shylie/llvm-project
1
0
Fork 0
Code Issues Pull Requests Actions 6 Packages Projects Releases Wiki Activity
llvm-project/clang/test/OpenMP/nvptx_teams_codegen.cpp
Joseph Huber 4376fbd793
[OpenMP] Move OpenMP implicit argument to the end and reformat (#185989) 
Summary:
We use this `dyn_ptr` argument in Clang/OpenMP to handle the
`KernelLaunchEnvironment`. This is a per-kernel argument used to share
some information. Currenetly, it's prepended to the argument list and we
generate storage for it in the runtime.

This is bad for a few reasons:
1. It changes the ABI by shifting user arguments
2. It cannot be trivially be left uninitialized if unused
3. The runtime must allocate its own memory for it

This PR changes it to be appended instead. Additionally, space for this
is always emitted. This means the OMPIRBuilder itself will provide the
storage, we simply need to populate it in the runtime if it is used.
This means that if it's unused we don't always pay the cost and it's
easier for non-OpenMP users to ignore it.

Backward compatibility is maintained by auto-upgrading the kernel
arguments. In `libomptarget` we completely allocate a new buffer to
store this in the new format. The plugins still need to respect the old
ABI of the called device object, so we simply rotate it if it's the old
version.
2026-03-12 18:08:22 -05:00

422 lines
25 KiB
C++
Raw Blame History

// NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py UTC_ARGS: --function-signature --include-generated-funcs --replace-value-regex "__omp_offloading_[0-9a-z]+_[0-9a-z]+" "reduction_size[.].+[.]" "pl_cond[.].+[.|,]" --prefix-filecheck-ir-name _
// Test target codegen - host bc file has to be created first.
// RUN: %clang_cc1 -DCK1 -verify -fopenmp -x c++ -triple powerpc64le-unknown-unknown -fopenmp-targets=nvptx64-nvidia-cuda -emit-llvm-bc %s -o %t-ppc-host.bc
// RUN: %clang_cc1 -DCK1 -verify -fopenmp -x c++ -triple nvptx64-unknown-unknown -fopenmp-targets=nvptx64-nvidia-cuda -emit-llvm %s -fopenmp-is-target-device -fopenmp-host-ir-file-path %t-ppc-host.bc -o - | FileCheck %s --check-prefix=CHECK1
// RUN: %clang_cc1 -DCK1 -verify -fopenmp -x c++ -triple i386-unknown-unknown -fopenmp-targets=nvptx-nvidia-cuda -emit-llvm-bc %s -o %t-x86-host.bc
// RUN: %clang_cc1 -DCK1 -verify -fopenmp -x c++ -triple nvptx-unknown-unknown -fopenmp-targets=nvptx-nvidia-cuda -emit-llvm %s -fopenmp-is-target-device -fopenmp-host-ir-file-path %t-x86-host.bc -o - | FileCheck %s --check-prefix=CHECK2
// expected-no-diagnostics
#ifndef HEADER
#define HEADER
#ifdef CK1
template <typename T>
int tmain(T argc) {
#pragma omp target
#pragma omp teams
argc = 0;
return 0;
}
int main (int argc, char **argv) {
#pragma omp target
#pragma omp teams
{
argc = 0;
}
return tmain(argv);
}
// only nvptx side: do not outline teams region and do not call fork_teams
// target region in template
#endif // CK1
// Test target codegen - host bc file has to be created first.
// RUN: %clang_cc1 -DCK2 -verify -fopenmp -x c++ -triple powerpc64le-unknown-unknown -fopenmp-targets=nvptx64-nvidia-cuda -emit-llvm-bc %s -o %t-ppc-host.bc
// RUN: %clang_cc1 -DCK2 -verify -fopenmp -x c++ -triple nvptx64-unknown-unknown -fopenmp-targets=nvptx64-nvidia-cuda -emit-llvm %s -fopenmp-is-target-device -fopenmp-host-ir-file-path %t-ppc-host.bc -o - | FileCheck %s --check-prefix=CHECK3
