shylie/llvm-project
1
0
Fork 0
Code Issues Pull Requests Actions 6 Packages Projects Releases Wiki Activity
llvm-project/clang/test/OpenMP/target_parallel_num_threads_codegen.cpp
David Tenty 196c144d0b [clang][CodeGenCXX] Improve handling of itanium ABI member function alignment requirements 
The itanium ABI for certain platforms requires a minimum alignments for
member function pointers to reserve certain bits for distinguishing
virtual and non-virtual functions.

Our implementation of this however depends on the alignment of the
function involved, which may however not reflect the true alignment of
function pointers on certain targets for which the alignment is
independent of the function (e.g. AIX). Worse, the 2-byte alignment
we use may be less than the ABI minimum for the target, and in the case
we are using explicit sections will result in invalid codegen.

This patch attempts to correct this situation by considering the target
alignment of function pointers as part of making the decision about
whether we need to adjust the function alignment to conform to the ABI.
Targets which do not provide the function ptr alignment information
will return a value of 1 when queried and will conservatively retain
the old alignment.

Differential Revision: https://reviews.llvm.org/D147184
2023-07-06 10:35:26 -04:00

1711 lines
123 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 host codegen.
// RUN: %clang_cc1 -verify -fopenmp -fopenmp-version=45 -x c++ -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-llvm %s -o - | FileCheck %s --check-prefix=CHECK1
// RUN: %clang_cc1 -fopenmp -fopenmp-version=45 -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-pch -o %t %s
// RUN: %clang_cc1 -fopenmp -fopenmp-version=45 -x c++ -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --check-prefix=CHECK1
// RUN: %clang_cc1 -verify -fopenmp -fopenmp-version=45 -x c++ -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-llvm %s -o - | FileCheck %s --check-prefix=CHECK3
// RUN: %clang_cc1 -fopenmp -fopenmp-version=45 -x c++ -std=c++11 -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-pch -o %t %s
// RUN: %clang_cc1 -fopenmp -fopenmp-version=45 -x c++ -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --check-prefix=CHECK3
// RUN: %clang_cc1 -verify -fopenmp-simd -fopenmp-version=45 -x c++ -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-llvm %s -o - | FileCheck %s --implicit-check-not="{{__kmpc|__tgt}}"
// RUN: %clang_cc1 -fopenmp-simd -fopenmp-version=45 -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-pch -o %t %s
// RUN: %clang_cc1 -fopenmp-simd -fopenmp-version=45 -x c++ -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --implicit-check-not="{{__kmpc|__tgt}}"
// RUN: %clang_cc1 -verify -fopenmp-simd -fopenmp-version=45 -x c++ -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-llvm %s -o - | FileCheck %s --implicit-check-not="{{__kmpc|__tgt}}"
// RUN: %clang_cc1 -fopenmp-simd -fopenmp-version=45 -x c++ -std=c++11 -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-pch -o %t %s
// RUN: %clang_cc1 -fopenmp-simd -fopenmp-version=45 -x c++ -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --implicit-check-not="{{__kmpc|__tgt}}"
// Test target codegen - host bc file has to be created first.
// RUN: %clang_cc1 -verify -fopenmp -fopenmp-version=45 -x c++ -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-llvm-bc %s -o %t-ppc-host.bc
// RUN: %clang_cc1 -verify -fopenmp -fopenmp-version=45 -x c++ -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-llvm %s -fopenmp-is-device -fopenmp-host-ir-file-path %t-ppc-host.bc -o - | FileCheck %s --check-prefix=CHECK9
// RUN: %clang_cc1 -fopenmp -fopenmp-version=45 -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-pch -fopenmp-is-device -fopenmp-host-ir-file-path %t-ppc-host.bc -o %t %s
// RUN: %clang_cc1 -fopenmp -fopenmp-version=45 -x c++ -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -std=c++11 -fopenmp-is-device -fopenmp-host-ir-file-path %t-ppc-host.bc -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --check-prefix=CHECK9
// RUN: %clang_cc1 -verify -fopenmp -fopenmp-version=45 -x c++ -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-llvm-bc %s -o %t-x86-host.bc
// RUN: %clang_cc1 -verify -fopenmp -fopenmp-version=45 -x c++ -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-llvm %s -fopenmp-is-device -fopenmp-host-ir-file-path %t-x86-host.bc -o - | FileCheck %s --check-prefix=CHECK11
// RUN: %clang_cc1 -fopenmp -fopenmp-version=45 -x c++ -std=c++11 -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-pch -fopenmp-is-device -fopenmp-host-ir-file-path %t-x86-host.bc -o %t %s
// RUN: %clang_cc1 -fopenmp -fopenmp-version=45 -x c++ -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -std=c++11 -fopenmp-is-device -fopenmp-host-ir-file-path %t-x86-host.bc -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --check-prefix=CHECK11
// RUN: %clang_cc1 -verify -fopenmp-simd -fopenmp-version=45 -x c++ -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-llvm-bc %s -o %t-ppc-host.bc
// RUN: %clang_cc1 -verify -fopenmp-simd -fopenmp-version=45 -x c++ -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-llvm %s -fopenmp-is-device -fopenmp-host-ir-file-path %t-ppc-host.bc -o - | FileCheck %s --implicit-check-not="{{__kmpc|__tgt}}"
// RUN: %clang_cc1 -fopenmp-simd -fopenmp-version=45 -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-pch -fopenmp-is-device -fopenmp-host-ir-file-path %t-ppc-host.bc -o %t %s
// RUN: %clang_cc1 -fopenmp-simd -fopenmp-version=45 -x c++ -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -std=c++11 -fopenmp-is-device -fopenmp-host-ir-file-path %t-ppc-host.bc -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --implicit-check-not="{{__kmpc|__tgt}}"
// RUN: %clang_cc1 -verify -fopenmp-simd -fopenmp-version=45 -x c++ -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-llvm-bc %s -o %t-x86-host.bc
// RUN: %clang_cc1 -verify -fopenmp-simd -fopenmp-version=45 -x c++ -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-llvm %s -fopenmp-is-device -fopenmp-host-ir-file-path %t-x86-host.bc -o - | FileCheck %s --implicit-check-not="{{__kmpc|__tgt}}"
// RUN: %clang_cc1 -fopenmp-simd -fopenmp-version=45 -x c++ -std=c++11 -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-pch -fopenmp-is-device -fopenmp-host-ir-file-path %t-x86-host.bc -o %t %s
// RUN: %clang_cc1 -fopenmp-simd -fopenmp-version=45 -x c++ -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -std=c++11 -fopenmp-is-device -fopenmp-host-ir-file-path %t-x86-host.bc -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --implicit-check-not="{{__kmpc|__tgt}}"
// Test host codegen.
// RUN: %clang_cc1 -verify -fopenmp -x c++ -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-llvm %s -o - | FileCheck %s --check-prefix=CHECK1
// RUN: %clang_cc1 -fopenmp -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-pch -o %t %s
// RUN: %clang_cc1 -fopenmp -x c++ -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --check-prefix=CHECK1
// RUN: %clang_cc1 -verify -fopenmp -x c++ -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-llvm %s -o - | FileCheck %s --check-prefix=CHECK3
// RUN: %clang_cc1 -fopenmp -x c++ -std=c++11 -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-pch -o %t %s
// RUN: %clang_cc1 -fopenmp -x c++ -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --check-prefix=CHECK3
// RUN: %clang_cc1 -verify -fopenmp-simd -x c++ -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-llvm %s -o - | FileCheck %s --implicit-check-not="{{__kmpc|__tgt}}"
// RUN: %clang_cc1 -fopenmp-simd -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-pch -o %t %s
// RUN: %clang_cc1 -fopenmp-simd -x c++ -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --implicit-check-not="{{__kmpc|__tgt}}"
// RUN: %clang_cc1 -verify -fopenmp-simd -x c++ -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-llvm %s -o - | FileCheck %s --implicit-check-not="{{__kmpc|__tgt}}"
// RUN: %clang_cc1 -fopenmp-simd -x c++ -std=c++11 -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-pch -o %t %s
// RUN: %clang_cc1 -fopenmp-simd -x c++ -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --implicit-check-not="{{__kmpc|__tgt}}"
// Test target codegen - host bc file has to be created first.
// RUN: %clang_cc1 -verify -fopenmp -x c++ -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-llvm-bc %s -o %t-ppc-host.bc
// RUN: %clang_cc1 -verify -fopenmp -x c++ -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-llvm %s -fopenmp-is-device -fopenmp-host-ir-file-path %t-ppc-host.bc -o - | FileCheck %s --check-prefix=CHECK9
// RUN: %clang_cc1 -fopenmp -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-pch -fopenmp-is-device -fopenmp-host-ir-file-path %t-ppc-host.bc -o %t %s
// RUN: %clang_cc1 -fopenmp -x c++ -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -std=c++11 -fopenmp-is-device -fopenmp-host-ir-file-path %t-ppc-host.bc -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --check-prefix=CHECK9
// RUN: %clang_cc1 -verify -fopenmp -x c++ -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-llvm-bc %s -o %t-x86-host.bc
// RUN: %clang_cc1 -verify -fopenmp -x c++ -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-llvm %s -fopenmp-is-device -fopenmp-host-ir-file-path %t-x86-host.bc -o - | FileCheck %s --check-prefix=CHECK11
// RUN: %clang_cc1 -fopenmp -x c++ -std=c++11 -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-pch -fopenmp-is-device -fopenmp-host-ir-file-path %t-x86-host.bc -o %t %s
// RUN: %clang_cc1 -fopenmp -x c++ -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -std=c++11 -fopenmp-is-device -fopenmp-host-ir-file-path %t-x86-host.bc -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --check-prefix=CHECK11
// RUN: %clang_cc1 -verify -fopenmp-simd -x c++ -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-llvm-bc %s -o %t-ppc-host.bc
// RUN: %clang_cc1 -verify -fopenmp-simd -x c++ -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-llvm %s -fopenmp-is-device -fopenmp-host-ir-file-path %t-ppc-host.bc -o - | FileCheck %s --implicit-check-not="{{__kmpc|__tgt}}"
// RUN: %clang_cc1 -fopenmp-simd -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-pch -fopenmp-is-device -fopenmp-host-ir-file-path %t-ppc-host.bc -o %t %s
// RUN: %clang_cc1 -fopenmp-simd -x c++ -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -std=c++11 -fopenmp-is-device -fopenmp-host-ir-file-path %t-ppc-host.bc -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --implicit-check-not="{{__kmpc|__tgt}}"
// RUN: %clang_cc1 -verify -fopenmp-simd -x c++ -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-llvm-bc %s -o %t-x86-host.bc
// RUN: %clang_cc1 -verify -fopenmp-simd -x c++ -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-llvm %s -fopenmp-is-device -fopenmp-host-ir-file-path %t-x86-host.bc -o - | FileCheck %s --implicit-check-not="{{__kmpc|__tgt}}"
// RUN: %clang_cc1 -fopenmp-simd -x c++ -std=c++11 -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-pch -fopenmp-is-device -fopenmp-host-ir-file-path %t-x86-host.bc -o %t %s
// RUN: %clang_cc1 -fopenmp-simd -x c++ -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -std=c++11 -fopenmp-is-device -fopenmp-host-ir-file-path %t-x86-host.bc -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --implicit-check-not="{{__kmpc|__tgt}}"
// expected-no-diagnostics
#ifndef HEADER
#define HEADER
// We have 6 target regions
// Check target registration is registered as a Ctor.
template<typename tx>
tx ftemplate(int n) {
tx a = 0;
#pragma omp target parallel num_threads(tx(20))
{
}
short b = 1;
#pragma omp target parallel num_threads(b)
{
a += b;
}
return a;
}
static
int fstatic(int n) {
#pragma omp target parallel num_threads(n)
{
}
#pragma omp target parallel num_threads(32+n)
{
}
return n+1;
}
struct S1 {
double a;
int r1(int n){
int b = 1;
#pragma omp target parallel num_threads(n-b)
{
this->a = (double)b + 1.5;
}
#pragma omp target parallel num_threads(1024)
{
this->a = 2.5;
}
return (int)a;
}
};
int bar(int n){
int a = 0;
S1 S;
a += S.r1(n);
a += fstatic(n);
a += ftemplate<int>(n);
return a;
}
// Check that the offloading functions are emitted and that the parallel function
// is appropriately guarded.
#endif
// CHECK1-LABEL: define {{[^@]+}}@_Z3bari
// CHECK1-SAME: (i32 noundef signext [[N:%.*]]) #[[ATTR0:[0-9]+]] {
// CHECK1-NEXT: entry:
// CHECK1-NEXT: [[N_ADDR:%.*]] = alloca i32, align 4
// CHECK1-NEXT: [[A:%.*]] = alloca i32, align 4
// CHECK1-NEXT: [[S:%.*]] = alloca [[STRUCT_S1:%.*]], align 8
// CHECK1-NEXT: store i32 [[N]], ptr [[N_ADDR]], align 4
// CHECK1-NEXT: store i32 0, ptr [[A]], align 4
// CHECK1-NEXT: [[TMP0:%.*]] = load i32, ptr [[N_ADDR]], align 4
// CHECK1-NEXT: [[CALL:%.*]] = call noundef signext i32 @_ZN2S12r1Ei(ptr noundef nonnull align 8 dereferenceable(8) [[S]], i32 noundef signext [[TMP0]])
// CHECK1-NEXT: [[TMP1:%.*]] = load i32, ptr [[A]], align 4
// CHECK1-NEXT: [[ADD:%.*]] = add nsw i32 [[TMP1]], [[CALL]]
// CHECK1-NEXT: store i32 [[ADD]], ptr [[A]], align 4
// CHECK1-NEXT: [[TMP2:%.*]] = load i32, ptr [[N_ADDR]], align 4
// CHECK1-NEXT: [[CALL1:%.*]] = call noundef signext i32 @_ZL7fstatici(i32 noundef signext [[TMP2]])
// CHECK1-NEXT: [[TMP3:%.*]] = load i32, ptr [[A]], align 4
