier_init(i8*, i32, i32, i32, i8*)
 

 
 ; CHECK: Function Attrs: nounwind
 ; CHECK-NEXT: declare void @__kmpc_omp_wait_deps(%struct.ident_t*, i32, i32, i8*, i32, i8*)
 
 ; CHECK: Function Attrs: nounwind
 ; CHECK-NEXT: declare i32 @__kmpc_cancellationpoint(%struct.ident_t*, i32, i32)
 
 ; CHECK: Function Attrs: nounwind
-; CHECK-NEXT: declare void @__kmpc_push_num_teams(%struct.ident_t*, i32, i32, i32)
+; CHECK-NEXT: declare i32 @__kmpc_push_num_teams(%struct.ident_t*, i32, i32, i32)
 
 ; CHECK: Function Attrs: nounwind
 ; CHECK-NEXT: declare void @__kmpc_fork_teams(%struct.ident_t*, i32, void (i32*, i32*, ...)*, ...)
 
 ; CHECK: Function Attrs: nounwind
 ; CHECK-NEXT: declare void @__kmpc_taskloop(%struct.ident_t*, i32, i8*, i32, i64*, i64*, i64, i32, i32, i64, i8*)
 
 ; CHECK: Function Attrs: nounwind
-; CHECK-NEXT: declare i8* @__kmpc_omp_target_task_alloc(%struct.ident_t*, i32, i32, i64, i64, i8*, i64)
+; CHECK-NEXT: declare i8* @__kmpc_omp_target_task_alloc(%struct.ident_t*, i32, i32, i64, i64, i32 (i32, i8*)*, i64)
 
 ; CHECK: Function Attrs: nounwind
 ; CHECK-NEXT: declare i8* @__kmpc_taskred_modifier_init(i8*, i32, i32, i32, i8*)
 
 ; CHECK: Function Attrs: nounwind
 ; CHECK-NEXT: declare i8* @__kmpc_taskred_init(i32, i32, i8*)
 
 ; CHECK: Function Attrs: nounwind
-; CHECK-NEXT: declare void @__kmpc_task_reduction_modifier_fini(i8*, i32, i32)
+; CHECK-NEXT: declare void @__kmpc_task_reduction_modifier_fini(%struct.ident_t*, i32, i32)
 
 ; CHECK: Function Attrs: nounwind
 ; CHECK-NEXT: declare void @__kmpc_copyprivate(%struct.ident_t*, i32, i64, i8*, void (i8*, i8*)*, i32)
 
 ; CHECK: Function Attrs: nounwind
 ; CHECK-NEXT: declare i8* @__kmpc_threadprivate_cached(%struct.ident_t*, i32, i8*, i64, i8***)
 
 ; CHECK: Function Attrs: nounwind

 
 ; CHECK: Function Attrs: nounwind
 ; CHECK-NEXT: declare i8* @__kmpc_alloc(i32, i64, i8*)
 
 ; CHECK: Function Attrs: nounwind
 ; CHECK-NEXT: declare void @__kmpc_free(i32, i8*, i8*)
 
 ; CHECK: Function Attrs: nounwind
+; CHECK-NEXT: declare void @__kmpc_push_target_tripcount(i64, i64)
+
+; CHECK: Function Attrs: nounwind
+; CHECK-NEXT: declare i32 @__tgt_target(i64, i8*, i32, i8**, i8**, i64*, i64*)
+
+; CHECK: Function Attrs: nounwind
+; CHECK-NEXT: declare i32 @__tgt_target_nowait(i64, i8*, i32, i8**, i8**, i64*, i64*)
+
+; CHECK: Function Attrs: nounwind
+; CHECK-NEXT: declare i32 @__tgt_target_teams(i64, i8*, i32, i8**, i8**, i64*, i64*, i32, i32)
+
+; CHECK: Function Attrs: nounwind
+; CHECK-NEXT: declare void @__tgt_register_requires(i64)
+
+; CHECK: Function Attrs: nounwind
+; CHECK-NEXT: declare void @__tgt_target_data_begin(i64, i32, i8**, i8**, i64*, i64*)
+
+; CHECK: Function Attrs: nounwind
+; CHECK-NEXT: declare void @__tgt_target_data_begin_nowait(i64, i32, i8**, i8**, i64*, i64*)
+
+; CHECK: Function Attrs: nounwind
+; CHECK-NEXT: declare void @__tgt_target_data_end(i64, i32, i8**, i8**, i64*, i64*)
+
+; CHECK: Function Attrs: nounwind
+; CHECK-NEXT: declare void @__tgt_target_data_end_nowait(i64, i32, i8**, i8**, i64*, i64*)
+
+; CHECK: Function Attrs: nounwind
+; CHECK-NEXT: declare void @__tgt_target_data_update(i64, i32, i8**, i8**, i64*, i64*)
+
+; CHECK: Function Attrs: nounwind
+; CHECK-NEXT: declare void @__tgt_target_data_update_nowait(i64, i32, i8**, i8**, i64*, i64*)
+
+; CHECK: Function Attrs: nounwind
+; CHECK-NEXT: declare i64 @__tgt_mapper_num_components(i8*)
+
+; CHECK: Function Attrs: nounwind
+; CHECK-NEXT: declare void @__tgt_push_mapper_component(i8*, i8*, i8*, i64, i64)
+
+; CHECK: Function Attrs: nounwind
 ; CHECK-NEXT: declare i8* @__kmpc_task_allow_completion_event(%struct.ident_t*, i32, i8*)
 
 ; CHECK: Function Attrs: nounwind
 ; CHECK-NEXT: declare i8* @__kmpc_task_reduction_get_th_data(i32, i8*, i8*)
 
 ; CHECK: Function Attrs: nounwind
 ; CHECK-NEXT: declare i8* @__kmpc_task_reduction_init(i32, i32, i8*)
 
 ; CHECK: Function Attrs: nounwind
 ; CHECK-NEXT: declare i8* @__kmpc_task_reduction_modifier_init(i8*, i32, i32, i32, i8*)
 
+; CHECK: Function Attrs: nounwind
+; CHECK-NEXT: declare void @__kmpc_proxy_task_completed_ooo(i8*)
+
 ; OPTIMISTIC: ; Function Attrs: inaccessiblememonly nofree nosync nounwind writeonly
 ; OPTIMISTIC-NEXT: declare dso_local void @omp_set_num_threads(i32)
 
 ; OPTIMISTIC: ; Function Attrs: inaccessiblememonly nofree nosync nounwind writeonly
 ; OPTIMISTIC-NEXT: declare dso_local void @omp_set_dynamic(i32)
 
 ; OPTIMISTIC: ; Function Attrs: inaccessiblememonly nofree nosync nounwind writeonly
 ; OPTIMISTIC-NEXT: declare dso_local void @omp_set_nested(i32)