// RUN: %clang_cc1 -DCK2 -verify -fopenmp -x c++ -triple i386-unknown-unknown -fopenmp-targets=nvptx-nvidia-cuda -emit-llvm-bc %s -o %t-x86-host.bc
// RUN: %clang_cc1 -DCK2 -verify -fopenmp -x c++ -triple nvptx-unknown-unknown -fopenmp-targets=nvptx-nvidia-cuda -emit-llvm %s -fopenmp-is-target-device -fopenmp-host-ir-file-path %t-x86-host.bc -o - | FileCheck %s --check-prefix=CHECK4
// expected-no-diagnostics
#ifdef CK2
template <typename T>
int tmain(T argc) {
int a = 10;
int b = 5;
#pragma omp target
#pragma omp teams num_teams(a) thread_limit(b)
{
argc = 0;
}
return 0;
}
int main (int argc, char **argv) {
int a = 20;
int b = 5;
#pragma omp target
#pragma omp teams num_teams(a) thread_limit(b)
{
argc = 0;
}
return tmain(argv);
}
#endif // CK2
#endif
// CHECK1-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}_main_l23
// CHECK1-SAME: (i64 noundef [[ARGC:%.*]], ptr noalias noundef [[DYN_PTR:%.*]]) #[[ATTR0:[0-9]+]] {
// CHECK1-NEXT: entry:
// CHECK1-NEXT: [[ARGC_ADDR:%.*]] = alloca i64, align 8
// CHECK1-NEXT: [[DYN_PTR_ADDR:%.*]] = alloca ptr, align 8
// CHECK1-NEXT: [[DOTZERO_ADDR:%.*]] = alloca i32, align 4
// CHECK1-NEXT: [[DOTTHREADID_TEMP_:%.*]] = alloca i32, align 4
// CHECK1-NEXT: store i64 [[ARGC]], ptr [[ARGC_ADDR]], align 8
// CHECK1-NEXT: store ptr [[DYN_PTR]], ptr [[DYN_PTR_ADDR]], align 8
// CHECK1-NEXT: [[TMP0:%.*]] = call i32 @__kmpc_target_init(ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}_main_l23_kernel_environment, ptr [[DYN_PTR]])
// CHECK1-NEXT: [[EXEC_USER_CODE:%.*]] = icmp eq i32 [[TMP0]], -1
// CHECK1-NEXT: br i1 [[EXEC_USER_CODE]], label [[USER_CODE_ENTRY:%.*]], label [[WORKER_EXIT:%.*]]
// CHECK1: user_code.entry:
// CHECK1-NEXT: [[TMP1:%.*]] = load i32, ptr [[ARGC_ADDR]], align 4
// CHECK1-NEXT: [[ARGC1:%.*]] = call align 8 ptr @__kmpc_alloc_shared(i64 4)
// CHECK1-NEXT: store i32 [[TMP1]], ptr [[ARGC1]], align 4
// CHECK1-NEXT: [[TMP2:%.*]] = call i32 @__kmpc_global_thread_num(ptr @[[GLOB1:[0-9]+]])
// CHECK1-NEXT: store i32 0, ptr [[DOTZERO_ADDR]], align 4
// CHECK1-NEXT: store i32 [[TMP2]], ptr [[DOTTHREADID_TEMP_]], align 4
// CHECK1-NEXT: call void @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}_main_l23_omp_outlined(ptr [[DOTTHREADID_TEMP_]], ptr [[DOTZERO_ADDR]], ptr [[ARGC1]]) #[[ATTR3:[0-9]+]]
// CHECK1-NEXT: call void @__kmpc_free_shared(ptr [[ARGC1]], i64 4)
// CHECK1-NEXT: call void @__kmpc_target_deinit()
// CHECK1-NEXT: ret void
// CHECK1: worker.exit:
// CHECK1-NEXT: ret void
//
//
// CHECK1-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}_main_l23_omp_outlined
// CHECK1-SAME: (ptr noalias noundef [[DOTGLOBAL_TID_:%.*]], ptr noalias noundef [[DOTBOUND_TID_:%.*]], ptr noundef nonnull align 4 dereferenceable(4) [[ARGC:%.*]]) #[[ATTR2:[0-9]+]] {
// CHECK1-NEXT: entry:
// CHECK1-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca ptr, align 8
// CHECK1-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca ptr, align 8
// CHECK1-NEXT: [[ARGC_ADDR:%.*]] = alloca ptr, align 8
// CHECK1-NEXT: store ptr [[DOTGLOBAL_TID_]], ptr [[DOTGLOBAL_TID__ADDR]], align 8
// CHECK1-NEXT: store ptr [[DOTBOUND_TID_]], ptr [[DOTBOUND_TID__ADDR]], align 8
// CHECK1-NEXT: store ptr [[ARGC]], ptr [[ARGC_ADDR]], align 8
// CHECK1-NEXT: [[TMP0:%.*]] = load ptr, ptr [[ARGC_ADDR]], align 8, !nonnull [[META5:![0-9]+]], !align [[META6:![0-9]+]]