// CHECK1-NEXT: [[ADD2:%.*]] = add nsw i32 [[TMP3]], [[CALL1]]
// CHECK1-NEXT: store i32 [[ADD2]], ptr [[A]], align 4
// CHECK1-NEXT: [[TMP4:%.*]] = load i32, ptr [[N_ADDR]], align 4
// CHECK1-NEXT: [[CALL3:%.*]] = call noundef signext i32 @_Z9ftemplateIiET_i(i32 noundef signext [[TMP4]])
// CHECK1-NEXT: [[TMP5:%.*]] = load i32, ptr [[A]], align 4
// CHECK1-NEXT: [[ADD4:%.*]] = add nsw i32 [[TMP5]], [[CALL3]]
// CHECK1-NEXT: store i32 [[ADD4]], ptr [[A]], align 4
// CHECK1-NEXT: [[TMP6:%.*]] = load i32, ptr [[A]], align 4
// CHECK1-NEXT: ret i32 [[TMP6]]
//
//
// CHECK1-LABEL: define {{[^@]+}}@_ZN2S12r1Ei
// CHECK1-SAME: (ptr noundef nonnull align 8 dereferenceable(8) [[THIS:%.*]], i32 noundef signext [[N:%.*]]) #[[ATTR0]] comdat {
// CHECK1-NEXT: entry:
// CHECK1-NEXT: [[THIS_ADDR:%.*]] = alloca ptr, align 8
// CHECK1-NEXT: [[N_ADDR:%.*]] = alloca i32, align 4
// CHECK1-NEXT: [[B:%.*]] = alloca i32, align 4
// CHECK1-NEXT: [[DOTCAPTURE_EXPR_:%.*]] = alloca i32, align 4
// CHECK1-NEXT: [[B_CASTED:%.*]] = alloca i64, align 8
// CHECK1-NEXT: [[DOTCAPTURE_EXPR__CASTED:%.*]] = alloca i64, align 8
// CHECK1-NEXT: [[DOTOFFLOAD_BASEPTRS:%.*]] = alloca [3 x ptr], align 8
// CHECK1-NEXT: [[DOTOFFLOAD_PTRS:%.*]] = alloca [3 x ptr], align 8
// CHECK1-NEXT: [[DOTOFFLOAD_MAPPERS:%.*]] = alloca [3 x ptr], align 8
// CHECK1-NEXT: [[KERNEL_ARGS:%.*]] = alloca [[STRUCT___TGT_KERNEL_ARGUMENTS:%.*]], align 8
// CHECK1-NEXT: [[DOTOFFLOAD_BASEPTRS3:%.*]] = alloca [1 x ptr], align 8
// CHECK1-NEXT: [[DOTOFFLOAD_PTRS4:%.*]] = alloca [1 x ptr], align 8
// CHECK1-NEXT: [[DOTOFFLOAD_MAPPERS5:%.*]] = alloca [1 x ptr], align 8
// CHECK1-NEXT: [[KERNEL_ARGS6:%.*]] = alloca [[STRUCT___TGT_KERNEL_ARGUMENTS]], align 8
// CHECK1-NEXT: store ptr [[THIS]], ptr [[THIS_ADDR]], align 8
// CHECK1-NEXT: store i32 [[N]], ptr [[N_ADDR]], align 4
// CHECK1-NEXT: [[THIS1:%.*]] = load ptr, ptr [[THIS_ADDR]], align 8
// CHECK1-NEXT: store i32 1, ptr [[B]], align 4
// CHECK1-NEXT: [[TMP0:%.*]] = load i32, ptr [[N_ADDR]], align 4
// CHECK1-NEXT: [[TMP1:%.*]] = load i32, ptr [[B]], align 4
// CHECK1-NEXT: [[SUB:%.*]] = sub nsw i32 [[TMP0]], [[TMP1]]
// CHECK1-NEXT: store i32 [[SUB]], ptr [[DOTCAPTURE_EXPR_]], align 4
// CHECK1-NEXT: [[TMP2:%.*]] = load i32, ptr [[B]], align 4
// CHECK1-NEXT: store i32 [[TMP2]], ptr [[B_CASTED]], align 4
// CHECK1-NEXT: [[TMP3:%.*]] = load i64, ptr [[B_CASTED]], align 8
// CHECK1-NEXT: [[TMP4:%.*]] = load i32, ptr [[DOTCAPTURE_EXPR_]], align 4
// CHECK1-NEXT: store i32 [[TMP4]], ptr [[DOTCAPTURE_EXPR__CASTED]], align 4
// CHECK1-NEXT: [[TMP5:%.*]] = load i64, ptr [[DOTCAPTURE_EXPR__CASTED]], align 8
// CHECK1-NEXT: [[A:%.*]] = getelementptr inbounds [[STRUCT_S1:%.*]], ptr [[THIS1]], i32 0, i32 0
// CHECK1-NEXT: [[TMP6:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_BASEPTRS]], i32 0, i32 0
// CHECK1-NEXT: store ptr [[THIS1]], ptr [[TMP6]], align 8
// CHECK1-NEXT: [[TMP7:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_PTRS]], i32 0, i32 0
// CHECK1-NEXT: store ptr [[A]], ptr [[TMP7]], align 8
// CHECK1-NEXT: [[TMP8:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_MAPPERS]], i64 0, i64 0
// CHECK1-NEXT: store ptr null, ptr [[TMP8]], align 8
// CHECK1-NEXT: [[TMP9:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_BASEPTRS]], i32 0, i32 1
// CHECK1-NEXT: store i64 [[TMP3]], ptr [[TMP9]], align 8
// CHECK1-NEXT: [[TMP10:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_PTRS]], i32 0, i32 1
// CHECK1-NEXT: store i64 [[TMP3]], ptr [[TMP10]], align 8
// CHECK1-NEXT: [[TMP11:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_MAPPERS]], i64 0, i64 1
// CHECK1-NEXT: store ptr null, ptr [[TMP11]], align 8
// CHECK1-NEXT: [[TMP12:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_BASEPTRS]], i32 0, i32 2
// CHECK1-NEXT: store i64 [[TMP5]], ptr [[TMP12]], align 8
// CHECK1-NEXT: [[TMP13:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_PTRS]], i32 0, i32 2
// CHECK1-NEXT: store i64 [[TMP5]], ptr [[TMP13]], align 8
// CHECK1-NEXT: [[TMP14:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_MAPPERS]], i64 0, i64 2
// CHECK1-NEXT: store ptr null, ptr [[TMP14]], align 8
// CHECK1-NEXT: [[TMP15:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_BASEPTRS]], i32 0, i32 0
// CHECK1-NEXT: [[TMP16:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_PTRS]], i32 0, i32 0
// CHECK1-NEXT: [[TMP17:%.*]] = load i32, ptr [[DOTCAPTURE_EXPR_]], align 4
// CHECK1-NEXT: [[TMP18:%.*]] = insertvalue [3 x i32] zeroinitializer, i32 [[TMP17]], 0
// CHECK1-NEXT: [[TMP19:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 0
// CHECK1-NEXT: store i32 2, ptr [[TMP19]], align 4
// CHECK1-NEXT: [[TMP20:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 1
// CHECK1-NEXT: store i32 3, ptr [[TMP20]], align 4
// CHECK1-NEXT: [[TMP21:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 2
// CHECK1-NEXT: store ptr [[TMP15]], ptr [[TMP21]], align 8
// CHECK1-NEXT: [[TMP22:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 3
// CHECK1-NEXT: store ptr [[TMP16]], ptr [[TMP22]], align 8
// CHECK1-NEXT: [[TMP23:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 4
// CHECK1-NEXT: store ptr @.offload_sizes, ptr [[TMP23]], align 8
// CHECK1-NEXT: [[TMP24:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 5
// CHECK1-NEXT: store ptr @.offload_maptypes, ptr [[TMP24]], align 8
// CHECK1-NEXT: [[TMP25:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 6
// CHECK1-NEXT: store ptr null, ptr [[TMP25]], align 8
// CHECK1-NEXT: [[TMP26:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 7
// CHECK1-NEXT: store ptr null, ptr [[TMP26]], align 8
// CHECK1-NEXT: [[TMP27:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 8
// CHECK1-NEXT: store i64 0, ptr [[TMP27]], align 8
// CHECK1-NEXT: [[TMP28:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 9
// CHECK1-NEXT: store i64 0, ptr [[TMP28]], align 8
// CHECK1-NEXT: [[TMP29:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 10
// CHECK1-NEXT: store [3 x i32] [i32 1, i32 0, i32 0], ptr [[TMP29]], align 4
// CHECK1-NEXT: [[TMP30:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 11
// CHECK1-NEXT: store [3 x i32] [[TMP18]], ptr [[TMP30]], align 4
// CHECK1-NEXT: [[TMP31:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 12
// CHECK1-NEXT: store i32 0, ptr [[TMP31]], align 4
// CHECK1-NEXT: [[TMP32:%.*]] = call i32 @__tgt_target_kernel(ptr @[[GLOB1:[0-9]+]], i64 -1, i32 1, i32 [[TMP17]], ptr @.{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZN2S12r1Ei_l121.region_id, ptr [[KERNEL_ARGS]])
// CHECK1-NEXT: [[TMP33:%.*]] = icmp ne i32 [[TMP32]], 0
// CHECK1-NEXT: br i1 [[TMP33]], label [[OMP_OFFLOAD_FAILED:%.*]], label [[OMP_OFFLOAD_CONT:%.*]]
// CHECK1: omp_offload.failed:
// CHECK1-NEXT: call void @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZN2S12r1Ei_l121(ptr [[THIS1]], i64 [[TMP3]], i64 [[TMP5]]) #[[ATTR3:[0-9]+]]
// CHECK1-NEXT: br label [[OMP_OFFLOAD_CONT]]
// CHECK1: omp_offload.cont:
// CHECK1-NEXT: [[A2:%.*]] = getelementptr inbounds [[STRUCT_S1]], ptr [[THIS1]], i32 0, i32 0
// CHECK1-NEXT: [[TMP34:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_BASEPTRS3]], i32 0, i32 0
// CHECK1-NEXT: store ptr [[THIS1]], ptr [[TMP34]], align 8
// CHECK1-NEXT: [[TMP35:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_PTRS4]], i32 0, i32 0
// CHECK1-NEXT: store ptr [[A2]], ptr [[TMP35]], align 8
// CHECK1-NEXT: [[TMP36:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_MAPPERS5]], i64 0, i64 0
// CHECK1-NEXT: store ptr null, ptr [[TMP36]], align 8
// CHECK1-NEXT: [[TMP37:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_BASEPTRS3]], i32 0, i32 0
// CHECK1-NEXT: [[TMP38:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_PTRS4]], i32 0, i32 0
// CHECK1-NEXT: [[TMP39:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 0
// CHECK1-NEXT: store i32 2, ptr [[TMP39]], align 4
// CHECK1-NEXT: [[TMP40:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 1
// CHECK1-NEXT: store i32 1, ptr [[TMP40]], align 4
// CHECK1-NEXT: [[TMP41:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 2
// CHECK1-NEXT: store ptr [[TMP37]], ptr [[TMP41]], align 8
// CHECK1-NEXT: [[TMP42:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 3
// CHECK1-NEXT: store ptr [[TMP38]], ptr [[TMP42]], align 8
// CHECK1-NEXT: [[TMP43:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 4
// CHECK1-NEXT: store ptr @.offload_sizes.1, ptr [[TMP43]], align 8
// CHECK1-NEXT: [[TMP44:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 5
// CHECK1-NEXT: store ptr @.offload_maptypes.2, ptr [[TMP44]], align 8
// CHECK1-NEXT: [[TMP45:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 6
// CHECK1-NEXT: store ptr null, ptr [[TMP45]], align 8
// CHECK1-NEXT: [[TMP46:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 7
// CHECK1-NEXT: store ptr null, ptr [[TMP46]], align 8
// CHECK1-NEXT: [[TMP47:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 8
// CHECK1-NEXT: store i64 0, ptr [[TMP47]], align 8
// CHECK1-NEXT: [[TMP48:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 9
// CHECK1-NEXT: store i64 0, ptr [[TMP48]], align 8
// CHECK1-NEXT: [[TMP49:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 10
// CHECK1-NEXT: store [3 x i32] [i32 1, i32 0, i32 0], ptr [[TMP49]], align 4
// CHECK1-NEXT: [[TMP50:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 11
// CHECK1-NEXT: store [3 x i32] [i32 1024, i32 0, i32 0], ptr [[TMP50]], align 4
// CHECK1-NEXT: [[TMP51:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 12
// CHECK1-NEXT: store i32 0, ptr [[TMP51]], align 4
// CHECK1-NEXT: [[TMP52:%.*]] = call i32 @__tgt_target_kernel(ptr @[[GLOB1]], i64 -1, i32 1, i32 1024, ptr @.{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZN2S12r1Ei_l126.region_id, ptr [[KERNEL_ARGS6]])
// CHECK1-NEXT: [[TMP53:%.*]] = icmp ne i32 [[TMP52]], 0
// CHECK1-NEXT: br i1 [[TMP53]], label [[OMP_OFFLOAD_FAILED7:%.*]], label [[OMP_OFFLOAD_CONT8:%.*]]
// CHECK1: omp_offload.failed7:
// CHECK1-NEXT: call void @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZN2S12r1Ei_l126(ptr [[THIS1]]) #[[ATTR3]]
// CHECK1-NEXT: br label [[OMP_OFFLOAD_CONT8]]
// CHECK1: omp_offload.cont8:
// CHECK1-NEXT: [[A9:%.*]] = getelementptr inbounds [[STRUCT_S1]], ptr [[THIS1]], i32 0, i32 0
// CHECK1-NEXT: [[TMP54:%.*]] = load double, ptr [[A9]], align 8
// CHECK1-NEXT: [[CONV:%.*]] = fptosi double [[TMP54]] to i32
// CHECK1-NEXT: ret i32 [[CONV]]
//
//
// CHECK1-LABEL: define {{[^@]+}}@_ZL7fstatici
// CHECK1-SAME: (i32 noundef signext [[N:%.*]]) #[[ATTR0]] {
// CHECK1-NEXT: entry:
// CHECK1-NEXT: [[N_ADDR:%.*]] = alloca i32, align 4
// CHECK1-NEXT: [[DOTCAPTURE_EXPR_:%.*]] = alloca i32, align 4
// CHECK1-NEXT: [[DOTCAPTURE_EXPR__CASTED:%.*]] = alloca i64, align 8
// CHECK1-NEXT: [[DOTOFFLOAD_BASEPTRS:%.*]] = alloca [1 x ptr], align 8
// CHECK1-NEXT: [[DOTOFFLOAD_PTRS:%.*]] = alloca [1 x ptr], align 8
// CHECK1-NEXT: [[DOTOFFLOAD_MAPPERS:%.*]] = alloca [1 x ptr], align 8
// CHECK1-NEXT: [[KERNEL_ARGS:%.*]] = alloca [[STRUCT___TGT_KERNEL_ARGUMENTS:%.*]], align 8
// CHECK1-NEXT: [[DOTCAPTURE_EXPR_1:%.*]] = alloca i32, align 4
// CHECK1-NEXT: [[DOTCAPTURE_EXPR__CASTED2:%.*]] = alloca i64, align 8
// CHECK1-NEXT: [[DOTOFFLOAD_BASEPTRS3:%.*]] = alloca [1 x ptr], align 8
// CHECK1-NEXT: [[DOTOFFLOAD_PTRS4:%.*]] = alloca [1 x ptr], align 8
// CHECK1-NEXT: [[DOTOFFLOAD_MAPPERS5:%.*]] = alloca [1 x ptr], align 8
// CHECK1-NEXT: [[KERNEL_ARGS6:%.*]] = alloca [[STRUCT___TGT_KERNEL_ARGUMENTS]], align 8
// CHECK1-NEXT: store i32 [[N]], ptr [[N_ADDR]], align 4
// CHECK1-NEXT: [[TMP0:%.*]] = load i32, ptr [[N_ADDR]], align 4
// CHECK1-NEXT: store i32 [[TMP0]], ptr [[DOTCAPTURE_EXPR_]], align 4
// CHECK1-NEXT: [[TMP1:%.*]] = load i32, ptr [[DOTCAPTURE_EXPR_]], align 4
// CHECK1-NEXT: store i32 [[TMP1]], ptr [[DOTCAPTURE_EXPR__CASTED]], align 4
// CHECK1-NEXT: [[TMP2:%.*]] = load i64, ptr [[DOTCAPTURE_EXPR__CASTED]], align 8
// CHECK1-NEXT: [[TMP3:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_BASEPTRS]], i32 0, i32 0
// CHECK1-NEXT: store i64 [[TMP2]], ptr [[TMP3]], align 8
// CHECK1-NEXT: [[TMP4:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_PTRS]], i32 0, i32 0
// CHECK1-NEXT: store i64 [[TMP2]], ptr [[TMP4]], align 8
// CHECK1-NEXT: [[TMP5:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_MAPPERS]], i64 0, i64 0
// CHECK1-NEXT: store ptr null, ptr [[TMP5]], align 8
// CHECK1-NEXT: [[TMP6:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_BASEPTRS]], i32 0, i32 0
// CHECK1-NEXT: [[TMP7:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_PTRS]], i32 0, i32 0
// CHECK1-NEXT: [[TMP8:%.*]] = load i32, ptr [[DOTCAPTURE_EXPR_]], align 4
// CHECK1-NEXT: [[TMP9:%.*]] = insertvalue [3 x i32] zeroinitializer, i32 [[TMP8]], 0
// CHECK1-NEXT: [[TMP10:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 0
// CHECK1-NEXT: store i32 2, ptr [[TMP10]], align 4
// CHECK1-NEXT: [[TMP11:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 1
// CHECK1-NEXT: store i32 1, ptr [[TMP11]], align 4
// CHECK1-NEXT: [[TMP12:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 2
// CHECK1-NEXT: store ptr [[TMP6]], ptr [[TMP12]], align 8
// CHECK1-NEXT: [[TMP13:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 3
// CHECK1-NEXT: store ptr [[TMP7]], ptr [[TMP13]], align 8