// CHECK1-NEXT: store i32 0, ptr [[TMP0]], align 4
// CHECK1-NEXT: ret void
//
//
// CHECK1-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z5tmainIPPcEiT__l15
// CHECK1-SAME: (ptr noundef [[ARGC:%.*]], ptr noalias noundef [[DYN_PTR:%.*]]) #[[ATTR0]] {
// CHECK1-NEXT: entry:
// CHECK1-NEXT: [[ARGC_ADDR:%.*]] = alloca ptr, align 8
// CHECK1-NEXT: [[DYN_PTR_ADDR:%.*]] = alloca ptr, align 8
// CHECK1-NEXT: [[DOTZERO_ADDR:%.*]] = alloca i32, align 4
// CHECK1-NEXT: [[DOTTHREADID_TEMP_:%.*]] = alloca i32, align 4
// CHECK1-NEXT: store ptr [[ARGC]], ptr [[ARGC_ADDR]], align 8
// CHECK1-NEXT: store ptr [[DYN_PTR]], ptr [[DYN_PTR_ADDR]], align 8
// CHECK1-NEXT: [[TMP0:%.*]] = call i32 @__kmpc_target_init(ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z5tmainIPPcEiT__l15_kernel_environment, ptr [[DYN_PTR]])
// CHECK1-NEXT: [[EXEC_USER_CODE:%.*]] = icmp eq i32 [[TMP0]], -1
// CHECK1-NEXT: br i1 [[EXEC_USER_CODE]], label [[USER_CODE_ENTRY:%.*]], label [[WORKER_EXIT:%.*]]
// CHECK1: user_code.entry:
// CHECK1-NEXT: [[TMP1:%.*]] = load ptr, ptr [[ARGC_ADDR]], align 8
// CHECK1-NEXT: [[ARGC1:%.*]] = call align 8 ptr @__kmpc_alloc_shared(i64 8)
// CHECK1-NEXT: store ptr [[TMP1]], ptr [[ARGC1]], align 8
// CHECK1-NEXT: [[TMP2:%.*]] = call i32 @__kmpc_global_thread_num(ptr @[[GLOB1]])
// CHECK1-NEXT: store i32 0, ptr [[DOTZERO_ADDR]], align 4
// CHECK1-NEXT: store i32 [[TMP2]], ptr [[DOTTHREADID_TEMP_]], align 4
// CHECK1-NEXT: call void @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z5tmainIPPcEiT__l15_omp_outlined(ptr [[DOTTHREADID_TEMP_]], ptr [[DOTZERO_ADDR]], ptr [[ARGC1]]) #[[ATTR3]]
// CHECK1-NEXT: call void @__kmpc_free_shared(ptr [[ARGC1]], i64 8)
// CHECK1-NEXT: call void @__kmpc_target_deinit()
// CHECK1-NEXT: ret void
// CHECK1: worker.exit:
// CHECK1-NEXT: ret void
//
//
// CHECK1-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z5tmainIPPcEiT__l15_omp_outlined
// CHECK1-SAME: (ptr noalias noundef [[DOTGLOBAL_TID_:%.*]], ptr noalias noundef [[DOTBOUND_TID_:%.*]], ptr noundef nonnull align 8 dereferenceable(8) [[ARGC:%.*]]) #[[ATTR2]] {
// CHECK1-NEXT: entry:
// CHECK1-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca ptr, align 8
// CHECK1-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca ptr, align 8
// CHECK1-NEXT: [[ARGC_ADDR:%.*]] = alloca ptr, align 8
// CHECK1-NEXT: store ptr [[DOTGLOBAL_TID_]], ptr [[DOTGLOBAL_TID__ADDR]], align 8
// CHECK1-NEXT: store ptr [[DOTBOUND_TID_]], ptr [[DOTBOUND_TID__ADDR]], align 8
// CHECK1-NEXT: store ptr [[ARGC]], ptr [[ARGC_ADDR]], align 8
// CHECK1-NEXT: [[TMP0:%.*]] = load ptr, ptr [[ARGC_ADDR]], align 8, !nonnull [[META5]], !align [[META7:![0-9]+]]
// CHECK1-NEXT: store ptr null, ptr [[TMP0]], align 8
// CHECK1-NEXT: ret void
//
//
// CHECK2-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}_main_l23
// CHECK2-SAME: (i32 noundef [[ARGC:%.*]], ptr noalias noundef [[DYN_PTR:%.*]]) #[[ATTR0:[0-9]+]] {
// CHECK2-NEXT: entry:
// CHECK2-NEXT: [[ARGC_ADDR:%.*]] = alloca i32, align 4
// CHECK2-NEXT: [[DYN_PTR_ADDR:%.*]] = alloca ptr, align 4
// CHECK2-NEXT: [[DOTZERO_ADDR:%.*]] = alloca i32, align 4
// CHECK2-NEXT: [[DOTTHREADID_TEMP_:%.*]] = alloca i32, align 4
// CHECK2-NEXT: store i32 [[ARGC]], ptr [[ARGC_ADDR]], align 4
// CHECK2-NEXT: store ptr [[DYN_PTR]], ptr [[DYN_PTR_ADDR]], align 4
// CHECK2-NEXT: [[TMP0:%.*]] = call i32 @__kmpc_target_init(ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}_main_l23_kernel_environment, ptr [[DYN_PTR]])
// CHECK2-NEXT: [[EXEC_USER_CODE:%.*]] = icmp eq i32 [[TMP0]], -1
// CHECK2-NEXT: br i1 [[EXEC_USER_CODE]], label [[USER_CODE_ENTRY:%.*]], label [[WORKER_EXIT:%.*]]
// CHECK2: user_code.entry:
// CHECK2-NEXT: [[TMP1:%.*]] = load i32, ptr [[ARGC_ADDR]], align 4
// CHECK2-NEXT: [[ARGC1:%.*]] = call align 8 ptr @__kmpc_alloc_shared(i32 4)
// CHECK2-NEXT: store i32 [[TMP1]], ptr [[ARGC1]], align 4
// CHECK2-NEXT: [[TMP2:%.*]] = call i32 @__kmpc_global_thread_num(ptr @[[GLOB1:[0-9]+]])
// CHECK2-NEXT: store i32 0, ptr [[DOTZERO_ADDR]], align 4
// CHECK2-NEXT: store i32 [[TMP2]], ptr [[DOTTHREADID_TEMP_]], align 4
// CHECK2-NEXT: call void @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}_main_l23_omp_outlined(ptr [[DOTTHREADID_TEMP_]], ptr [[DOTZERO_ADDR]], ptr [[ARGC1]]) #[[ATTR3:[0-9]+]]
// CHECK2-NEXT: call void @__kmpc_free_shared(ptr [[ARGC1]], i32 4)
// CHECK2-NEXT: call void @__kmpc_target_deinit()
// CHECK2-NEXT: ret void
// CHECK2: worker.exit:
// CHECK2-NEXT: ret void
//
//
// CHECK2-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}_main_l23_omp_outlined
// CHECK2-SAME: (ptr noalias noundef [[DOTGLOBAL_TID_:%.*]], ptr noalias noundef [[DOTBOUND_TID_:%.*]], ptr noundef nonnull align 4 dereferenceable(4) [[ARGC:%.*]]) #[[ATTR2:[0-9]+]] {
// CHECK2-NEXT: entry:
// CHECK2-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca ptr, align 4
// CHECK2-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca ptr, align 4
// CHECK2-NEXT: [[ARGC_ADDR:%.*]] = alloca ptr, align 4
// CHECK2-NEXT: store ptr [[DOTGLOBAL_TID_]], ptr [[DOTGLOBAL_TID__ADDR]], align 4
// CHECK2-NEXT: store ptr [[DOTBOUND_TID_]], ptr [[DOTBOUND_TID__ADDR]], align 4
// CHECK2-NEXT: store ptr [[ARGC]], ptr [[ARGC_ADDR]], align 4
// CHECK2-NEXT: [[TMP0:%.*]] = load ptr, ptr [[ARGC_ADDR]], align 4, !nonnull [[META5:![0-9]+]], !align [[META6:![0-9]+]]
// CHECK2-NEXT: store i32 0, ptr [[TMP0]], align 4
// CHECK2-NEXT: ret void
//
//
// CHECK2-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z5tmainIPPcEiT__l15
// CHECK2-SAME: (ptr noundef [[ARGC:%.*]], ptr noalias noundef [[DYN_PTR:%.*]]) #[[ATTR0]] {
// CHECK2-NEXT: entry:
// CHECK2-NEXT: [[ARGC_ADDR:%.*]] = alloca ptr, align 4
// CHECK2-NEXT: [[DYN_PTR_ADDR:%.*]] = alloca ptr, align 4
// CHECK2-NEXT: [[DOTZERO_ADDR:%.*]] = alloca i32, align 4
// CHECK2-NEXT: [[DOTTHREADID_TEMP_:%.*]] = alloca i32, align 4
// CHECK2-NEXT: store ptr [[ARGC]], ptr [[ARGC_ADDR]], align 4
// CHECK2-NEXT: store ptr [[DYN_PTR]], ptr [[DYN_PTR_ADDR]], align 4
// CHECK2-NEXT: [[TMP0:%.*]] = call i32 @__kmpc_target_init(ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z5tmainIPPcEiT__l15_kernel_environment, ptr [[DYN_PTR]])
// CHECK2-NEXT: [[EXEC_USER_CODE:%.*]] = icmp eq i32 [[TMP0]], -1
// CHECK2-NEXT: br i1 [[EXEC_USER_CODE]], label [[USER_CODE_ENTRY:%.*]], label [[WORKER_EXIT:%.*]]
// CHECK2: user_code.entry:
// CHECK2-NEXT: [[TMP1:%.*]] = load ptr, ptr [[ARGC_ADDR]], align 4
// CHECK2-NEXT: [[ARGC1:%.*]] = call align 8 ptr @__kmpc_alloc_shared(i32 4)
// CHECK2-NEXT: store ptr [[TMP1]], ptr [[ARGC1]], align 4