// CHECK1-NEXT: [[TMP14:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 4
// CHECK1-NEXT: store ptr @.offload_sizes.3, ptr [[TMP14]], align 8
// CHECK1-NEXT: [[TMP15:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 5
// CHECK1-NEXT: store ptr @.offload_maptypes.4, ptr [[TMP15]], align 8
// CHECK1-NEXT: [[TMP16:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 6
// CHECK1-NEXT: store ptr null, ptr [[TMP16]], align 8
// CHECK1-NEXT: [[TMP17:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 7
// CHECK1-NEXT: store ptr null, ptr [[TMP17]], align 8
// CHECK1-NEXT: [[TMP18:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 8
// CHECK1-NEXT: store i64 0, ptr [[TMP18]], align 8
// CHECK1-NEXT: [[TMP19:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 9
// CHECK1-NEXT: store i64 0, ptr [[TMP19]], align 8
// CHECK1-NEXT: [[TMP20:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 10
// CHECK1-NEXT: store [3 x i32] [i32 1, i32 0, i32 0], ptr [[TMP20]], align 4
// CHECK1-NEXT: [[TMP21:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 11
// CHECK1-NEXT: store [3 x i32] [[TMP9]], ptr [[TMP21]], align 4
// CHECK1-NEXT: [[TMP22:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 12
// CHECK1-NEXT: store i32 0, ptr [[TMP22]], align 4
// CHECK1-NEXT: [[TMP23:%.*]] = call i32 @__tgt_target_kernel(ptr @[[GLOB1]], i64 -1, i32 1, i32 [[TMP8]], ptr @.{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZL7fstatici_l104.region_id, ptr [[KERNEL_ARGS]])
// CHECK1-NEXT: [[TMP24:%.*]] = icmp ne i32 [[TMP23]], 0
// CHECK1-NEXT: br i1 [[TMP24]], label [[OMP_OFFLOAD_FAILED:%.*]], label [[OMP_OFFLOAD_CONT:%.*]]
// CHECK1: omp_offload.failed:
// CHECK1-NEXT: call void @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZL7fstatici_l104(i64 [[TMP2]]) #[[ATTR3]]
// CHECK1-NEXT: br label [[OMP_OFFLOAD_CONT]]
// CHECK1: omp_offload.cont:
// CHECK1-NEXT: [[TMP25:%.*]] = load i32, ptr [[N_ADDR]], align 4
// CHECK1-NEXT: [[ADD:%.*]] = add nsw i32 32, [[TMP25]]
// CHECK1-NEXT: store i32 [[ADD]], ptr [[DOTCAPTURE_EXPR_1]], align 4
// CHECK1-NEXT: [[TMP26:%.*]] = load i32, ptr [[DOTCAPTURE_EXPR_1]], align 4
// CHECK1-NEXT: store i32 [[TMP26]], ptr [[DOTCAPTURE_EXPR__CASTED2]], align 4
// CHECK1-NEXT: [[TMP27:%.*]] = load i64, ptr [[DOTCAPTURE_EXPR__CASTED2]], align 8
// CHECK1-NEXT: [[TMP28:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_BASEPTRS3]], i32 0, i32 0
// CHECK1-NEXT: store i64 [[TMP27]], ptr [[TMP28]], align 8
// CHECK1-NEXT: [[TMP29:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_PTRS4]], i32 0, i32 0
// CHECK1-NEXT: store i64 [[TMP27]], ptr [[TMP29]], align 8
// CHECK1-NEXT: [[TMP30:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_MAPPERS5]], i64 0, i64 0
// CHECK1-NEXT: store ptr null, ptr [[TMP30]], align 8
// CHECK1-NEXT: [[TMP31:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_BASEPTRS3]], i32 0, i32 0
// CHECK1-NEXT: [[TMP32:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_PTRS4]], i32 0, i32 0
// CHECK1-NEXT: [[TMP33:%.*]] = load i32, ptr [[DOTCAPTURE_EXPR_1]], align 4
// CHECK1-NEXT: [[TMP34:%.*]] = insertvalue [3 x i32] zeroinitializer, i32 [[TMP33]], 0
// CHECK1-NEXT: [[TMP35:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 0
// CHECK1-NEXT: store i32 2, ptr [[TMP35]], align 4
// CHECK1-NEXT: [[TMP36:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 1
// CHECK1-NEXT: store i32 1, ptr [[TMP36]], align 4
// CHECK1-NEXT: [[TMP37:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 2
// CHECK1-NEXT: store ptr [[TMP31]], ptr [[TMP37]], align 8
// CHECK1-NEXT: [[TMP38:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 3
// CHECK1-NEXT: store ptr [[TMP32]], ptr [[TMP38]], align 8
// CHECK1-NEXT: [[TMP39:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 4
// CHECK1-NEXT: store ptr @.offload_sizes.5, ptr [[TMP39]], align 8
// CHECK1-NEXT: [[TMP40:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 5
// CHECK1-NEXT: store ptr @.offload_maptypes.6, ptr [[TMP40]], align 8
// CHECK1-NEXT: [[TMP41:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 6
// CHECK1-NEXT: store ptr null, ptr [[TMP41]], align 8
// CHECK1-NEXT: [[TMP42:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 7
// CHECK1-NEXT: store ptr null, ptr [[TMP42]], align 8
// CHECK1-NEXT: [[TMP43:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 8
// CHECK1-NEXT: store i64 0, ptr [[TMP43]], align 8
// CHECK1-NEXT: [[TMP44:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 9
// CHECK1-NEXT: store i64 0, ptr [[TMP44]], align 8
// CHECK1-NEXT: [[TMP45:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 10
// CHECK1-NEXT: store [3 x i32] [i32 1, i32 0, i32 0], ptr [[TMP45]], align 4
// CHECK1-NEXT: [[TMP46:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 11
// CHECK1-NEXT: store [3 x i32] [[TMP34]], ptr [[TMP46]], align 4
// CHECK1-NEXT: [[TMP47:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 12
// CHECK1-NEXT: store i32 0, ptr [[TMP47]], align 4
// CHECK1-NEXT: [[TMP48:%.*]] = call i32 @__tgt_target_kernel(ptr @[[GLOB1]], i64 -1, i32 1, i32 [[TMP33]], ptr @.{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZL7fstatici_l108.region_id, ptr [[KERNEL_ARGS6]])
// CHECK1-NEXT: [[TMP49:%.*]] = icmp ne i32 [[TMP48]], 0
// CHECK1-NEXT: br i1 [[TMP49]], label [[OMP_OFFLOAD_FAILED7:%.*]], label [[OMP_OFFLOAD_CONT8:%.*]]
// CHECK1: omp_offload.failed7:
// CHECK1-NEXT: call void @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZL7fstatici_l108(i64 [[TMP27]]) #[[ATTR3]]
// CHECK1-NEXT: br label [[OMP_OFFLOAD_CONT8]]
// CHECK1: omp_offload.cont8:
// CHECK1-NEXT: [[TMP50:%.*]] = load i32, ptr [[N_ADDR]], align 4
// CHECK1-NEXT: [[ADD9:%.*]] = add nsw i32 [[TMP50]], 1
// CHECK1-NEXT: ret i32 [[ADD9]]
//
//
// CHECK1-LABEL: define {{[^@]+}}@_Z9ftemplateIiET_i
// CHECK1-SAME: (i32 noundef signext [[N:%.*]]) #[[ATTR0]] comdat {
// CHECK1-NEXT: entry:
// CHECK1-NEXT: [[N_ADDR:%.*]] = alloca i32, align 4
// CHECK1-NEXT: [[A:%.*]] = alloca i32, align 4
// CHECK1-NEXT: [[KERNEL_ARGS:%.*]] = alloca [[STRUCT___TGT_KERNEL_ARGUMENTS:%.*]], align 8
// CHECK1-NEXT: [[B:%.*]] = alloca i16, align 2
// CHECK1-NEXT: [[DOTCAPTURE_EXPR_:%.*]] = alloca i16, align 2
// CHECK1-NEXT: [[A_CASTED:%.*]] = alloca i64, align 8
// CHECK1-NEXT: [[B_CASTED:%.*]] = alloca i64, align 8
// CHECK1-NEXT: [[DOTCAPTURE_EXPR__CASTED:%.*]] = alloca i64, align 8
// CHECK1-NEXT: [[DOTOFFLOAD_BASEPTRS:%.*]] = alloca [3 x ptr], align 8
// CHECK1-NEXT: [[DOTOFFLOAD_PTRS:%.*]] = alloca [3 x ptr], align 8
// CHECK1-NEXT: [[DOTOFFLOAD_MAPPERS:%.*]] = alloca [3 x ptr], align 8
// CHECK1-NEXT: [[KERNEL_ARGS1:%.*]] = alloca [[STRUCT___TGT_KERNEL_ARGUMENTS]], align 8
// CHECK1-NEXT: store i32 [[N]], ptr [[N_ADDR]], align 4
// CHECK1-NEXT: store i32 0, ptr [[A]], align 4
// CHECK1-NEXT: [[TMP0:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 0
// CHECK1-NEXT: store i32 2, ptr [[TMP0]], align 4
// CHECK1-NEXT: [[TMP1:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 1
// CHECK1-NEXT: store i32 0, ptr [[TMP1]], align 4
// CHECK1-NEXT: [[TMP2:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 2
// CHECK1-NEXT: store ptr null, ptr [[TMP2]], align 8
// CHECK1-NEXT: [[TMP3:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 3
// CHECK1-NEXT: store ptr null, ptr [[TMP3]], align 8
// CHECK1-NEXT: [[TMP4:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 4
// CHECK1-NEXT: store ptr null, ptr [[TMP4]], align 8
// CHECK1-NEXT: [[TMP5:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 5
// CHECK1-NEXT: store ptr null, ptr [[TMP5]], align 8
// CHECK1-NEXT: [[TMP6:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 6
// CHECK1-NEXT: store ptr null, ptr [[TMP6]], align 8
// CHECK1-NEXT: [[TMP7:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 7
// CHECK1-NEXT: store ptr null, ptr [[TMP7]], align 8
// CHECK1-NEXT: [[TMP8:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 8
// CHECK1-NEXT: store i64 0, ptr [[TMP8]], align 8
// CHECK1-NEXT: [[TMP9:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 9
// CHECK1-NEXT: store i64 0, ptr [[TMP9]], align 8
// CHECK1-NEXT: [[TMP10:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 10
// CHECK1-NEXT: store [3 x i32] [i32 1, i32 0, i32 0], ptr [[TMP10]], align 4
// CHECK1-NEXT: [[TMP11:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 11
// CHECK1-NEXT: store [3 x i32] [i32 20, i32 0, i32 0], ptr [[TMP11]], align 4
// CHECK1-NEXT: [[TMP12:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 12
// CHECK1-NEXT: store i32 0, ptr [[TMP12]], align 4
// CHECK1-NEXT: [[TMP13:%.*]] = call i32 @__tgt_target_kernel(ptr @[[GLOB1]], i64 -1, i32 1, i32 20, ptr @.{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIiET_i_l88.region_id, ptr [[KERNEL_ARGS]])
// CHECK1-NEXT: [[TMP14:%.*]] = icmp ne i32 [[TMP13]], 0
// CHECK1-NEXT: br i1 [[TMP14]], label [[OMP_OFFLOAD_FAILED:%.*]], label [[OMP_OFFLOAD_CONT:%.*]]
// CHECK1: omp_offload.failed:
// CHECK1-NEXT: call void @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIiET_i_l88() #[[ATTR3]]
// CHECK1-NEXT: br label [[OMP_OFFLOAD_CONT]]
// CHECK1: omp_offload.cont:
// CHECK1-NEXT: store i16 1, ptr [[B]], align 2
// CHECK1-NEXT: [[TMP15:%.*]] = load i16, ptr [[B]], align 2
// CHECK1-NEXT: store i16 [[TMP15]], ptr [[DOTCAPTURE_EXPR_]], align 2
// CHECK1-NEXT: [[TMP16:%.*]] = load i32, ptr [[A]], align 4
// CHECK1-NEXT: store i32 [[TMP16]], ptr [[A_CASTED]], align 4
// CHECK1-NEXT: [[TMP17:%.*]] = load i64, ptr [[A_CASTED]], align 8
// CHECK1-NEXT: [[TMP18:%.*]] = load i16, ptr [[B]], align 2
// CHECK1-NEXT: store i16 [[TMP18]], ptr [[B_CASTED]], align 2
// CHECK1-NEXT: [[TMP19:%.*]] = load i64, ptr [[B_CASTED]], align 8
// CHECK1-NEXT: [[TMP20:%.*]] = load i16, ptr [[DOTCAPTURE_EXPR_]], align 2
// CHECK1-NEXT: store i16 [[TMP20]], ptr [[DOTCAPTURE_EXPR__CASTED]], align 2
// CHECK1-NEXT: [[TMP21:%.*]] = load i64, ptr [[DOTCAPTURE_EXPR__CASTED]], align 8
// CHECK1-NEXT: [[TMP22:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_BASEPTRS]], i32 0, i32 0
// CHECK1-NEXT: store i64 [[TMP17]], ptr [[TMP22]], align 8
// CHECK1-NEXT: [[TMP23:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_PTRS]], i32 0, i32 0
// CHECK1-NEXT: store i64 [[TMP17]], ptr [[TMP23]], align 8
// CHECK1-NEXT: [[TMP24:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_MAPPERS]], i64 0, i64 0
// CHECK1-NEXT: store ptr null, ptr [[TMP24]], align 8
// CHECK1-NEXT: [[TMP25:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_BASEPTRS]], i32 0, i32 1
// CHECK1-NEXT: store i64 [[TMP19]], ptr [[TMP25]], align 8
// CHECK1-NEXT: [[TMP26:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_PTRS]], i32 0, i32 1
// CHECK1-NEXT: store i64 [[TMP19]], ptr [[TMP26]], align 8
// CHECK1-NEXT: [[TMP27:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_MAPPERS]], i64 0, i64 1
// CHECK1-NEXT: store ptr null, ptr [[TMP27]], align 8
// CHECK1-NEXT: [[TMP28:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_BASEPTRS]], i32 0, i32 2
// CHECK1-NEXT: store i64 [[TMP21]], ptr [[TMP28]], align 8
// CHECK1-NEXT: [[TMP29:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_PTRS]], i32 0, i32 2
// CHECK1-NEXT: store i64 [[TMP21]], ptr [[TMP29]], align 8
// CHECK1-NEXT: [[TMP30:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_MAPPERS]], i64 0, i64 2
// CHECK1-NEXT: store ptr null, ptr [[TMP30]], align 8
// CHECK1-NEXT: [[TMP31:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_BASEPTRS]], i32 0, i32 0
// CHECK1-NEXT: [[TMP32:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_PTRS]], i32 0, i32 0
// CHECK1-NEXT: [[TMP33:%.*]] = load i16, ptr [[DOTCAPTURE_EXPR_]], align 2
// CHECK1-NEXT: [[TMP34:%.*]] = zext i16 [[TMP33]] to i32
// CHECK1-NEXT: [[TMP35:%.*]] = insertvalue [3 x i32] zeroinitializer, i32 [[TMP34]], 0
// CHECK1-NEXT: [[TMP36:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS1]], i32 0, i32 0
// CHECK1-NEXT: store i32 2, ptr [[TMP36]], align 4
// CHECK1-NEXT: [[TMP37:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS1]], i32 0, i32 1
// CHECK1-NEXT: store i32 3, ptr [[TMP37]], align 4
// CHECK1-NEXT: [[TMP38:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS1]], i32 0, i32 2
// CHECK1-NEXT: store ptr [[TMP31]], ptr [[TMP38]], align 8
// CHECK1-NEXT: [[TMP39:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS1]], i32 0, i32 3
// CHECK1-NEXT: store ptr [[TMP32]], ptr [[TMP39]], align 8
// CHECK1-NEXT: [[TMP40:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS1]], i32 0, i32 4
// CHECK1-NEXT: store ptr @.offload_sizes.7, ptr [[TMP40]], align 8
// CHECK1-NEXT: [[TMP41:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS1]], i32 0, i32 5
// CHECK1-NEXT: store ptr @.offload_maptypes.8, ptr [[TMP41]], align 8
// CHECK1-NEXT: [[TMP42:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS1]], i32 0, i32 6
// CHECK1-NEXT: store ptr null, ptr [[TMP42]], align 8
// CHECK1-NEXT: [[TMP43:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS1]], i32 0, i32 7
// CHECK1-NEXT: store ptr null, ptr [[TMP43]], align 8
// CHECK1-NEXT: [[TMP44:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS1]], i32 0, i32 8
// CHECK1-NEXT: store i64 0, ptr [[TMP44]], align 8
// CHECK1-NEXT: [[TMP45:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS1]], i32 0, i32 9
// CHECK1-NEXT: store i64 0, ptr [[TMP45]], align 8