// CHECK2-NEXT: [[TMP2:%.*]] = call i32 @__kmpc_global_thread_num(ptr @[[GLOB1]])
// CHECK2-NEXT: store i32 0, ptr [[DOTZERO_ADDR]], align 4
// CHECK2-NEXT: store i32 [[TMP2]], ptr [[DOTTHREADID_TEMP_]], align 4
// CHECK2-NEXT: call void @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z5tmainIPPcEiT__l15_omp_outlined(ptr [[DOTTHREADID_TEMP_]], ptr [[DOTZERO_ADDR]], ptr [[ARGC1]]) #[[ATTR3]]
// CHECK2-NEXT: call void @__kmpc_free_shared(ptr [[ARGC1]], i32 4)
// CHECK2-NEXT: call void @__kmpc_target_deinit()
// CHECK2-NEXT: ret void
// CHECK2: worker.exit:
// CHECK2-NEXT: ret void
//
//
// CHECK2-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z5tmainIPPcEiT__l15_omp_outlined
// CHECK2-SAME: (ptr noalias noundef [[DOTGLOBAL_TID_:%.*]], ptr noalias noundef [[DOTBOUND_TID_:%.*]], ptr noundef nonnull align 4 dereferenceable(4) [[ARGC:%.*]]) #[[ATTR2]] {
// CHECK2-NEXT: entry:
// CHECK2-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca ptr, align 4
// CHECK2-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca ptr, align 4
// CHECK2-NEXT: [[ARGC_ADDR:%.*]] = alloca ptr, align 4
// CHECK2-NEXT: store ptr [[DOTGLOBAL_TID_]], ptr [[DOTGLOBAL_TID__ADDR]], align 4
// CHECK2-NEXT: store ptr [[DOTBOUND_TID_]], ptr [[DOTBOUND_TID__ADDR]], align 4
// CHECK2-NEXT: store ptr [[ARGC]], ptr [[ARGC_ADDR]], align 4
// CHECK2-NEXT: [[TMP0:%.*]] = load ptr, ptr [[ARGC_ADDR]], align 4, !nonnull [[META5]], !align [[META6]]
// CHECK2-NEXT: store ptr null, ptr [[TMP0]], align 4
// CHECK2-NEXT: ret void
//
//
// CHECK3-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}_main_l64
// CHECK3-SAME: (i64 noundef [[A:%.*]], i64 noundef [[B:%.*]], i64 noundef [[ARGC:%.*]], ptr noalias noundef [[DYN_PTR:%.*]]) #[[ATTR0:[0-9]+]] {
// CHECK3-NEXT: entry:
// CHECK3-NEXT: [[A_ADDR:%.*]] = alloca i64, align 8
// CHECK3-NEXT: [[B_ADDR:%.*]] = alloca i64, align 8
// CHECK3-NEXT: [[ARGC_ADDR:%.*]] = alloca i64, align 8
// CHECK3-NEXT: [[DYN_PTR_ADDR:%.*]] = alloca ptr, align 8
// CHECK3-NEXT: [[DOTZERO_ADDR:%.*]] = alloca i32, align 4
// CHECK3-NEXT: [[DOTTHREADID_TEMP_:%.*]] = alloca i32, align 4
// CHECK3-NEXT: store i64 [[A]], ptr [[A_ADDR]], align 8
// CHECK3-NEXT: store i64 [[B]], ptr [[B_ADDR]], align 8
// CHECK3-NEXT: store i64 [[ARGC]], ptr [[ARGC_ADDR]], align 8
// CHECK3-NEXT: store ptr [[DYN_PTR]], ptr [[DYN_PTR_ADDR]], align 8
// CHECK3-NEXT: [[TMP0:%.*]] = call i32 @__kmpc_target_init(ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}_main_l64_kernel_environment, ptr [[DYN_PTR]])
// CHECK3-NEXT: [[EXEC_USER_CODE:%.*]] = icmp eq i32 [[TMP0]], -1
// CHECK3-NEXT: br i1 [[EXEC_USER_CODE]], label [[USER_CODE_ENTRY:%.*]], label [[WORKER_EXIT:%.*]]
// CHECK3: user_code.entry:
// CHECK3-NEXT: [[TMP1:%.*]] = load i32, ptr [[ARGC_ADDR]], align 4
// CHECK3-NEXT: [[ARGC1:%.*]] = call align 8 ptr @__kmpc_alloc_shared(i64 4)
// CHECK3-NEXT: store i32 [[TMP1]], ptr [[ARGC1]], align 4
// CHECK3-NEXT: [[TMP2:%.*]] = call i32 @__kmpc_global_thread_num(ptr @[[GLOB1:[0-9]+]])
// CHECK3-NEXT: store i32 0, ptr [[DOTZERO_ADDR]], align 4
// CHECK3-NEXT: store i32 [[TMP2]], ptr [[DOTTHREADID_TEMP_]], align 4
// CHECK3-NEXT: call void @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}_main_l64_omp_outlined(ptr [[DOTTHREADID_TEMP_]], ptr [[DOTZERO_ADDR]], ptr [[ARGC1]]) #[[ATTR3:[0-9]+]]