// CHECK1-NEXT: [[TMP46:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS1]], i32 0, i32 10
// CHECK1-NEXT: store [3 x i32] [i32 1, i32 0, i32 0], ptr [[TMP46]], align 4
// CHECK1-NEXT: [[TMP47:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS1]], i32 0, i32 11
// CHECK1-NEXT: store [3 x i32] [[TMP35]], ptr [[TMP47]], align 4
// CHECK1-NEXT: [[TMP48:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS1]], i32 0, i32 12
// CHECK1-NEXT: store i32 0, ptr [[TMP48]], align 4
// CHECK1-NEXT: [[TMP49:%.*]] = call i32 @__tgt_target_kernel(ptr @[[GLOB1]], i64 -1, i32 1, i32 [[TMP34]], ptr @.{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIiET_i_l93.region_id, ptr [[KERNEL_ARGS1]])
// CHECK1-NEXT: [[TMP50:%.*]] = icmp ne i32 [[TMP49]], 0
// CHECK1-NEXT: br i1 [[TMP50]], label [[OMP_OFFLOAD_FAILED2:%.*]], label [[OMP_OFFLOAD_CONT3:%.*]]
// CHECK1: omp_offload.failed2:
// CHECK1-NEXT: call void @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIiET_i_l93(i64 [[TMP17]], i64 [[TMP19]], i64 [[TMP21]]) #[[ATTR3]]
// CHECK1-NEXT: br label [[OMP_OFFLOAD_CONT3]]
// CHECK1: omp_offload.cont3:
// CHECK1-NEXT: [[TMP51:%.*]] = load i32, ptr [[A]], align 4
// CHECK1-NEXT: ret i32 [[TMP51]]
//
//
// CHECK1-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZN2S12r1Ei_l121
// CHECK1-SAME: (ptr noundef [[THIS:%.*]], i64 noundef [[B:%.*]], i64 noundef [[DOTCAPTURE_EXPR_:%.*]]) #[[ATTR1:[0-9]+]] {
// CHECK1-NEXT: entry:
// CHECK1-NEXT: [[THIS_ADDR:%.*]] = alloca ptr, align 8
// CHECK1-NEXT: [[B_ADDR:%.*]] = alloca i64, align 8
// CHECK1-NEXT: [[DOTCAPTURE_EXPR__ADDR:%.*]] = alloca i64, align 8
// CHECK1-NEXT: [[B_CASTED:%.*]] = alloca i64, align 8
// CHECK1-NEXT: [[TMP0:%.*]] = call i32 @__kmpc_global_thread_num(ptr @[[GLOB1]])
// CHECK1-NEXT: store ptr [[THIS]], ptr [[THIS_ADDR]], align 8
// CHECK1-NEXT: store i64 [[B]], ptr [[B_ADDR]], align 8
// CHECK1-NEXT: store i64 [[DOTCAPTURE_EXPR_]], ptr [[DOTCAPTURE_EXPR__ADDR]], align 8
// CHECK1-NEXT: [[TMP1:%.*]] = load ptr, ptr [[THIS_ADDR]], align 8
// CHECK1-NEXT: [[TMP2:%.*]] = load i32, ptr [[DOTCAPTURE_EXPR__ADDR]], align 4
// CHECK1-NEXT: call void @__kmpc_push_num_threads(ptr @[[GLOB1]], i32 [[TMP0]], i32 [[TMP2]])
// CHECK1-NEXT: [[TMP3:%.*]] = load i32, ptr [[B_ADDR]], align 4
// CHECK1-NEXT: store i32 [[TMP3]], ptr [[B_CASTED]], align 4
// CHECK1-NEXT: [[TMP4:%.*]] = load i64, ptr [[B_CASTED]], align 8
// CHECK1-NEXT: call void (ptr, i32, ptr, ...) @__kmpc_fork_call(ptr @[[GLOB1]], i32 2, ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZN2S12r1Ei_l121.omp_outlined, ptr [[TMP1]], i64 [[TMP4]])
// CHECK1-NEXT: ret void
//
//
// CHECK1-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZN2S12r1Ei_l121.omp_outlined
// CHECK1-SAME: (ptr noalias noundef [[DOTGLOBAL_TID_:%.*]], ptr noalias noundef [[DOTBOUND_TID_:%.*]], ptr noundef [[THIS:%.*]], i64 noundef [[B:%.*]]) #[[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: [[THIS_ADDR:%.*]] = alloca ptr, align 8
// CHECK1-NEXT: [[B_ADDR:%.*]] = alloca i64, 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 [[THIS]], ptr [[THIS_ADDR]], align 8
// CHECK1-NEXT: store i64 [[B]], ptr [[B_ADDR]], align 8
// CHECK1-NEXT: [[TMP0:%.*]] = load ptr, ptr [[THIS_ADDR]], align 8
// CHECK1-NEXT: [[TMP1:%.*]] = load i32, ptr [[B_ADDR]], align 4
// CHECK1-NEXT: [[CONV:%.*]] = sitofp i32 [[TMP1]] to double
// CHECK1-NEXT: [[ADD:%.*]] = fadd double [[CONV]], 1.500000e+00
// CHECK1-NEXT: [[A:%.*]] = getelementptr inbounds [[STRUCT_S1:%.*]], ptr [[TMP0]], i32 0, i32 0
// CHECK1-NEXT: store double [[ADD]], ptr [[A]], align 8
// CHECK1-NEXT: ret void
//
//
// CHECK1-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZN2S12r1Ei_l126
// CHECK1-SAME: (ptr noundef [[THIS:%.*]]) #[[ATTR1]] {
// CHECK1-NEXT: entry:
// CHECK1-NEXT: [[THIS_ADDR:%.*]] = alloca ptr, align 8
// CHECK1-NEXT: [[TMP0:%.*]] = call i32 @__kmpc_global_thread_num(ptr @[[GLOB1]])
// CHECK1-NEXT: store ptr [[THIS]], ptr [[THIS_ADDR]], align 8
// CHECK1-NEXT: [[TMP1:%.*]] = load ptr, ptr [[THIS_ADDR]], align 8
// CHECK1-NEXT: call void @__kmpc_push_num_threads(ptr @[[GLOB1]], i32 [[TMP0]], i32 1024)
// CHECK1-NEXT: call void (ptr, i32, ptr, ...) @__kmpc_fork_call(ptr @[[GLOB1]], i32 1, ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZN2S12r1Ei_l126.omp_outlined, ptr [[TMP1]])
// CHECK1-NEXT: ret void
//
//
// CHECK1-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZN2S12r1Ei_l126.omp_outlined
// CHECK1-SAME: (ptr noalias noundef [[DOTGLOBAL_TID_:%.*]], ptr noalias noundef [[DOTBOUND_TID_:%.*]], ptr noundef [[THIS:%.*]]) #[[ATTR2]] {
// CHECK1-NEXT: entry:
// CHECK1-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca ptr, align 8
// CHECK1-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca ptr, align 8
// CHECK1-NEXT: [[THIS_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 [[THIS]], ptr [[THIS_ADDR]], align 8
// CHECK1-NEXT: [[TMP0:%.*]] = load ptr, ptr [[THIS_ADDR]], align 8
// CHECK1-NEXT: [[A:%.*]] = getelementptr inbounds [[STRUCT_S1:%.*]], ptr [[TMP0]], i32 0, i32 0
// CHECK1-NEXT: store double 2.500000e+00, ptr [[A]], align 8
// CHECK1-NEXT: ret void
//
//
// CHECK1-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZL7fstatici_l104
// CHECK1-SAME: (i64 noundef [[DOTCAPTURE_EXPR_:%.*]]) #[[ATTR1]] {
// CHECK1-NEXT: entry:
// CHECK1-NEXT: [[DOTCAPTURE_EXPR__ADDR:%.*]] = alloca i64, align 8
// CHECK1-NEXT: [[TMP0:%.*]] = call i32 @__kmpc_global_thread_num(ptr @[[GLOB1]])
// CHECK1-NEXT: store i64 [[DOTCAPTURE_EXPR_]], ptr [[DOTCAPTURE_EXPR__ADDR]], align 8
// CHECK1-NEXT: [[TMP1:%.*]] = load i32, ptr [[DOTCAPTURE_EXPR__ADDR]], align 4
// CHECK1-NEXT: call void @__kmpc_push_num_threads(ptr @[[GLOB1]], i32 [[TMP0]], i32 [[TMP1]])
// CHECK1-NEXT: call void (ptr, i32, ptr, ...) @__kmpc_fork_call(ptr @[[GLOB1]], i32 0, ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZL7fstatici_l104.omp_outlined)
// CHECK1-NEXT: ret void
//
//
// CHECK1-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZL7fstatici_l104.omp_outlined
// CHECK1-SAME: (ptr noalias noundef [[DOTGLOBAL_TID_:%.*]], ptr noalias noundef [[DOTBOUND_TID_:%.*]]) #[[ATTR2]] {
// CHECK1-NEXT: entry:
// CHECK1-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca ptr, align 8
// CHECK1-NEXT: [[DOTBOUND_TID__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: ret void
//
//
// CHECK1-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZL7fstatici_l108
// CHECK1-SAME: (i64 noundef [[DOTCAPTURE_EXPR_:%.*]]) #[[ATTR1]] {
// CHECK1-NEXT: entry:
// CHECK1-NEXT: [[DOTCAPTURE_EXPR__ADDR:%.*]] = alloca i64, align 8
// CHECK1-NEXT: [[TMP0:%.*]] = call i32 @__kmpc_global_thread_num(ptr @[[GLOB1]])
// CHECK1-NEXT: store i64 [[DOTCAPTURE_EXPR_]], ptr [[DOTCAPTURE_EXPR__ADDR]], align 8
// CHECK1-NEXT: [[TMP1:%.*]] = load i32, ptr [[DOTCAPTURE_EXPR__ADDR]], align 4
// CHECK1-NEXT: call void @__kmpc_push_num_threads(ptr @[[GLOB1]], i32 [[TMP0]], i32 [[TMP1]])
// CHECK1-NEXT: call void (ptr, i32, ptr, ...) @__kmpc_fork_call(ptr @[[GLOB1]], i32 0, ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZL7fstatici_l108.omp_outlined)
// CHECK1-NEXT: ret void
//
//
// CHECK1-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZL7fstatici_l108.omp_outlined
// CHECK1-SAME: (ptr noalias noundef [[DOTGLOBAL_TID_:%.*]], ptr noalias noundef [[DOTBOUND_TID_:%.*]]) #[[ATTR2]] {
// CHECK1-NEXT: entry:
// CHECK1-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca ptr, align 8
// CHECK1-NEXT: [[DOTBOUND_TID__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: ret void
//
//
// CHECK1-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIiET_i_l88
// CHECK1-SAME: () #[[ATTR1]] {
// CHECK1-NEXT: entry:
// CHECK1-NEXT: [[TMP0:%.*]] = call i32 @__kmpc_global_thread_num(ptr @[[GLOB1]])
// CHECK1-NEXT: call void @__kmpc_push_num_threads(ptr @[[GLOB1]], i32 [[TMP0]], i32 20)
// CHECK1-NEXT: call void (ptr, i32, ptr, ...) @__kmpc_fork_call(ptr @[[GLOB1]], i32 0, ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIiET_i_l88.omp_outlined)
// CHECK1-NEXT: ret void
//
//
// CHECK1-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIiET_i_l88.omp_outlined
// CHECK1-SAME: (ptr noalias noundef [[DOTGLOBAL_TID_:%.*]], ptr noalias noundef [[DOTBOUND_TID_:%.*]]) #[[ATTR2]] {
// CHECK1-NEXT: entry:
// CHECK1-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca ptr, align 8
// CHECK1-NEXT: [[DOTBOUND_TID__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: ret void
//
//
// CHECK1-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIiET_i_l93
// CHECK1-SAME: (i64 noundef [[A:%.*]], i64 noundef [[B:%.*]], i64 noundef [[DOTCAPTURE_EXPR_:%.*]]) #[[ATTR1]] {
// CHECK1-NEXT: entry:
// CHECK1-NEXT: [[A_ADDR:%.*]] = alloca i64, align 8
// CHECK1-NEXT: [[B_ADDR:%.*]] = alloca i64, align 8
// CHECK1-NEXT: [[DOTCAPTURE_EXPR__ADDR:%.*]] = alloca i64, align 8
// CHECK1-NEXT: [[A_CASTED:%.*]] = alloca i64, align 8
// CHECK1-NEXT: [[B_CASTED:%.*]] = alloca i64, align 8
// CHECK1-NEXT: [[TMP0:%.*]] = call i32 @__kmpc_global_thread_num(ptr @[[GLOB1]])
// CHECK1-NEXT: store i64 [[A]], ptr [[A_ADDR]], align 8
// CHECK1-NEXT: store i64 [[B]], ptr [[B_ADDR]], align 8
// CHECK1-NEXT: store i64 [[DOTCAPTURE_EXPR_]], ptr [[DOTCAPTURE_EXPR__ADDR]], align 8
// CHECK1-NEXT: [[TMP1:%.*]] = load i16, ptr [[DOTCAPTURE_EXPR__ADDR]], align 2
// CHECK1-NEXT: [[TMP2:%.*]] = sext i16 [[TMP1]] to i32
// CHECK1-NEXT: call void @__kmpc_push_num_threads(ptr @[[GLOB1]], i32 [[TMP0]], i32 [[TMP2]])
// CHECK1-NEXT: [[TMP3:%.*]] = load i32, ptr [[A_ADDR]], align 4
// CHECK1-NEXT: store i32 [[TMP3]], ptr [[A_CASTED]], align 4
// CHECK1-NEXT: [[TMP4:%.*]] = load i64, ptr [[A_CASTED]], align 8
// CHECK1-NEXT: [[TMP5:%.*]] = load i16, ptr [[B_ADDR]], align 2
// CHECK1-NEXT: store i16 [[TMP5]], ptr [[B_CASTED]], align 2
// CHECK1-NEXT: [[TMP6:%.*]] = load i64, ptr [[B_CASTED]], align 8
// CHECK1-NEXT: call void (ptr, i32, ptr, ...) @__kmpc_fork_call(ptr @[[GLOB1]], i32 2, ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIiET_i_l93.omp_outlined, i64 [[TMP4]], i64 [[TMP6]])
// CHECK1-NEXT: ret void
//
//
// CHECK1-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIiET_i_l93.omp_outlined
// CHECK1-SAME: (ptr noalias noundef [[DOTGLOBAL_TID_:%.*]], ptr noalias noundef [[DOTBOUND_TID_:%.*]], i64 noundef [[A:%.*]], i64 noundef [[B:%.*]]) #[[ATTR2]] {
// CHECK1-NEXT: entry:
// CHECK1-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca ptr, align 8
// CHECK1-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca ptr, align 8
// CHECK1-NEXT: [[A_ADDR:%.*]] = alloca i64, align 8
// CHECK1-NEXT: [[B_ADDR:%.*]] = alloca i64, 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 i64 [[A]], ptr [[A_ADDR]], align 8
// CHECK1-NEXT: store i64 [[B]], ptr [[B_ADDR]], align 8
// CHECK1-NEXT: [[TMP0:%.*]] = load i16, ptr [[B_ADDR]], align 2
// CHECK1-NEXT: [[CONV:%.*]] = sext i16 [[TMP0]] to i32
// CHECK1-NEXT: [[TMP1:%.*]] = load i32, ptr [[A_ADDR]], align 4
// CHECK1-NEXT: [[ADD:%.*]] = add nsw i32 [[TMP1]], [[CONV]]
// CHECK1-NEXT: store i32 [[ADD]], ptr [[A_ADDR]], align 4
// CHECK1-NEXT: ret void
//
//
// CHECK1-LABEL: define {{[^@]+}}@.omp_offloading.requires_reg
// CHECK1-SAME: () #[[ATTR4:[0-9]+]] {
// CHECK1-NEXT: entry:
// CHECK1-NEXT: call void @__tgt_register_requires(i64 1)
// CHECK1-NEXT: ret void
//
//
// CHECK3-LABEL: define {{[^@]+}}@_Z3bari
// CHECK3-SAME: (i32 noundef [[N:%.*]]) #[[ATTR0:[0-9]+]] {
// CHECK3-NEXT: entry:
// CHECK3-NEXT: [[N_ADDR:%.*]] = alloca i32, align 4
// CHECK3-NEXT: [[A:%.*]] = alloca i32, align 4
// CHECK3-NEXT: [[S:%.*]] = alloca [[STRUCT_S1:%.*]], align 4
// CHECK3-NEXT: store i32 [[N]], ptr [[N_ADDR]], align 4
// CHECK3-NEXT: store i32 0, ptr [[A]], align 4
// CHECK3-NEXT: [[TMP0:%.*]] = load i32, ptr [[N_ADDR]], align 4
// CHECK3-NEXT: [[CALL:%.*]] = call noundef i32 @_ZN2S12r1Ei(ptr noundef nonnull align 4 dereferenceable(8) [[S]], i32 noundef [[TMP0]])
// CHECK3-NEXT: [[TMP1:%.*]] = load i32, ptr [[A]], align 4
// CHECK3-NEXT: [[ADD:%.*]] = add nsw i32 [[TMP1]], [[CALL]]
// CHECK3-NEXT: store i32 [[ADD]], ptr [[A]], align 4
// CHECK3-NEXT: [[TMP2:%.*]] = load i32, ptr [[N_ADDR]], align 4
// CHECK3-NEXT: [[CALL1:%.*]] = call noundef i32 @_ZL7fstatici(i32 noundef [[TMP2]])
// CHECK3-NEXT: [[TMP3:%.*]] = load i32, ptr [[A]], align 4
// CHECK3-NEXT: [[ADD2:%.*]] = add nsw i32 [[TMP3]], [[CALL1]]
// CHECK3-NEXT: store i32 [[ADD2]], ptr [[A]], align 4
// CHECK3-NEXT: [[TMP4:%.*]] = load i32, ptr [[N_ADDR]], align 4
// CHECK3-NEXT: [[CALL3:%.*]] = call noundef i32 @_Z9ftemplateIiET_i(i32 noundef [[TMP4]])
// CHECK3-NEXT: [[TMP5:%.*]] = load i32, ptr [[A]], align 4
// CHECK3-NEXT: [[ADD4:%.*]] = add nsw i32 [[TMP5]], [[CALL3]]
// CHECK3-NEXT: store i32 [[ADD4]], ptr [[A]], align 4
// CHECK3-NEXT: [[TMP6:%.*]] = load i32, ptr [[A]], align 4
// CHECK3-NEXT: ret i32 [[TMP6]]
//
//
// CHECK3-LABEL: define {{[^@]+}}@_ZN2S12r1Ei
// CHECK3-SAME: (ptr noundef nonnull align 4 dereferenceable(8) [[THIS:%.*]], i32 noundef [[N:%.*]]) #[[ATTR0]] comdat align 2 {
// CHECK3-NEXT: entry:
// CHECK3-NEXT: [[THIS_ADDR:%.*]] = alloca ptr, align 4
// CHECK3-NEXT: [[N_ADDR:%.*]] = alloca i32, align 4
// CHECK3-NEXT: [[B:%.*]] = alloca i32, align 4
// CHECK3-NEXT: [[DOTCAPTURE_EXPR_:%.*]] = alloca i32, align 4
// CHECK3-NEXT: [[B_CASTED:%.*]] = alloca i32, align 4
// CHECK3-NEXT: [[DOTCAPTURE_EXPR__CASTED:%.*]] = alloca i32, align 4
// CHECK3-NEXT: [[DOTOFFLOAD_BASEPTRS:%.*]] = alloca [3 x ptr], align 4
// CHECK3-NEXT: [[DOTOFFLOAD_PTRS:%.*]] = alloca [3 x ptr], align 4
// CHECK3-NEXT: [[DOTOFFLOAD_MAPPERS:%.*]] = alloca [3 x ptr], align 4
// CHECK3-NEXT: [[KERNEL_ARGS:%.*]] = alloca [[STRUCT___TGT_KERNEL_ARGUMENTS:%.*]], align 8
// CHECK3-NEXT: [[DOTOFFLOAD_BASEPTRS3:%.*]] = alloca [1 x ptr], align 4
// CHECK3-NEXT: [[DOTOFFLOAD_PTRS4:%.*]] = alloca [1 x ptr], align 4
// CHECK3-NEXT: [[DOTOFFLOAD_MAPPERS5:%.*]] = alloca [1 x ptr], align 4
// CHECK3-NEXT: [[KERNEL_ARGS6:%.*]] = alloca [[STRUCT___TGT_KERNEL_ARGUMENTS]], align 8
// CHECK3-NEXT: store ptr [[THIS]], ptr [[THIS_ADDR]], align 4
// CHECK3-NEXT: store i32 [[N]], ptr [[N_ADDR]], align 4
// CHECK3-NEXT: [[THIS1:%.*]] = load ptr, ptr [[THIS_ADDR]], align 4
// CHECK3-NEXT: store i32 1, ptr [[B]], align 4
// CHECK3-NEXT: [[TMP0:%.*]] = load i32, ptr [[N_ADDR]], align 4
// CHECK3-NEXT: [[TMP1:%.*]] = load i32, ptr [[B]], align 4
// CHECK3-NEXT: [[SUB:%.*]] = sub nsw i32 [[TMP0]], [[TMP1]]
// CHECK3-NEXT: store i32 [[SUB]], ptr [[DOTCAPTURE_EXPR_]], align 4
// CHECK3-NEXT: [[TMP2:%.*]] = load i32, ptr [[B]], align 4
// CHECK3-NEXT: store i32 [[TMP2]], ptr [[B_CASTED]], align 4
// CHECK3-NEXT: [[TMP3:%.*]] = load i32, ptr [[B_CASTED]], align 4
// CHECK3-NEXT: [[TMP4:%.*]] = load i32, ptr [[DOTCAPTURE_EXPR_]], align 4
// CHECK3-NEXT: store i32 [[TMP4]], ptr [[DOTCAPTURE_EXPR__CASTED]], align 4
// CHECK3-NEXT: [[TMP5:%.*]] = load i32, ptr [[DOTCAPTURE_EXPR__CASTED]], align 4
// CHECK3-NEXT: [[A:%.*]] = getelementptr inbounds [[STRUCT_S1:%.*]], ptr [[THIS1]], i32 0, i32 0
// CHECK3-NEXT: [[TMP6:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_BASEPTRS]], i32 0, i32 0
// CHECK3-NEXT: store ptr [[THIS1]], ptr [[TMP6]], align 4
// CHECK3-NEXT: [[TMP7:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_PTRS]], i32 0, i32 0
// CHECK3-NEXT: store ptr [[A]], ptr [[TMP7]], align 4
// CHECK3-NEXT: [[TMP8:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_MAPPERS]], i32 0, i32 0
// CHECK3-NEXT: store ptr null, ptr [[TMP8]], align 4
// CHECK3-NEXT: [[TMP9:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_BASEPTRS]], i32 0, i32 1
// CHECK3-NEXT: store i32 [[TMP3]], ptr [[TMP9]], align 4
// CHECK3-NEXT: [[TMP10:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_PTRS]], i32 0, i32 1
// CHECK3-NEXT: store i32 [[TMP3]], ptr [[TMP10]], align 4
// CHECK3-NEXT: [[TMP11:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_MAPPERS]], i32 0, i32 1
// CHECK3-NEXT: store ptr null, ptr [[TMP11]], align 4
// CHECK3-NEXT: [[TMP12:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_BASEPTRS]], i32 0, i32 2
// CHECK3-NEXT: store i32 [[TMP5]], ptr [[TMP12]], align 4
// CHECK3-NEXT: [[TMP13:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_PTRS]], i32 0, i32 2
// CHECK3-NEXT: store i32 [[TMP5]], ptr [[TMP13]], align 4
// CHECK3-NEXT: [[TMP14:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_MAPPERS]], i32 0, i32 2
// CHECK3-NEXT: store ptr null, ptr [[TMP14]], align 4
// CHECK3-NEXT: [[TMP15:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_BASEPTRS]], i32 0, i32 0
// CHECK3-NEXT: [[TMP16:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_PTRS]], i32 0, i32 0
// CHECK3-NEXT: [[TMP17:%.*]] = load i32, ptr [[DOTCAPTURE_EXPR_]], align 4
// CHECK3-NEXT: [[TMP18:%.*]] = insertvalue [3 x i32] zeroinitializer, i32 [[TMP17]], 0
// CHECK3-NEXT: [[TMP19:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 0
// CHECK3-NEXT: store i32 2, ptr [[TMP19]], align 4
// CHECK3-NEXT: [[TMP20:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 1
// CHECK3-NEXT: store i32 3, ptr [[TMP20]], align 4
// CHECK3-NEXT: [[TMP21:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 2
// CHECK3-NEXT: store ptr [[TMP15]], ptr [[TMP21]], align 4
// CHECK3-NEXT: [[TMP22:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 3
// CHECK3-NEXT: store ptr [[TMP16]], ptr [[TMP22]], align 4
// CHECK3-NEXT: [[TMP23:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 4
// CHECK3-NEXT: store ptr @.offload_sizes, ptr [[TMP23]], align 4
// CHECK3-NEXT: [[TMP24:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 5
// CHECK3-NEXT: store ptr @.offload_maptypes, ptr [[TMP24]], align 4
// CHECK3-NEXT: [[TMP25:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 6
// CHECK3-NEXT: store ptr null, ptr [[TMP25]], align 4
// CHECK3-NEXT: [[TMP26:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 7
// CHECK3-NEXT: store ptr null, ptr [[TMP26]], align 4
// CHECK3-NEXT: [[TMP27:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 8
// CHECK3-NEXT: store i64 0, ptr [[TMP27]], align 8
// CHECK3-NEXT: [[TMP28:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 9
// CHECK3-NEXT: store i64 0, ptr [[TMP28]], align 8
// CHECK3-NEXT: [[TMP29:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 10
// CHECK3-NEXT: store [3 x i32] [i32 1, i32 0, i32 0], ptr [[TMP29]], align 4
// CHECK3-NEXT: [[TMP30:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 11
// CHECK3-NEXT: store [3 x i32] [[TMP18]], ptr [[TMP30]], align 4
// CHECK3-NEXT: [[TMP31:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 12
// CHECK3-NEXT: store i32 0, ptr [[TMP31]], align 4
// CHECK3-NEXT: [[TMP32:%.*]] = call i32 @__tgt_target_kernel(ptr @[[GLOB1:[0-9]+]], i64 -1, i32 1, i32 [[TMP17]], ptr @.{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZN2S12r1Ei_l121.region_id, ptr [[KERNEL_ARGS]])
// CHECK3-NEXT: [[TMP33:%.*]] = icmp ne i32 [[TMP32]], 0
// CHECK3-NEXT: br i1 [[TMP33]], label [[OMP_OFFLOAD_FAILED:%.*]], label [[OMP_OFFLOAD_CONT:%.*]]
// CHECK3: omp_offload.failed:
// CHECK3-NEXT: call void @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZN2S12r1Ei_l121(ptr [[THIS1]], i32 [[TMP3]], i32 [[TMP5]]) #[[ATTR3:[0-9]+]]
// CHECK3-NEXT: br label [[OMP_OFFLOAD_CONT]]
// CHECK3: omp_offload.cont:
// CHECK3-NEXT: [[A2:%.*]] = getelementptr inbounds [[STRUCT_S1]], ptr [[THIS1]], i32 0, i32 0
// CHECK3-NEXT: [[TMP34:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_BASEPTRS3]], i32 0, i32 0
// CHECK3-NEXT: store ptr [[THIS1]], ptr [[TMP34]], align 4
// CHECK3-NEXT: [[TMP35:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_PTRS4]], i32 0, i32 0
// CHECK3-NEXT: store ptr [[A2]], ptr [[TMP35]], align 4
// CHECK3-NEXT: [[TMP36:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_MAPPERS5]], i32 0, i32 0
// CHECK3-NEXT: store ptr null, ptr [[TMP36]], align 4
// CHECK3-NEXT: [[TMP37:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_BASEPTRS3]], i32 0, i32 0
// CHECK3-NEXT: [[TMP38:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_PTRS4]], i32 0, i32 0
// CHECK3-NEXT: [[TMP39:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 0
// CHECK3-NEXT: store i32 2, ptr [[TMP39]], align 4
// CHECK3-NEXT: [[TMP40:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 1
// CHECK3-NEXT: store i32 1, ptr [[TMP40]], align 4
// CHECK3-NEXT: [[TMP41:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 2
// CHECK3-NEXT: store ptr [[TMP37]], ptr [[TMP41]], align 4
// CHECK3-NEXT: [[TMP42:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 3
// CHECK3-NEXT: store ptr [[TMP38]], ptr [[TMP42]], align 4
// CHECK3-NEXT: [[TMP43:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 4
// CHECK3-NEXT: store ptr @.offload_sizes.1, ptr [[TMP43]], align 4
// CHECK3-NEXT: [[TMP44:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 5
// CHECK3-NEXT: store ptr @.offload_maptypes.2, ptr [[TMP44]], align 4
// CHECK3-NEXT: [[TMP45:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 6
// CHECK3-NEXT: store ptr null, ptr [[TMP45]], align 4
// CHECK3-NEXT: [[TMP46:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 7
// CHECK3-NEXT: store ptr null, ptr [[TMP46]], align 4
// CHECK3-NEXT: [[TMP47:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 8
// CHECK3-NEXT: store i64 0, ptr [[TMP47]], align 8
// CHECK3-NEXT: [[TMP48:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 9
// CHECK3-NEXT: store i64 0, ptr [[TMP48]], align 8
// CHECK3-NEXT: [[TMP49:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 10
// CHECK3-NEXT: store [3 x i32] [i32 1, i32 0, i32 0], ptr [[TMP49]], align 4
// CHECK3-NEXT: [[TMP50:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 11
// CHECK3-NEXT: store [3 x i32] [i32 1024, i32 0, i32 0], ptr [[TMP50]], align 4
// CHECK3-NEXT: [[TMP51:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 12
// CHECK3-NEXT: store i32 0, ptr [[TMP51]], align 4
// CHECK3-NEXT: [[TMP52:%.*]] = call i32 @__tgt_target_kernel(ptr @[[GLOB1]], i64 -1, i32 1, i32 1024, ptr @.{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZN2S12r1Ei_l126.region_id, ptr [[KERNEL_ARGS6]])
// CHECK3-NEXT: [[TMP53:%.*]] = icmp ne i32 [[TMP52]], 0
// CHECK3-NEXT: br i1 [[TMP53]], label [[OMP_OFFLOAD_FAILED7:%.*]], label [[OMP_OFFLOAD_CONT8:%.*]]
// CHECK3: omp_offload.failed7:
// CHECK3-NEXT: call void @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZN2S12r1Ei_l126(ptr [[THIS1]]) #[[ATTR3]]
// CHECK3-NEXT: br label [[OMP_OFFLOAD_CONT8]]
// CHECK3: omp_offload.cont8:
// CHECK3-NEXT: [[A9:%.*]] = getelementptr inbounds [[STRUCT_S1]], ptr [[THIS1]], i32 0, i32 0
// CHECK3-NEXT: [[TMP54:%.*]] = load double, ptr [[A9]], align 4
// CHECK3-NEXT: [[CONV:%.*]] = fptosi double [[TMP54]] to i32
// CHECK3-NEXT: ret i32 [[CONV]]
//
//
// CHECK3-LABEL: define {{[^@]+}}@_ZL7fstatici
// CHECK3-SAME: (i32 noundef [[N:%.*]]) #[[ATTR0]] {
// CHECK3-NEXT: entry:
// CHECK3-NEXT: [[N_ADDR:%.*]] = alloca i32, align 4
// CHECK3-NEXT: [[DOTCAPTURE_EXPR_:%.*]] = alloca i32, align 4
// CHECK3-NEXT: [[DOTCAPTURE_EXPR__CASTED:%.*]] = alloca i32, align 4
// CHECK3-NEXT: [[DOTOFFLOAD_BASEPTRS:%.*]] = alloca [1 x ptr], align 4
// CHECK3-NEXT: [[DOTOFFLOAD_PTRS:%.*]] = alloca [1 x ptr], align 4
// CHECK3-NEXT: [[DOTOFFLOAD_MAPPERS:%.*]] = alloca [1 x ptr], align 4
// CHECK3-NEXT: [[KERNEL_ARGS:%.*]] = alloca [[STRUCT___TGT_KERNEL_ARGUMENTS:%.*]], align 8
// CHECK3-NEXT: [[DOTCAPTURE_EXPR_1:%.*]] = alloca i32, align 4
// CHECK3-NEXT: [[DOTCAPTURE_EXPR__CASTED2:%.*]] = alloca i32, align 4
// CHECK3-NEXT: [[DOTOFFLOAD_BASEPTRS3:%.*]] = alloca [1 x ptr], align 4
// CHECK3-NEXT: [[DOTOFFLOAD_PTRS4:%.*]] = alloca [1 x ptr], align 4
// CHECK3-NEXT: [[DOTOFFLOAD_MAPPERS5:%.*]] = alloca [1 x ptr], align 4
// CHECK3-NEXT: [[KERNEL_ARGS6:%.*]] = alloca [[STRUCT___TGT_KERNEL_ARGUMENTS]], align 8
// CHECK3-NEXT: store i32 [[N]], ptr [[N_ADDR]], align 4
// CHECK3-NEXT: [[TMP0:%.*]] = load i32, ptr [[N_ADDR]], align 4
// CHECK3-NEXT: store i32 [[TMP0]], ptr [[DOTCAPTURE_EXPR_]], align 4
// CHECK3-NEXT: [[TMP1:%.*]] = load i32, ptr [[DOTCAPTURE_EXPR_]], align 4
// CHECK3-NEXT: store i32 [[TMP1]], ptr [[DOTCAPTURE_EXPR__CASTED]], align 4
// CHECK3-NEXT: [[TMP2:%.*]] = load i32, ptr [[DOTCAPTURE_EXPR__CASTED]], align 4
// CHECK3-NEXT: [[TMP3:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_BASEPTRS]], i32 0, i32 0
// CHECK3-NEXT: store i32 [[TMP2]], ptr [[TMP3]], align 4
// CHECK3-NEXT: [[TMP4:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_PTRS]], i32 0, i32 0
// CHECK3-NEXT: store i32 [[TMP2]], ptr [[TMP4]], align 4
// CHECK3-NEXT: [[TMP5:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_MAPPERS]], i32 0, i32 0
// CHECK3-NEXT: store ptr null, ptr [[TMP5]], align 4
// CHECK3-NEXT: [[TMP6:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_BASEPTRS]], i32 0, i32 0
// CHECK3-NEXT: [[TMP7:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_PTRS]], i32 0, i32 0
// CHECK3-NEXT: [[TMP8:%.*]] = load i32, ptr [[DOTCAPTURE_EXPR_]], align 4
// CHECK3-NEXT: [[TMP9:%.*]] = insertvalue [3 x i32] zeroinitializer, i32 [[TMP8]], 0
// CHECK3-NEXT: [[TMP10:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 0
// CHECK3-NEXT: store i32 2, ptr [[TMP10]], align 4
// CHECK3-NEXT: [[TMP11:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 1
// CHECK3-NEXT: store i32 1, ptr [[TMP11]], align 4
// CHECK3-NEXT: [[TMP12:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 2
// CHECK3-NEXT: store ptr [[TMP6]], ptr [[TMP12]], align 4
// CHECK3-NEXT: [[TMP13:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 3
// CHECK3-NEXT: store ptr [[TMP7]], ptr [[TMP13]], align 4
// CHECK3-NEXT: [[TMP14:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 4
// CHECK3-NEXT: store ptr @.offload_sizes.3, ptr [[TMP14]], align 4
// CHECK3-NEXT: [[TMP15:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 5
// CHECK3-NEXT: store ptr @.offload_maptypes.4, ptr [[TMP15]], align 4
// CHECK3-NEXT: [[TMP16:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 6
// CHECK3-NEXT: store ptr null, ptr [[TMP16]], align 4
// CHECK3-NEXT: [[TMP17:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 7
// CHECK3-NEXT: store ptr null, ptr [[TMP17]], align 4
// CHECK3-NEXT: [[TMP18:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 8
// CHECK3-NEXT: store i64 0, ptr [[TMP18]], align 8
// CHECK3-NEXT: [[TMP19:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 9
// CHECK3-NEXT: store i64 0, ptr [[TMP19]], align 8
// CHECK3-NEXT: [[TMP20:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 10
// CHECK3-NEXT: store [3 x i32] [i32 1, i32 0, i32 0], ptr [[TMP20]], align 4
// CHECK3-NEXT: [[TMP21:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 11