// CHECK3-NEXT: call void @__kmpc_free_shared(ptr [[ARGC1]], i64 4)
// CHECK3-NEXT: call void @__kmpc_target_deinit()
// CHECK3-NEXT: ret void
// CHECK3: worker.exit:
// CHECK3-NEXT: ret void
//
//
// CHECK3-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}_main_l64_omp_outlined
// CHECK3-SAME: (ptr noalias noundef [[DOTGLOBAL_TID_:%.*]], ptr noalias noundef [[DOTBOUND_TID_:%.*]], ptr noundef nonnull align 4 dereferenceable(4) [[ARGC:%.*]]) #[[ATTR2:[0-9]+]] {
// CHECK3-NEXT: entry:
// CHECK3-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca ptr, align 8
// CHECK3-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca ptr, align 8
// CHECK3-NEXT: [[ARGC_ADDR:%.*]] = alloca ptr, align 8
// CHECK3-NEXT: store ptr [[DOTGLOBAL_TID_]], ptr [[DOTGLOBAL_TID__ADDR]], align 8
// CHECK3-NEXT: store ptr [[DOTBOUND_TID_]], ptr [[DOTBOUND_TID__ADDR]], align 8
// CHECK3-NEXT: store ptr [[ARGC]], ptr [[ARGC_ADDR]], align 8
// CHECK3-NEXT: [[TMP0:%.*]] = load ptr, ptr [[ARGC_ADDR]], align 8, !nonnull [[META5:![0-9]+]], !align [[META6:![0-9]+]]
// CHECK3-NEXT: store i32 0, ptr [[TMP0]], align 4
// CHECK3-NEXT: ret void
//
//
// CHECK3-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z5tmainIPPcEiT__l53
// CHECK3-SAME: (i64 noundef [[A:%.*]], i64 noundef [[B:%.*]], ptr noundef [[ARGC:%.*]], ptr noalias noundef [[DYN_PTR:%.*]]) #[[ATTR0]] {
// CHECK3-NEXT: entry:
// CHECK3-NEXT: [[A_ADDR:%.*]] = alloca i64, align 8
// CHECK3-NEXT: [[B_ADDR:%.*]] = alloca i64, align 8
// CHECK3-NEXT: [[ARGC_ADDR:%.*]] = alloca ptr, align 8
// CHECK3-NEXT: [[DYN_PTR_ADDR:%.*]] = alloca ptr, align 8
// CHECK3-NEXT: [[DOTZERO_ADDR:%.*]] = alloca i32, align 4
// CHECK3-NEXT: [[DOTTHREADID_TEMP_:%.*]] = alloca i32, align 4
// CHECK3-NEXT: store i64 [[A]], ptr [[A_ADDR]], align 8
// CHECK3-NEXT: store i64 [[B]], ptr [[B_ADDR]], align 8
// CHECK3-NEXT: store ptr [[ARGC]], ptr [[ARGC_ADDR]], align 8
// CHECK3-NEXT: store ptr [[DYN_PTR]], ptr [[DYN_PTR_ADDR]], align 8
// CHECK3-NEXT: [[TMP0:%.*]] = call i32 @__kmpc_target_init(ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z5tmainIPPcEiT__l53_kernel_environment, ptr [[DYN_PTR]])
// CHECK3-NEXT: [[EXEC_USER_CODE:%.*]] = icmp eq i32 [[TMP0]], -1
// CHECK3-NEXT: br i1 [[EXEC_USER_CODE]], label [[USER_CODE_ENTRY:%.*]], label [[WORKER_EXIT:%.*]]
// CHECK3: user_code.entry:
// CHECK3-NEXT: [[TMP1:%.*]] = load ptr, ptr [[ARGC_ADDR]], align 8
// CHECK3-NEXT: [[ARGC1:%.*]] = call align 8 ptr @__kmpc_alloc_shared(i64 8)
// CHECK3-NEXT: store ptr [[TMP1]], ptr [[ARGC1]], align 8
// CHECK3-NEXT: [[TMP2:%.*]] = call i32 @__kmpc_global_thread_num(ptr @[[GLOB1]])
// CHECK3-NEXT: store i32 0, ptr [[DOTZERO_ADDR]], align 4
// CHECK3-NEXT: store i32 [[TMP2]], ptr [[DOTTHREADID_TEMP_]], align 4
// CHECK3-NEXT: call void @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z5tmainIPPcEiT__l53_omp_outlined(ptr [[DOTTHREADID_TEMP_]], ptr [[DOTZERO_ADDR]], ptr [[ARGC1]]) #[[ATTR3]]
// CHECK3-NEXT: call void @__kmpc_free_shared(ptr [[ARGC1]], i64 8)
// CHECK3-NEXT: call void @__kmpc_target_deinit()
// CHECK3-NEXT: ret void
// CHECK3: worker.exit:
// CHECK3-NEXT: ret void
//
//
// CHECK3-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z5tmainIPPcEiT__l53_omp_outlined
// CHECK3-SAME: (ptr noalias noundef [[DOTGLOBAL_TID_:%.*]], ptr noalias noundef [[DOTBOUND_TID_:%.*]], ptr noundef nonnull align 8 dereferenceable(8) [[ARGC:%.*]]) #[[ATTR2]] {
// CHECK3-NEXT: entry:
// CHECK3-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca ptr, align 8
// CHECK3-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca ptr, align 8
// CHECK3-NEXT: [[ARGC_ADDR:%.*]] = alloca ptr, align 8
// CHECK3-NEXT: store ptr [[DOTGLOBAL_TID_]], ptr [[DOTGLOBAL_TID__ADDR]], align 8
// CHECK3-NEXT: store ptr [[DOTBOUND_TID_]], ptr [[DOTBOUND_TID__ADDR]], align 8
// CHECK3-NEXT: store ptr [[ARGC]], ptr [[ARGC_ADDR]], align 8
// CHECK3-NEXT: [[TMP0:%.*]] = load ptr, ptr [[ARGC_ADDR]], align 8, !nonnull [[META5]], !align [[META7:![0-9]+]]
// CHECK3-NEXT: store ptr null, ptr [[TMP0]], align 8
// CHECK3-NEXT: ret void
//
//
// CHECK4-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}_main_l64
// CHECK4-SAME: (i32 noundef [[A:%.*]], i32 noundef [[B:%.*]], i32 noundef [[ARGC:%.*]], ptr noalias noundef [[DYN_PTR:%.*]]) #[[ATTR0:[0-9]+]] {
// CHECK4-NEXT: entry:
// CHECK4-NEXT: [[A_ADDR:%.*]] = alloca i32, align 4
// CHECK4-NEXT: [[B_ADDR:%.*]] = alloca i32, align 4
// CHECK4-NEXT: [[ARGC_ADDR:%.*]] = alloca i32, align 4
// CHECK4-NEXT: [[DYN_PTR_ADDR:%.*]] = alloca ptr, align 4
// CHECK4-NEXT: [[DOTZERO_ADDR:%.*]] = alloca i32, align 4
// CHECK4-NEXT: [[DOTTHREADID_TEMP_:%.*]] = alloca i32, align 4
// CHECK4-NEXT: store i32 [[A]], ptr [[A_ADDR]], align 4
// CHECK4-NEXT: store i32 [[B]], ptr [[B_ADDR]], align 4
// CHECK4-NEXT: store i32 [[ARGC]], ptr [[ARGC_ADDR]], align 4
// CHECK4-NEXT: store ptr [[DYN_PTR]], ptr [[DYN_PTR_ADDR]], align 4
// CHECK4-NEXT: [[TMP0:%.*]] = call i32 @__kmpc_target_init(ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}_main_l64_kernel_environment, ptr [[DYN_PTR]])
// CHECK4-NEXT: [[EXEC_USER_CODE:%.*]] = icmp eq i32 [[TMP0]], -1
// CHECK4-NEXT: br i1 [[EXEC_USER_CODE]], label [[USER_CODE_ENTRY:%.*]], label [[WORKER_EXIT:%.*]]
// CHECK4: user_code.entry:
// CHECK4-NEXT: [[TMP1:%.*]] = load i32, ptr [[ARGC_ADDR]], align 4
// CHECK4-NEXT: [[ARGC1:%.*]] = call align 8 ptr @__kmpc_alloc_shared(i32 4)
// CHECK4-NEXT: store i32 [[TMP1]], ptr [[ARGC1]], align 4
// CHECK4-NEXT: [[TMP2:%.*]] = call i32 @__kmpc_global_thread_num(ptr @[[GLOB1:[0-9]+]])
// CHECK4-NEXT: store i32 0, ptr [[DOTZERO_ADDR]], align 4
// CHECK4-NEXT: store i32 [[TMP2]], ptr [[DOTTHREADID_TEMP_]], align 4
// CHECK4-NEXT: call void @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}_main_l64_omp_outlined(ptr [[DOTTHREADID_TEMP_]], ptr [[DOTZERO_ADDR]], ptr [[ARGC1]]) #[[ATTR3:[0-9]+]]
// CHECK4-NEXT: call void @__kmpc_free_shared(ptr [[ARGC1]], i32 4)
// CHECK4-NEXT: call void @__kmpc_target_deinit()
// CHECK4-NEXT: ret void
// CHECK4: worker.exit:
// CHECK4-NEXT: ret void
//
//
// CHECK4-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}_main_l64_omp_outlined
// CHECK4-SAME: (ptr noalias noundef [[DOTGLOBAL_TID_:%.*]], ptr noalias noundef [[DOTBOUND_TID_:%.*]], ptr noundef nonnull align 4 dereferenceable(4) [[ARGC:%.*]]) #[[ATTR2:[0-9]+]] {
// CHECK4-NEXT: entry:
// CHECK4-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca ptr, align 4
// CHECK4-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca ptr, align 4
// CHECK4-NEXT: [[ARGC_ADDR:%.*]] = alloca ptr, align 4
// CHECK4-NEXT: store ptr [[DOTGLOBAL_TID_]], ptr [[DOTGLOBAL_TID__ADDR]], align 4