// CHECK3-NEXT: store [3 x i32] [[TMP9]], ptr [[TMP21]], align 4
// CHECK3-NEXT: [[TMP22:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 12
// CHECK3-NEXT: store i32 0, ptr [[TMP22]], align 4
// CHECK3-NEXT: [[TMP23:%.*]] = call i32 @__tgt_target_kernel(ptr @[[GLOB1]], i64 -1, i32 1, i32 [[TMP8]], ptr @.{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZL7fstatici_l104.region_id, ptr [[KERNEL_ARGS]])
// CHECK3-NEXT: [[TMP24:%.*]] = icmp ne i32 [[TMP23]], 0
// CHECK3-NEXT: br i1 [[TMP24]], label [[OMP_OFFLOAD_FAILED:%.*]], label [[OMP_OFFLOAD_CONT:%.*]]
// CHECK3: omp_offload.failed:
// CHECK3-NEXT: call void @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZL7fstatici_l104(i32 [[TMP2]]) #[[ATTR3]]
// CHECK3-NEXT: br label [[OMP_OFFLOAD_CONT]]
// CHECK3: omp_offload.cont:
// CHECK3-NEXT: [[TMP25:%.*]] = load i32, ptr [[N_ADDR]], align 4
// CHECK3-NEXT: [[ADD:%.*]] = add nsw i32 32, [[TMP25]]
// CHECK3-NEXT: store i32 [[ADD]], ptr [[DOTCAPTURE_EXPR_1]], align 4
// CHECK3-NEXT: [[TMP26:%.*]] = load i32, ptr [[DOTCAPTURE_EXPR_1]], align 4
// CHECK3-NEXT: store i32 [[TMP26]], ptr [[DOTCAPTURE_EXPR__CASTED2]], align 4
// CHECK3-NEXT: [[TMP27:%.*]] = load i32, ptr [[DOTCAPTURE_EXPR__CASTED2]], align 4
// CHECK3-NEXT: [[TMP28:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_BASEPTRS3]], i32 0, i32 0
// CHECK3-NEXT: store i32 [[TMP27]], ptr [[TMP28]], align 4
// CHECK3-NEXT: [[TMP29:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_PTRS4]], i32 0, i32 0
// CHECK3-NEXT: store i32 [[TMP27]], ptr [[TMP29]], align 4
// CHECK3-NEXT: [[TMP30:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_MAPPERS5]], i32 0, i32 0
// CHECK3-NEXT: store ptr null, ptr [[TMP30]], align 4
// CHECK3-NEXT: [[TMP31:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_BASEPTRS3]], i32 0, i32 0
// CHECK3-NEXT: [[TMP32:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_PTRS4]], i32 0, i32 0
// CHECK3-NEXT: [[TMP33:%.*]] = load i32, ptr [[DOTCAPTURE_EXPR_1]], align 4
// CHECK3-NEXT: [[TMP34:%.*]] = insertvalue [3 x i32] zeroinitializer, i32 [[TMP33]], 0
// CHECK3-NEXT: [[TMP35:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 0
// CHECK3-NEXT: store i32 2, ptr [[TMP35]], align 4
// CHECK3-NEXT: [[TMP36:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 1
// CHECK3-NEXT: store i32 1, ptr [[TMP36]], align 4
// CHECK3-NEXT: [[TMP37:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 2
// CHECK3-NEXT: store ptr [[TMP31]], ptr [[TMP37]], align 4
// CHECK3-NEXT: [[TMP38:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 3
// CHECK3-NEXT: store ptr [[TMP32]], ptr [[TMP38]], align 4
// CHECK3-NEXT: [[TMP39:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 4
// CHECK3-NEXT: store ptr @.offload_sizes.5, ptr [[TMP39]], align 4
// CHECK3-NEXT: [[TMP40:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 5
// CHECK3-NEXT: store ptr @.offload_maptypes.6, ptr [[TMP40]], align 4
// CHECK3-NEXT: [[TMP41:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 6
// CHECK3-NEXT: store ptr null, ptr [[TMP41]], align 4
// CHECK3-NEXT: [[TMP42:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 7
// CHECK3-NEXT: store ptr null, ptr [[TMP42]], align 4
// CHECK3-NEXT: [[TMP43:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 8
// CHECK3-NEXT: store i64 0, ptr [[TMP43]], align 8
// CHECK3-NEXT: [[TMP44:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 9
// CHECK3-NEXT: store i64 0, ptr [[TMP44]], align 8
// CHECK3-NEXT: [[TMP45:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 10
// CHECK3-NEXT: store [3 x i32] [i32 1, i32 0, i32 0], ptr [[TMP45]], align 4
// CHECK3-NEXT: [[TMP46:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 11
// CHECK3-NEXT: store [3 x i32] [[TMP34]], ptr [[TMP46]], align 4
// CHECK3-NEXT: [[TMP47:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS6]], i32 0, i32 12
// CHECK3-NEXT: store i32 0, ptr [[TMP47]], align 4
// CHECK3-NEXT: [[TMP48:%.*]] = call i32 @__tgt_target_kernel(ptr @[[GLOB1]], i64 -1, i32 1, i32 [[TMP33]], ptr @.{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZL7fstatici_l108.region_id, ptr [[KERNEL_ARGS6]])
// CHECK3-NEXT: [[TMP49:%.*]] = icmp ne i32 [[TMP48]], 0
// CHECK3-NEXT: br i1 [[TMP49]], label [[OMP_OFFLOAD_FAILED7:%.*]], label [[OMP_OFFLOAD_CONT8:%.*]]
// CHECK3: omp_offload.failed7:
// CHECK3-NEXT: call void @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZL7fstatici_l108(i32 [[TMP27]]) #[[ATTR3]]
// CHECK3-NEXT: br label [[OMP_OFFLOAD_CONT8]]
// CHECK3: omp_offload.cont8:
// CHECK3-NEXT: [[TMP50:%.*]] = load i32, ptr [[N_ADDR]], align 4
// CHECK3-NEXT: [[ADD9:%.*]] = add nsw i32 [[TMP50]], 1
// CHECK3-NEXT: ret i32 [[ADD9]]
//
//
// CHECK3-LABEL: define {{[^@]+}}@_Z9ftemplateIiET_i
// CHECK3-SAME: (i32 noundef [[N:%.*]]) #[[ATTR0]] comdat {
// CHECK3-NEXT: entry:
// CHECK3-NEXT: [[N_ADDR:%.*]] = alloca i32, align 4
// CHECK3-NEXT: [[A:%.*]] = alloca i32, align 4
// CHECK3-NEXT: [[KERNEL_ARGS:%.*]] = alloca [[STRUCT___TGT_KERNEL_ARGUMENTS:%.*]], align 8
// CHECK3-NEXT: [[B:%.*]] = alloca i16, align 2
// CHECK3-NEXT: [[DOTCAPTURE_EXPR_:%.*]] = alloca i16, align 2
// CHECK3-NEXT: [[A_CASTED:%.*]] = alloca i32, align 4
// CHECK3-NEXT: [[B_CASTED:%.*]] = alloca i32, align 4
// CHECK3-NEXT: [[DOTCAPTURE_EXPR__CASTED:%.*]] = alloca i32, align 4
// CHECK3-NEXT: [[DOTOFFLOAD_BASEPTRS:%.*]] = alloca [3 x ptr], align 4
// CHECK3-NEXT: [[DOTOFFLOAD_PTRS:%.*]] = alloca [3 x ptr], align 4
// CHECK3-NEXT: [[DOTOFFLOAD_MAPPERS:%.*]] = alloca [3 x ptr], align 4
// CHECK3-NEXT: [[KERNEL_ARGS1:%.*]] = alloca [[STRUCT___TGT_KERNEL_ARGUMENTS]], align 8
// CHECK3-NEXT: store i32 [[N]], ptr [[N_ADDR]], align 4
// CHECK3-NEXT: store i32 0, ptr [[A]], align 4
// CHECK3-NEXT: [[TMP0:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 0
// CHECK3-NEXT: store i32 2, ptr [[TMP0]], align 4
// CHECK3-NEXT: [[TMP1:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 1
// CHECK3-NEXT: store i32 0, ptr [[TMP1]], align 4
// CHECK3-NEXT: [[TMP2:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 2
// CHECK3-NEXT: store ptr null, ptr [[TMP2]], align 4
// CHECK3-NEXT: [[TMP3:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 3
// CHECK3-NEXT: store ptr null, ptr [[TMP3]], align 4
// CHECK3-NEXT: [[TMP4:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 4
// CHECK3-NEXT: store ptr null, ptr [[TMP4]], align 4
// CHECK3-NEXT: [[TMP5:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 5
// CHECK3-NEXT: store ptr null, ptr [[TMP5]], align 4
// CHECK3-NEXT: [[TMP6:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 6
// CHECK3-NEXT: store ptr null, ptr [[TMP6]], align 4
// CHECK3-NEXT: [[TMP7:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 7
// CHECK3-NEXT: store ptr null, ptr [[TMP7]], align 4
// CHECK3-NEXT: [[TMP8:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 8
// CHECK3-NEXT: store i64 0, ptr [[TMP8]], align 8
// CHECK3-NEXT: [[TMP9:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 9
// CHECK3-NEXT: store i64 0, ptr [[TMP9]], align 8
// CHECK3-NEXT: [[TMP10:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 10
// CHECK3-NEXT: store [3 x i32] [i32 1, i32 0, i32 0], ptr [[TMP10]], align 4
// CHECK3-NEXT: [[TMP11:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 11
// CHECK3-NEXT: store [3 x i32] [i32 20, i32 0, i32 0], ptr [[TMP11]], align 4
// CHECK3-NEXT: [[TMP12:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 12
// CHECK3-NEXT: store i32 0, ptr [[TMP12]], align 4
// CHECK3-NEXT: [[TMP13:%.*]] = call i32 @__tgt_target_kernel(ptr @[[GLOB1]], i64 -1, i32 1, i32 20, ptr @.{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIiET_i_l88.region_id, ptr [[KERNEL_ARGS]])
// CHECK3-NEXT: [[TMP14:%.*]] = icmp ne i32 [[TMP13]], 0
// CHECK3-NEXT: br i1 [[TMP14]], label [[OMP_OFFLOAD_FAILED:%.*]], label [[OMP_OFFLOAD_CONT:%.*]]
// CHECK3: omp_offload.failed:
// CHECK3-NEXT: call void @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIiET_i_l88() #[[ATTR3]]
// CHECK3-NEXT: br label [[OMP_OFFLOAD_CONT]]
// CHECK3: omp_offload.cont:
// CHECK3-NEXT: store i16 1, ptr [[B]], align 2
// CHECK3-NEXT: [[TMP15:%.*]] = load i16, ptr [[B]], align 2
// CHECK3-NEXT: store i16 [[TMP15]], ptr [[DOTCAPTURE_EXPR_]], align 2
// CHECK3-NEXT: [[TMP16:%.*]] = load i32, ptr [[A]], align 4
// CHECK3-NEXT: store i32 [[TMP16]], ptr [[A_CASTED]], align 4
// CHECK3-NEXT: [[TMP17:%.*]] = load i32, ptr [[A_CASTED]], align 4
// CHECK3-NEXT: [[TMP18:%.*]] = load i16, ptr [[B]], align 2
// CHECK3-NEXT: store i16 [[TMP18]], ptr [[B_CASTED]], align 2
// CHECK3-NEXT: [[TMP19:%.*]] = load i32, ptr [[B_CASTED]], align 4
// CHECK3-NEXT: [[TMP20:%.*]] = load i16, ptr [[DOTCAPTURE_EXPR_]], align 2
// CHECK3-NEXT: store i16 [[TMP20]], ptr [[DOTCAPTURE_EXPR__CASTED]], align 2
// CHECK3-NEXT: [[TMP21:%.*]] = load i32, ptr [[DOTCAPTURE_EXPR__CASTED]], align 4
// CHECK3-NEXT: [[TMP22:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_BASEPTRS]], i32 0, i32 0
// CHECK3-NEXT: store i32 [[TMP17]], ptr [[TMP22]], align 4
// CHECK3-NEXT: [[TMP23:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_PTRS]], i32 0, i32 0
// CHECK3-NEXT: store i32 [[TMP17]], ptr [[TMP23]], align 4
// CHECK3-NEXT: [[TMP24:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_MAPPERS]], i32 0, i32 0
// CHECK3-NEXT: store ptr null, ptr [[TMP24]], align 4
// CHECK3-NEXT: [[TMP25:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_BASEPTRS]], i32 0, i32 1
// CHECK3-NEXT: store i32 [[TMP19]], ptr [[TMP25]], align 4
// CHECK3-NEXT: [[TMP26:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_PTRS]], i32 0, i32 1
// CHECK3-NEXT: store i32 [[TMP19]], ptr [[TMP26]], align 4
// CHECK3-NEXT: [[TMP27:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_MAPPERS]], i32 0, i32 1
// CHECK3-NEXT: store ptr null, ptr [[TMP27]], align 4
// CHECK3-NEXT: [[TMP28:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_BASEPTRS]], i32 0, i32 2
// CHECK3-NEXT: store i32 [[TMP21]], ptr [[TMP28]], align 4
// CHECK3-NEXT: [[TMP29:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_PTRS]], i32 0, i32 2
// CHECK3-NEXT: store i32 [[TMP21]], ptr [[TMP29]], align 4
// CHECK3-NEXT: [[TMP30:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_MAPPERS]], i32 0, i32 2
// CHECK3-NEXT: store ptr null, ptr [[TMP30]], align 4
// CHECK3-NEXT: [[TMP31:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_BASEPTRS]], i32 0, i32 0
// CHECK3-NEXT: [[TMP32:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_PTRS]], i32 0, i32 0
// CHECK3-NEXT: [[TMP33:%.*]] = load i16, ptr [[DOTCAPTURE_EXPR_]], align 2
// CHECK3-NEXT: [[TMP34:%.*]] = zext i16 [[TMP33]] to i32
// CHECK3-NEXT: [[TMP35:%.*]] = insertvalue [3 x i32] zeroinitializer, i32 [[TMP34]], 0
// CHECK3-NEXT: [[TMP36:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS1]], i32 0, i32 0
// CHECK3-NEXT: store i32 2, ptr [[TMP36]], align 4
// CHECK3-NEXT: [[TMP37:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS1]], i32 0, i32 1
// CHECK3-NEXT: store i32 3, ptr [[TMP37]], align 4
// CHECK3-NEXT: [[TMP38:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS1]], i32 0, i32 2
// CHECK3-NEXT: store ptr [[TMP31]], ptr [[TMP38]], align 4
// CHECK3-NEXT: [[TMP39:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS1]], i32 0, i32 3
// CHECK3-NEXT: store ptr [[TMP32]], ptr [[TMP39]], align 4
// CHECK3-NEXT: [[TMP40:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS1]], i32 0, i32 4
// CHECK3-NEXT: store ptr @.offload_sizes.7, ptr [[TMP40]], align 4
// CHECK3-NEXT: [[TMP41:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS1]], i32 0, i32 5
// CHECK3-NEXT: store ptr @.offload_maptypes.8, ptr [[TMP41]], align 4
// CHECK3-NEXT: [[TMP42:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS1]], i32 0, i32 6
// CHECK3-NEXT: store ptr null, ptr [[TMP42]], align 4
// CHECK3-NEXT: [[TMP43:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS1]], i32 0, i32 7
// CHECK3-NEXT: store ptr null, ptr [[TMP43]], align 4
// CHECK3-NEXT: [[TMP44:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS1]], i32 0, i32 8
// CHECK3-NEXT: store i64 0, ptr [[TMP44]], align 8
// CHECK3-NEXT: [[TMP45:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS1]], i32 0, i32 9
// CHECK3-NEXT: store i64 0, ptr [[TMP45]], align 8
// CHECK3-NEXT: [[TMP46:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS1]], i32 0, i32 10
// CHECK3-NEXT: store [3 x i32] [i32 1, i32 0, i32 0], ptr [[TMP46]], align 4
// CHECK3-NEXT: [[TMP47:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS1]], i32 0, i32 11
// CHECK3-NEXT: store [3 x i32] [[TMP35]], ptr [[TMP47]], align 4
// CHECK3-NEXT: [[TMP48:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS1]], i32 0, i32 12
// CHECK3-NEXT: store i32 0, ptr [[TMP48]], align 4
// CHECK3-NEXT: [[TMP49:%.*]] = call i32 @__tgt_target_kernel(ptr @[[GLOB1]], i64 -1, i32 1, i32 [[TMP34]], ptr @.{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIiET_i_l93.region_id, ptr [[KERNEL_ARGS1]])
// CHECK3-NEXT: [[TMP50:%.*]] = icmp ne i32 [[TMP49]], 0
// CHECK3-NEXT: br i1 [[TMP50]], label [[OMP_OFFLOAD_FAILED2:%.*]], label [[OMP_OFFLOAD_CONT3:%.*]]
// CHECK3: omp_offload.failed2:
// CHECK3-NEXT: call void @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIiET_i_l93(i32 [[TMP17]], i32 [[TMP19]], i32 [[TMP21]]) #[[ATTR3]]