// CHECK4-NEXT: store ptr [[DOTBOUND_TID_]], ptr [[DOTBOUND_TID__ADDR]], align 4
// CHECK4-NEXT: store ptr [[ARGC]], ptr [[ARGC_ADDR]], align 4
// CHECK4-NEXT: [[TMP0:%.*]] = load ptr, ptr [[ARGC_ADDR]], align 4, !nonnull [[META5:![0-9]+]], !align [[META6:![0-9]+]]
// CHECK4-NEXT: store i32 0, ptr [[TMP0]], align 4
// CHECK4-NEXT: ret void
//
//
// CHECK4-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z5tmainIPPcEiT__l53
// CHECK4-SAME: (i32 noundef [[A:%.*]], i32 noundef [[B:%.*]], ptr noundef [[ARGC:%.*]], ptr noalias noundef [[DYN_PTR:%.*]]) #[[ATTR0]] {
// CHECK4-NEXT: entry:
// CHECK4-NEXT: [[A_ADDR:%.*]] = alloca i32, align 4
// CHECK4-NEXT: [[B_ADDR:%.*]] = alloca i32, align 4
// CHECK4-NEXT: [[ARGC_ADDR:%.*]] = alloca ptr, align 4
// CHECK4-NEXT: [[DYN_PTR_ADDR:%.*]] = alloca ptr, align 4
// CHECK4-NEXT: [[DOTZERO_ADDR:%.*]] = alloca i32, align 4
// CHECK4-NEXT: [[DOTTHREADID_TEMP_:%.*]] = alloca i32, align 4
// CHECK4-NEXT: store i32 [[A]], ptr [[A_ADDR]], align 4
// CHECK4-NEXT: store i32 [[B]], ptr [[B_ADDR]], align 4
// CHECK4-NEXT: store ptr [[ARGC]], ptr [[ARGC_ADDR]], align 4
// CHECK4-NEXT: store ptr [[DYN_PTR]], ptr [[DYN_PTR_ADDR]], align 4
// CHECK4-NEXT: [[TMP0:%.*]] = call i32 @__kmpc_target_init(ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z5tmainIPPcEiT__l53_kernel_environment, ptr [[DYN_PTR]])
// CHECK4-NEXT: [[EXEC_USER_CODE:%.*]] = icmp eq i32 [[TMP0]], -1
// CHECK4-NEXT: br i1 [[EXEC_USER_CODE]], label [[USER_CODE_ENTRY:%.*]], label [[WORKER_EXIT:%.*]]
// CHECK4: user_code.entry:
// CHECK4-NEXT: [[TMP1:%.*]] = load ptr, ptr [[ARGC_ADDR]], align 4
// CHECK4-NEXT: [[ARGC1:%.*]] = call align 8 ptr @__kmpc_alloc_shared(i32 4)
// CHECK4-NEXT: store ptr [[TMP1]], ptr [[ARGC1]], align 4
// CHECK4-NEXT: [[TMP2:%.*]] = call i32 @__kmpc_global_thread_num(ptr @[[GLOB1]])
// CHECK4-NEXT: store i32 0, ptr [[DOTZERO_ADDR]], align 4
// CHECK4-NEXT: store i32 [[TMP2]], ptr [[DOTTHREADID_TEMP_]], align 4
// CHECK4-NEXT: call void @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z5tmainIPPcEiT__l53_omp_outlined(ptr [[DOTTHREADID_TEMP_]], ptr [[DOTZERO_ADDR]], ptr [[ARGC1]]) #[[ATTR3]]
// CHECK4-NEXT: call void @__kmpc_free_shared(ptr [[ARGC1]], i32 4)
// CHECK4-NEXT: call void @__kmpc_target_deinit()
// CHECK4-NEXT: ret void
// CHECK4: worker.exit:
// CHECK4-NEXT: ret void
//
//
// CHECK4-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z5tmainIPPcEiT__l53_omp_outlined
// CHECK4-SAME: (ptr noalias noundef [[DOTGLOBAL_TID_:%.*]], ptr noalias noundef [[DOTBOUND_TID_:%.*]], ptr noundef nonnull align 4 dereferenceable(4) [[ARGC:%.*]]) #[[ATTR2]] {
// CHECK4-NEXT: entry:
// CHECK4-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca ptr, align 4
// CHECK4-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca ptr, align 4
// CHECK4-NEXT: [[ARGC_ADDR:%.*]] = alloca ptr, align 4
// CHECK4-NEXT: store ptr [[DOTGLOBAL_TID_]], ptr [[DOTGLOBAL_TID__ADDR]], align 4
// CHECK4-NEXT: store ptr [[DOTBOUND_TID_]], ptr [[DOTBOUND_TID__ADDR]], align 4
// CHECK4-NEXT: store ptr [[ARGC]], ptr [[ARGC_ADDR]], align 4
// CHECK4-NEXT: [[TMP0:%.*]] = load ptr, ptr [[ARGC_ADDR]], align 4, !nonnull [[META5]], !align [[META6]]
// CHECK4-NEXT: store ptr null, ptr [[TMP0]], align 4
// CHECK4-NEXT: ret void
//
Reference in New Issue View Git Blame Copy Permalink