// CHECK3-NEXT: br label [[OMP_OFFLOAD_CONT3]]
// CHECK3: omp_offload.cont3:
// CHECK3-NEXT: [[TMP51:%.*]] = load i32, ptr [[A]], align 4
// CHECK3-NEXT: ret i32 [[TMP51]]
//
//
// CHECK3-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZN2S12r1Ei_l121
// CHECK3-SAME: (ptr noundef [[THIS:%.*]], i32 noundef [[B:%.*]], i32 noundef [[DOTCAPTURE_EXPR_:%.*]]) #[[ATTR1:[0-9]+]] {
// CHECK3-NEXT: entry:
// CHECK3-NEXT: [[THIS_ADDR:%.*]] = alloca ptr, align 4
// CHECK3-NEXT: [[B_ADDR:%.*]] = alloca i32, align 4
// CHECK3-NEXT: [[DOTCAPTURE_EXPR__ADDR:%.*]] = alloca i32, align 4
// CHECK3-NEXT: [[B_CASTED:%.*]] = alloca i32, align 4
// CHECK3-NEXT: [[TMP0:%.*]] = call i32 @__kmpc_global_thread_num(ptr @[[GLOB1]])
// CHECK3-NEXT: store ptr [[THIS]], ptr [[THIS_ADDR]], align 4
// CHECK3-NEXT: store i32 [[B]], ptr [[B_ADDR]], align 4
// CHECK3-NEXT: store i32 [[DOTCAPTURE_EXPR_]], ptr [[DOTCAPTURE_EXPR__ADDR]], align 4
// CHECK3-NEXT: [[TMP1:%.*]] = load ptr, ptr [[THIS_ADDR]], align 4
// CHECK3-NEXT: [[TMP2:%.*]] = load i32, ptr [[DOTCAPTURE_EXPR__ADDR]], align 4
// CHECK3-NEXT: call void @__kmpc_push_num_threads(ptr @[[GLOB1]], i32 [[TMP0]], i32 [[TMP2]])
// CHECK3-NEXT: [[TMP3:%.*]] = load i32, ptr [[B_ADDR]], align 4
// CHECK3-NEXT: store i32 [[TMP3]], ptr [[B_CASTED]], align 4
// CHECK3-NEXT: [[TMP4:%.*]] = load i32, ptr [[B_CASTED]], align 4
// CHECK3-NEXT: call void (ptr, i32, ptr, ...) @__kmpc_fork_call(ptr @[[GLOB1]], i32 2, ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZN2S12r1Ei_l121.omp_outlined, ptr [[TMP1]], i32 [[TMP4]])
// CHECK3-NEXT: ret void
//
//
// CHECK3-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZN2S12r1Ei_l121.omp_outlined
// CHECK3-SAME: (ptr noalias noundef [[DOTGLOBAL_TID_:%.*]], ptr noalias noundef [[DOTBOUND_TID_:%.*]], ptr noundef [[THIS:%.*]], i32 noundef [[B:%.*]]) #[[ATTR2:[0-9]+]] {
// CHECK3-NEXT: entry:
// CHECK3-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca ptr, align 4
// CHECK3-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca ptr, align 4
// CHECK3-NEXT: [[THIS_ADDR:%.*]] = alloca ptr, align 4
// CHECK3-NEXT: [[B_ADDR:%.*]] = alloca i32, align 4
// CHECK3-NEXT: store ptr [[DOTGLOBAL_TID_]], ptr [[DOTGLOBAL_TID__ADDR]], align 4
// CHECK3-NEXT: store ptr [[DOTBOUND_TID_]], ptr [[DOTBOUND_TID__ADDR]], align 4
// CHECK3-NEXT: store ptr [[THIS]], ptr [[THIS_ADDR]], align 4
// CHECK3-NEXT: store i32 [[B]], ptr [[B_ADDR]], align 4
// CHECK3-NEXT: [[TMP0:%.*]] = load ptr, ptr [[THIS_ADDR]], align 4
// CHECK3-NEXT: [[TMP1:%.*]] = load i32, ptr [[B_ADDR]], align 4
// CHECK3-NEXT: [[CONV:%.*]] = sitofp i32 [[TMP1]] to double
// CHECK3-NEXT: [[ADD:%.*]] = fadd double [[CONV]], 1.500000e+00
// CHECK3-NEXT: [[A:%.*]] = getelementptr inbounds [[STRUCT_S1:%.*]], ptr [[TMP0]], i32 0, i32 0
// CHECK3-NEXT: store double [[ADD]], ptr [[A]], align 4
// CHECK3-NEXT: ret void
//
//
// CHECK3-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZN2S12r1Ei_l126
// CHECK3-SAME: (ptr noundef [[THIS:%.*]]) #[[ATTR1]] {
// CHECK3-NEXT: entry:
// CHECK3-NEXT: [[THIS_ADDR:%.*]] = alloca ptr, align 4
// CHECK3-NEXT: [[TMP0:%.*]] = call i32 @__kmpc_global_thread_num(ptr @[[GLOB1]])
// CHECK3-NEXT: store ptr [[THIS]], ptr [[THIS_ADDR]], align 4
// CHECK3-NEXT: [[TMP1:%.*]] = load ptr, ptr [[THIS_ADDR]], align 4
// CHECK3-NEXT: call void @__kmpc_push_num_threads(ptr @[[GLOB1]], i32 [[TMP0]], i32 1024)
// CHECK3-NEXT: call void (ptr, i32, ptr, ...) @__kmpc_fork_call(ptr @[[GLOB1]], i32 1, ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZN2S12r1Ei_l126.omp_outlined, ptr [[TMP1]])
// CHECK3-NEXT: ret void
//
//
// CHECK3-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZN2S12r1Ei_l126.omp_outlined
// CHECK3-SAME: (ptr noalias noundef [[DOTGLOBAL_TID_:%.*]], ptr noalias noundef [[DOTBOUND_TID_:%.*]], ptr noundef [[THIS:%.*]]) #[[ATTR2]] {
// CHECK3-NEXT: entry:
// CHECK3-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca ptr, align 4
// CHECK3-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca ptr, align 4
// CHECK3-NEXT: [[THIS_ADDR:%.*]] = alloca ptr, align 4
// CHECK3-NEXT: store ptr [[DOTGLOBAL_TID_]], ptr [[DOTGLOBAL_TID__ADDR]], align 4
// CHECK3-NEXT: store ptr [[DOTBOUND_TID_]], ptr [[DOTBOUND_TID__ADDR]], align 4
// CHECK3-NEXT: store ptr [[THIS]], ptr [[THIS_ADDR]], align 4
// CHECK3-NEXT: [[TMP0:%.*]] = load ptr, ptr [[THIS_ADDR]], align 4
// CHECK3-NEXT: [[A:%.*]] = getelementptr inbounds [[STRUCT_S1:%.*]], ptr [[TMP0]], i32 0, i32 0
// CHECK3-NEXT: store double 2.500000e+00, ptr [[A]], align 4
// CHECK3-NEXT: ret void
//
//
// CHECK3-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZL7fstatici_l104
// CHECK3-SAME: (i32 noundef [[DOTCAPTURE_EXPR_:%.*]]) #[[ATTR1]] {
// CHECK3-NEXT: entry:
// CHECK3-NEXT: [[DOTCAPTURE_EXPR__ADDR:%.*]] = alloca i32, align 4
// CHECK3-NEXT: [[TMP0:%.*]] = call i32 @__kmpc_global_thread_num(ptr @[[GLOB1]])
// CHECK3-NEXT: store i32 [[DOTCAPTURE_EXPR_]], ptr [[DOTCAPTURE_EXPR__ADDR]], align 4
// CHECK3-NEXT: [[TMP1:%.*]] = load i32, ptr [[DOTCAPTURE_EXPR__ADDR]], align 4
// CHECK3-NEXT: call void @__kmpc_push_num_threads(ptr @[[GLOB1]], i32 [[TMP0]], i32 [[TMP1]])
// CHECK3-NEXT: call void (ptr, i32, ptr, ...) @__kmpc_fork_call(ptr @[[GLOB1]], i32 0, ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZL7fstatici_l104.omp_outlined)
// CHECK3-NEXT: ret void
//
//
// CHECK3-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZL7fstatici_l104.omp_outlined
// CHECK3-SAME: (ptr noalias noundef [[DOTGLOBAL_TID_:%.*]], ptr noalias noundef [[DOTBOUND_TID_:%.*]]) #[[ATTR2]] {
// CHECK3-NEXT: entry:
// CHECK3-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca ptr, align 4
// CHECK3-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca ptr, align 4
// CHECK3-NEXT: store ptr [[DOTGLOBAL_TID_]], ptr [[DOTGLOBAL_TID__ADDR]], align 4
// CHECK3-NEXT: store ptr [[DOTBOUND_TID_]], ptr [[DOTBOUND_TID__ADDR]], align 4
// CHECK3-NEXT: ret void
//
//
// CHECK3-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZL7fstatici_l108
// CHECK3-SAME: (i32 noundef [[DOTCAPTURE_EXPR_:%.*]]) #[[ATTR1]] {
// CHECK3-NEXT: entry:
// CHECK3-NEXT: [[DOTCAPTURE_EXPR__ADDR:%.*]] = alloca i32, align 4
// CHECK3-NEXT: [[TMP0:%.*]] = call i32 @__kmpc_global_thread_num(ptr @[[GLOB1]])
// CHECK3-NEXT: store i32 [[DOTCAPTURE_EXPR_]], ptr [[DOTCAPTURE_EXPR__ADDR]], align 4
// CHECK3-NEXT: [[TMP1:%.*]] = load i32, ptr [[DOTCAPTURE_EXPR__ADDR]], align 4
// CHECK3-NEXT: call void @__kmpc_push_num_threads(ptr @[[GLOB1]], i32 [[TMP0]], i32 [[TMP1]])
// CHECK3-NEXT: call void (ptr, i32, ptr, ...) @__kmpc_fork_call(ptr @[[GLOB1]], i32 0, ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZL7fstatici_l108.omp_outlined)
// CHECK3-NEXT: ret void
//
//
// CHECK3-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZL7fstatici_l108.omp_outlined
// CHECK3-SAME: (ptr noalias noundef [[DOTGLOBAL_TID_:%.*]], ptr noalias noundef [[DOTBOUND_TID_:%.*]]) #[[ATTR2]] {
// CHECK3-NEXT: entry:
// CHECK3-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca ptr, align 4
// CHECK3-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca ptr, align 4
// CHECK3-NEXT: store ptr [[DOTGLOBAL_TID_]], ptr [[DOTGLOBAL_TID__ADDR]], align 4
// CHECK3-NEXT: store ptr [[DOTBOUND_TID_]], ptr [[DOTBOUND_TID__ADDR]], align 4
// CHECK3-NEXT: ret void
//
//
// CHECK3-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIiET_i_l88
// CHECK3-SAME: () #[[ATTR1]] {
// CHECK3-NEXT: entry:
// CHECK3-NEXT: [[TMP0:%.*]] = call i32 @__kmpc_global_thread_num(ptr @[[GLOB1]])
// CHECK3-NEXT: call void @__kmpc_push_num_threads(ptr @[[GLOB1]], i32 [[TMP0]], i32 20)
// CHECK3-NEXT: call void (ptr, i32, ptr, ...) @__kmpc_fork_call(ptr @[[GLOB1]], i32 0, ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIiET_i_l88.omp_outlined)
// CHECK3-NEXT: ret void
//
//
// CHECK3-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIiET_i_l88.omp_outlined
// CHECK3-SAME: (ptr noalias noundef [[DOTGLOBAL_TID_:%.*]], ptr noalias noundef [[DOTBOUND_TID_:%.*]]) #[[ATTR2]] {
// CHECK3-NEXT: entry:
// CHECK3-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca ptr, align 4
// CHECK3-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca ptr, align 4
// CHECK3-NEXT: store ptr [[DOTGLOBAL_TID_]], ptr [[DOTGLOBAL_TID__ADDR]], align 4
// CHECK3-NEXT: store ptr [[DOTBOUND_TID_]], ptr [[DOTBOUND_TID__ADDR]], align 4
// CHECK3-NEXT: ret void
//
//
// CHECK3-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIiET_i_l93
// CHECK3-SAME: (i32 noundef [[A:%.*]], i32 noundef [[B:%.*]], i32 noundef [[DOTCAPTURE_EXPR_:%.*]]) #[[ATTR1]] {
// CHECK3-NEXT: entry:
// CHECK3-NEXT: [[A_ADDR:%.*]] = alloca i32, align 4
// CHECK3-NEXT: [[B_ADDR:%.*]] = alloca i32, align 4
// CHECK3-NEXT: [[DOTCAPTURE_EXPR__ADDR:%.*]] = alloca i32, align 4
// CHECK3-NEXT: [[A_CASTED:%.*]] = alloca i32, align 4
// CHECK3-NEXT: [[B_CASTED:%.*]] = alloca i32, align 4
// CHECK3-NEXT: [[TMP0:%.*]] = call i32 @__kmpc_global_thread_num(ptr @[[GLOB1]])
// CHECK3-NEXT: store i32 [[A]], ptr [[A_ADDR]], align 4
// CHECK3-NEXT: store i32 [[B]], ptr [[B_ADDR]], align 4
// CHECK3-NEXT: store i32 [[DOTCAPTURE_EXPR_]], ptr [[DOTCAPTURE_EXPR__ADDR]], align 4
// CHECK3-NEXT: [[TMP1:%.*]] = load i16, ptr [[DOTCAPTURE_EXPR__ADDR]], align 2
// CHECK3-NEXT: [[TMP2:%.*]] = sext i16 [[TMP1]] to i32
// CHECK3-NEXT: call void @__kmpc_push_num_threads(ptr @[[GLOB1]], i32 [[TMP0]], i32 [[TMP2]])
// CHECK3-NEXT: [[TMP3:%.*]] = load i32, ptr [[A_ADDR]], align 4
// CHECK3-NEXT: store i32 [[TMP3]], ptr [[A_CASTED]], align 4
// CHECK3-NEXT: [[TMP4:%.*]] = load i32, ptr [[A_CASTED]], align 4
// CHECK3-NEXT: [[TMP5:%.*]] = load i16, ptr [[B_ADDR]], align 2
// CHECK3-NEXT: store i16 [[TMP5]], ptr [[B_CASTED]], align 2
// CHECK3-NEXT: [[TMP6:%.*]] = load i32, ptr [[B_CASTED]], align 4
// CHECK3-NEXT: call void (ptr, i32, ptr, ...) @__kmpc_fork_call(ptr @[[GLOB1]], i32 2, ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIiET_i_l93.omp_outlined, i32 [[TMP4]], i32 [[TMP6]])
// CHECK3-NEXT: ret void
//
//
// CHECK3-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIiET_i_l93.omp_outlined
// CHECK3-SAME: (ptr noalias noundef [[DOTGLOBAL_TID_:%.*]], ptr noalias noundef [[DOTBOUND_TID_:%.*]], i32 noundef [[A:%.*]], i32 noundef [[B:%.*]]) #[[ATTR2]] {
// CHECK3-NEXT: entry:
// CHECK3-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca ptr, align 4
// CHECK3-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca ptr, align 4
// CHECK3-NEXT: [[A_ADDR:%.*]] = alloca i32, align 4
// CHECK3-NEXT: [[B_ADDR:%.*]] = alloca i32, align 4
// CHECK3-NEXT: store ptr [[DOTGLOBAL_TID_]], ptr [[DOTGLOBAL_TID__ADDR]], align 4
// CHECK3-NEXT: store ptr [[DOTBOUND_TID_]], ptr [[DOTBOUND_TID__ADDR]], align 4
// CHECK3-NEXT: store i32 [[A]], ptr [[A_ADDR]], align 4
// CHECK3-NEXT: store i32 [[B]], ptr [[B_ADDR]], align 4
// CHECK3-NEXT: [[TMP0:%.*]] = load i16, ptr [[B_ADDR]], align 2
// CHECK3-NEXT: [[CONV:%.*]] = sext i16 [[TMP0]] to i32
// CHECK3-NEXT: [[TMP1:%.*]] = load i32, ptr [[A_ADDR]], align 4
// CHECK3-NEXT: [[ADD:%.*]] = add nsw i32 [[TMP1]], [[CONV]]
// CHECK3-NEXT: store i32 [[ADD]], ptr [[A_ADDR]], align 4
// CHECK3-NEXT: ret void
//
//
// CHECK3-LABEL: define {{[^@]+}}@.omp_offloading.requires_reg
// CHECK3-SAME: () #[[ATTR4:[0-9]+]] {
// CHECK3-NEXT: entry:
// CHECK3-NEXT: call void @__tgt_register_requires(i64 1)
// CHECK3-NEXT: ret void
//
//
// CHECK9-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZL7fstatici_l104
// CHECK9-SAME: (i64 noundef [[DOTCAPTURE_EXPR_:%.*]]) #[[ATTR0:[0-9]+]] {
// CHECK9-NEXT: entry:
// CHECK9-NEXT: [[DOTCAPTURE_EXPR__ADDR:%.*]] = alloca i64, align 8
// CHECK9-NEXT: [[TMP0:%.*]] = call i32 @__kmpc_global_thread_num(ptr @[[GLOB1:[0-9]+]])
// CHECK9-NEXT: store i64 [[DOTCAPTURE_EXPR_]], ptr [[DOTCAPTURE_EXPR__ADDR]], align 8
// CHECK9-NEXT: [[TMP1:%.*]] = load i32, ptr [[DOTCAPTURE_EXPR__ADDR]], align 4
// CHECK9-NEXT: call void @__kmpc_push_num_threads(ptr @[[GLOB1]], i32 [[TMP0]], i32 [[TMP1]])
// CHECK9-NEXT: call void (ptr, i32, ptr, ...) @__kmpc_fork_call(ptr @[[GLOB1]], i32 0, ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZL7fstatici_l104.omp_outlined)
// CHECK9-NEXT: ret void
//
//
// CHECK9-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZL7fstatici_l104.omp_outlined
// CHECK9-SAME: (ptr noalias noundef [[DOTGLOBAL_TID_:%.*]], ptr noalias noundef [[DOTBOUND_TID_:%.*]]) #[[ATTR1:[0-9]+]] {
// CHECK9-NEXT: entry:
// CHECK9-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca ptr, align 8
// CHECK9-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca ptr, align 8
// CHECK9-NEXT: store ptr [[DOTGLOBAL_TID_]], ptr [[DOTGLOBAL_TID__ADDR]], align 8
// CHECK9-NEXT: store ptr [[DOTBOUND_TID_]], ptr [[DOTBOUND_TID__ADDR]], align 8
// CHECK9-NEXT: ret void
//
//
// CHECK9-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZL7fstatici_l108
// CHECK9-SAME: (i64 noundef [[DOTCAPTURE_EXPR_:%.*]]) #[[ATTR0]] {
// CHECK9-NEXT: entry:
// CHECK9-NEXT: [[DOTCAPTURE_EXPR__ADDR:%.*]] = alloca i64, align 8
// CHECK9-NEXT: [[TMP0:%.*]] = call i32 @__kmpc_global_thread_num(ptr @[[GLOB1]])
// CHECK9-NEXT: store i64 [[DOTCAPTURE_EXPR_]], ptr [[DOTCAPTURE_EXPR__ADDR]], align 8
// CHECK9-NEXT: [[TMP1:%.*]] = load i32, ptr [[DOTCAPTURE_EXPR__ADDR]], align 4
// CHECK9-NEXT: call void @__kmpc_push_num_threads(ptr @[[GLOB1]], i32 [[TMP0]], i32 [[TMP1]])
// CHECK9-NEXT: call void (ptr, i32, ptr, ...) @__kmpc_fork_call(ptr @[[GLOB1]], i32 0, ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZL7fstatici_l108.omp_outlined)
// CHECK9-NEXT: ret void
//
//
// CHECK9-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZL7fstatici_l108.omp_outlined
// CHECK9-SAME: (ptr noalias noundef [[DOTGLOBAL_TID_:%.*]], ptr noalias noundef [[DOTBOUND_TID_:%.*]]) #[[ATTR1]] {
// CHECK9-NEXT: entry:
// CHECK9-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca ptr, align 8
// CHECK9-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca ptr, align 8
// CHECK9-NEXT: store ptr [[DOTGLOBAL_TID_]], ptr [[DOTGLOBAL_TID__ADDR]], align 8
// CHECK9-NEXT: store ptr [[DOTBOUND_TID_]], ptr [[DOTBOUND_TID__ADDR]], align 8
// CHECK9-NEXT: ret void
//
//
// CHECK9-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZN2S12r1Ei_l121
// CHECK9-SAME: (ptr noundef [[THIS:%.*]], i64 noundef [[B:%.*]], i64 noundef [[DOTCAPTURE_EXPR_:%.*]]) #[[ATTR0]] {
// CHECK9-NEXT: entry:
// CHECK9-NEXT: [[THIS_ADDR:%.*]] = alloca ptr, align 8
// CHECK9-NEXT: [[B_ADDR:%.*]] = alloca i64, align 8
// CHECK9-NEXT: [[DOTCAPTURE_EXPR__ADDR:%.*]] = alloca i64, align 8
// CHECK9-NEXT: [[B_CASTED:%.*]] = alloca i64, align 8
// CHECK9-NEXT: [[TMP0:%.*]] = call i32 @__kmpc_global_thread_num(ptr @[[GLOB1]])
// CHECK9-NEXT: store ptr [[THIS]], ptr [[THIS_ADDR]], align 8
// CHECK9-NEXT: store i64 [[B]], ptr [[B_ADDR]], align 8
// CHECK9-NEXT: store i64 [[DOTCAPTURE_EXPR_]], ptr [[DOTCAPTURE_EXPR__ADDR]], align 8
// CHECK9-NEXT: [[TMP1:%.*]] = load ptr, ptr [[THIS_ADDR]], align 8
// CHECK9-NEXT: [[TMP2:%.*]] = load i32, ptr [[DOTCAPTURE_EXPR__ADDR]], align 4
// CHECK9-NEXT: call void @__kmpc_push_num_threads(ptr @[[GLOB1]], i32 [[TMP0]], i32 [[TMP2]])
// CHECK9-NEXT: [[TMP3:%.*]] = load i32, ptr [[B_ADDR]], align 4
// CHECK9-NEXT: store i32 [[TMP3]], ptr [[B_CASTED]], align 4
// CHECK9-NEXT: [[TMP4:%.*]] = load i64, ptr [[B_CASTED]], align 8
// CHECK9-NEXT: call void (ptr, i32, ptr, ...) @__kmpc_fork_call(ptr @[[GLOB1]], i32 2, ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZN2S12r1Ei_l121.omp_outlined, ptr [[TMP1]], i64 [[TMP4]])
// CHECK9-NEXT: ret void
//
//
// CHECK9-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZN2S12r1Ei_l121.omp_outlined
// CHECK9-SAME: (ptr noalias noundef [[DOTGLOBAL_TID_:%.*]], ptr noalias noundef [[DOTBOUND_TID_:%.*]], ptr noundef [[THIS:%.*]], i64 noundef [[B:%.*]]) #[[ATTR1]] {
// CHECK9-NEXT: entry:
// CHECK9-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca ptr, align 8
// CHECK9-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca ptr, align 8
// CHECK9-NEXT: [[THIS_ADDR:%.*]] = alloca ptr, align 8
// CHECK9-NEXT: [[B_ADDR:%.*]] = alloca i64, align 8
// CHECK9-NEXT: store ptr [[DOTGLOBAL_TID_]], ptr [[DOTGLOBAL_TID__ADDR]], align 8
// CHECK9-NEXT: store ptr [[DOTBOUND_TID_]], ptr [[DOTBOUND_TID__ADDR]], align 8
// CHECK9-NEXT: store ptr [[THIS]], ptr [[THIS_ADDR]], align 8
// CHECK9-NEXT: store i64 [[B]], ptr [[B_ADDR]], align 8
// CHECK9-NEXT: [[TMP0:%.*]] = load ptr, ptr [[THIS_ADDR]], align 8
// CHECK9-NEXT: [[TMP1:%.*]] = load i32, ptr [[B_ADDR]], align 4
// CHECK9-NEXT: [[CONV:%.*]] = sitofp i32 [[TMP1]] to double
// CHECK9-NEXT: [[ADD:%.*]] = fadd double [[CONV]], 1.500000e+00
// CHECK9-NEXT: [[A:%.*]] = getelementptr inbounds [[STRUCT_S1:%.*]], ptr [[TMP0]], i32 0, i32 0
// CHECK9-NEXT: store double [[ADD]], ptr [[A]], align 8
// CHECK9-NEXT: ret void
//
//
// CHECK9-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZN2S12r1Ei_l126
// CHECK9-SAME: (ptr noundef [[THIS:%.*]]) #[[ATTR0]] {
// CHECK9-NEXT: entry:
// CHECK9-NEXT: [[THIS_ADDR:%.*]] = alloca ptr, align 8
// CHECK9-NEXT: [[TMP0:%.*]] = call i32 @__kmpc_global_thread_num(ptr @[[GLOB1]])
// CHECK9-NEXT: store ptr [[THIS]], ptr [[THIS_ADDR]], align 8
// CHECK9-NEXT: [[TMP1:%.*]] = load ptr, ptr [[THIS_ADDR]], align 8
// CHECK9-NEXT: call void @__kmpc_push_num_threads(ptr @[[GLOB1]], i32 [[TMP0]], i32 1024)
// CHECK9-NEXT: call void (ptr, i32, ptr, ...) @__kmpc_fork_call(ptr @[[GLOB1]], i32 1, ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZN2S12r1Ei_l126.omp_outlined, ptr [[TMP1]])
// CHECK9-NEXT: ret void
//
//
// CHECK9-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZN2S12r1Ei_l126.omp_outlined
// CHECK9-SAME: (ptr noalias noundef [[DOTGLOBAL_TID_:%.*]], ptr noalias noundef [[DOTBOUND_TID_:%.*]], ptr noundef [[THIS:%.*]]) #[[ATTR1]] {
// CHECK9-NEXT: entry:
// CHECK9-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca ptr, align 8
// CHECK9-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca ptr, align 8
// CHECK9-NEXT: [[THIS_ADDR:%.*]] = alloca ptr, align 8
// CHECK9-NEXT: store ptr [[DOTGLOBAL_TID_]], ptr [[DOTGLOBAL_TID__ADDR]], align 8
// CHECK9-NEXT: store ptr [[DOTBOUND_TID_]], ptr [[DOTBOUND_TID__ADDR]], align 8
// CHECK9-NEXT: store ptr [[THIS]], ptr [[THIS_ADDR]], align 8
// CHECK9-NEXT: [[TMP0:%.*]] = load ptr, ptr [[THIS_ADDR]], align 8
// CHECK9-NEXT: [[A:%.*]] = getelementptr inbounds [[STRUCT_S1:%.*]], ptr [[TMP0]], i32 0, i32 0
// CHECK9-NEXT: store double 2.500000e+00, ptr [[A]], align 8
// CHECK9-NEXT: ret void
//
//
// CHECK9-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIiET_i_l88
// CHECK9-SAME: () #[[ATTR0]] {
// CHECK9-NEXT: entry:
// CHECK9-NEXT: [[TMP0:%.*]] = call i32 @__kmpc_global_thread_num(ptr @[[GLOB1]])
// CHECK9-NEXT: call void @__kmpc_push_num_threads(ptr @[[GLOB1]], i32 [[TMP0]], i32 20)
// CHECK9-NEXT: call void (ptr, i32, ptr, ...) @__kmpc_fork_call(ptr @[[GLOB1]], i32 0, ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIiET_i_l88.omp_outlined)
// CHECK9-NEXT: ret void
//
//
// CHECK9-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIiET_i_l88.omp_outlined
// CHECK9-SAME: (ptr noalias noundef [[DOTGLOBAL_TID_:%.*]], ptr noalias noundef [[DOTBOUND_TID_:%.*]]) #[[ATTR1]] {
// CHECK9-NEXT: entry:
// CHECK9-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca ptr, align 8
// CHECK9-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca ptr, align 8
// CHECK9-NEXT: store ptr [[DOTGLOBAL_TID_]], ptr [[DOTGLOBAL_TID__ADDR]], align 8
// CHECK9-NEXT: store ptr [[DOTBOUND_TID_]], ptr [[DOTBOUND_TID__ADDR]], align 8
// CHECK9-NEXT: ret void
//
//
// CHECK9-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIiET_i_l93
// CHECK9-SAME: (i64 noundef [[A:%.*]], i64 noundef [[B:%.*]], i64 noundef [[DOTCAPTURE_EXPR_:%.*]]) #[[ATTR0]] {
// CHECK9-NEXT: entry:
// CHECK9-NEXT: [[A_ADDR:%.*]] = alloca i64, align 8
// CHECK9-NEXT: [[B_ADDR:%.*]] = alloca i64, align 8
// CHECK9-NEXT: [[DOTCAPTURE_EXPR__ADDR:%.*]] = alloca i64, align 8
// CHECK9-NEXT: [[A_CASTED:%.*]] = alloca i64, align 8
// CHECK9-NEXT: [[B_CASTED:%.*]] = alloca i64, align 8
// CHECK9-NEXT: [[TMP0:%.*]] = call i32 @__kmpc_global_thread_num(ptr @[[GLOB1]])
// CHECK9-NEXT: store i64 [[A]], ptr [[A_ADDR]], align 8
// CHECK9-NEXT: store i64 [[B]], ptr [[B_ADDR]], align 8
// CHECK9-NEXT: store i64 [[DOTCAPTURE_EXPR_]], ptr [[DOTCAPTURE_EXPR__ADDR]], align 8
// CHECK9-NEXT: [[TMP1:%.*]] = load i16, ptr [[DOTCAPTURE_EXPR__ADDR]], align 2
// CHECK9-NEXT: [[TMP2:%.*]] = sext i16 [[TMP1]] to i32
// CHECK9-NEXT: call void @__kmpc_push_num_threads(ptr @[[GLOB1]], i32 [[TMP0]], i32 [[TMP2]])
// CHECK9-NEXT: [[TMP3:%.*]] = load i32, ptr [[A_ADDR]], align 4
// CHECK9-NEXT: store i32 [[TMP3]], ptr [[A_CASTED]], align 4
// CHECK9-NEXT: [[TMP4:%.*]] = load i64, ptr [[A_CASTED]], align 8
// CHECK9-NEXT: [[TMP5:%.*]] = load i16, ptr [[B_ADDR]], align 2
// CHECK9-NEXT: store i16 [[TMP5]], ptr [[B_CASTED]], align 2
// CHECK9-NEXT: [[TMP6:%.*]] = load i64, ptr [[B_CASTED]], align 8
// CHECK9-NEXT: call void (ptr, i32, ptr, ...) @__kmpc_fork_call(ptr @[[GLOB1]], i32 2, ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIiET_i_l93.omp_outlined, i64 [[TMP4]], i64 [[TMP6]])
// CHECK9-NEXT: ret void
//
//
// CHECK9-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIiET_i_l93.omp_outlined
// CHECK9-SAME: (ptr noalias noundef [[DOTGLOBAL_TID_:%.*]], ptr noalias noundef [[DOTBOUND_TID_:%.*]], i64 noundef [[A:%.*]], i64 noundef [[B:%.*]]) #[[ATTR1]] {
// CHECK9-NEXT: entry:
// CHECK9-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca ptr, align 8
// CHECK9-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca ptr, align 8
// CHECK9-NEXT: [[A_ADDR:%.*]] = alloca i64, align 8
// CHECK9-NEXT: [[B_ADDR:%.*]] = alloca i64, align 8
// CHECK9-NEXT: store ptr [[DOTGLOBAL_TID_]], ptr [[DOTGLOBAL_TID__ADDR]], align 8
// CHECK9-NEXT: store ptr [[DOTBOUND_TID_]], ptr [[DOTBOUND_TID__ADDR]], align 8
// CHECK9-NEXT: store i64 [[A]], ptr [[A_ADDR]], align 8
// CHECK9-NEXT: store i64 [[B]], ptr [[B_ADDR]], align 8
// CHECK9-NEXT: [[TMP0:%.*]] = load i16, ptr [[B_ADDR]], align 2
// CHECK9-NEXT: [[CONV:%.*]] = sext i16 [[TMP0]] to i32
// CHECK9-NEXT: [[TMP1:%.*]] = load i32, ptr [[A_ADDR]], align 4
// CHECK9-NEXT: [[ADD:%.*]] = add nsw i32 [[TMP1]], [[CONV]]
// CHECK9-NEXT: store i32 [[ADD]], ptr [[A_ADDR]], align 4
// CHECK9-NEXT: ret void
//
//
// CHECK11-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZL7fstatici_l104
// CHECK11-SAME: (i32 noundef [[DOTCAPTURE_EXPR_:%.*]]) #[[ATTR0:[0-9]+]] {
// CHECK11-NEXT: entry:
// CHECK11-NEXT: [[DOTCAPTURE_EXPR__ADDR:%.*]] = alloca i32, align 4
// CHECK11-NEXT: [[TMP0:%.*]] = call i32 @__kmpc_global_thread_num(ptr @[[GLOB1:[0-9]+]])
// CHECK11-NEXT: store i32 [[DOTCAPTURE_EXPR_]], ptr [[DOTCAPTURE_EXPR__ADDR]], align 4
// CHECK11-NEXT: [[TMP1:%.*]] = load i32, ptr [[DOTCAPTURE_EXPR__ADDR]], align 4
// CHECK11-NEXT: call void @__kmpc_push_num_threads(ptr @[[GLOB1]], i32 [[TMP0]], i32 [[TMP1]])
// CHECK11-NEXT: call void (ptr, i32, ptr, ...) @__kmpc_fork_call(ptr @[[GLOB1]], i32 0, ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZL7fstatici_l104.omp_outlined)
// CHECK11-NEXT: ret void
//
//
// CHECK11-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZL7fstatici_l104.omp_outlined
// CHECK11-SAME: (ptr noalias noundef [[DOTGLOBAL_TID_:%.*]], ptr noalias noundef [[DOTBOUND_TID_:%.*]]) #[[ATTR1:[0-9]+]] {
// CHECK11-NEXT: entry:
// CHECK11-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca ptr, align 4
// CHECK11-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca ptr, align 4
// CHECK11-NEXT: store ptr [[DOTGLOBAL_TID_]], ptr [[DOTGLOBAL_TID__ADDR]], align 4
// CHECK11-NEXT: store ptr [[DOTBOUND_TID_]], ptr [[DOTBOUND_TID__ADDR]], align 4
// CHECK11-NEXT: ret void
//
//
// CHECK11-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZL7fstatici_l108
// CHECK11-SAME: (i32 noundef [[DOTCAPTURE_EXPR_:%.*]]) #[[ATTR0]] {
// CHECK11-NEXT: entry:
// CHECK11-NEXT: [[DOTCAPTURE_EXPR__ADDR:%.*]] = alloca i32, align 4
// CHECK11-NEXT: [[TMP0:%.*]] = call i32 @__kmpc_global_thread_num(ptr @[[GLOB1]])
// CHECK11-NEXT: store i32 [[DOTCAPTURE_EXPR_]], ptr [[DOTCAPTURE_EXPR__ADDR]], align 4
// CHECK11-NEXT: [[TMP1:%.*]] = load i32, ptr [[DOTCAPTURE_EXPR__ADDR]], align 4
// CHECK11-NEXT: call void @__kmpc_push_num_threads(ptr @[[GLOB1]], i32 [[TMP0]], i32 [[TMP1]])
// CHECK11-NEXT: call void (ptr, i32, ptr, ...) @__kmpc_fork_call(ptr @[[GLOB1]], i32 0, ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZL7fstatici_l108.omp_outlined)
// CHECK11-NEXT: ret void
//
//
// CHECK11-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZL7fstatici_l108.omp_outlined
// CHECK11-SAME: (ptr noalias noundef [[DOTGLOBAL_TID_:%.*]], ptr noalias noundef [[DOTBOUND_TID_:%.*]]) #[[ATTR1]] {
// CHECK11-NEXT: entry:
// CHECK11-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca ptr, align 4
// CHECK11-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca ptr, align 4
// CHECK11-NEXT: store ptr [[DOTGLOBAL_TID_]], ptr [[DOTGLOBAL_TID__ADDR]], align 4
// CHECK11-NEXT: store ptr [[DOTBOUND_TID_]], ptr [[DOTBOUND_TID__ADDR]], align 4
// CHECK11-NEXT: ret void
//
//
// CHECK11-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZN2S12r1Ei_l121
// CHECK11-SAME: (ptr noundef [[THIS:%.*]], i32 noundef [[B:%.*]], i32 noundef [[DOTCAPTURE_EXPR_:%.*]]) #[[ATTR0]] {
// CHECK11-NEXT: entry:
// CHECK11-NEXT: [[THIS_ADDR:%.*]] = alloca ptr, align 4
// CHECK11-NEXT: [[B_ADDR:%.*]] = alloca i32, align 4
// CHECK11-NEXT: [[DOTCAPTURE_EXPR__ADDR:%.*]] = alloca i32, align 4
// CHECK11-NEXT: [[B_CASTED:%.*]] = alloca i32, align 4
// CHECK11-NEXT: [[TMP0:%.*]] = call i32 @__kmpc_global_thread_num(ptr @[[GLOB1]])
// CHECK11-NEXT: store ptr [[THIS]], ptr [[THIS_ADDR]], align 4
// CHECK11-NEXT: store i32 [[B]], ptr [[B_ADDR]], align 4
// CHECK11-NEXT: store i32 [[DOTCAPTURE_EXPR_]], ptr [[DOTCAPTURE_EXPR__ADDR]], align 4
// CHECK11-NEXT: [[TMP1:%.*]] = load ptr, ptr [[THIS_ADDR]], align 4
// CHECK11-NEXT: [[TMP2:%.*]] = load i32, ptr [[DOTCAPTURE_EXPR__ADDR]], align 4
// CHECK11-NEXT: call void @__kmpc_push_num_threads(ptr @[[GLOB1]], i32 [[TMP0]], i32 [[TMP2]])
// CHECK11-NEXT: [[TMP3:%.*]] = load i32, ptr [[B_ADDR]], align 4
// CHECK11-NEXT: store i32 [[TMP3]], ptr [[B_CASTED]], align 4
// CHECK11-NEXT: [[TMP4:%.*]] = load i32, ptr [[B_CASTED]], align 4
// CHECK11-NEXT: call void (ptr, i32, ptr, ...) @__kmpc_fork_call(ptr @[[GLOB1]], i32 2, ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZN2S12r1Ei_l121.omp_outlined, ptr [[TMP1]], i32 [[TMP4]])
// CHECK11-NEXT: ret void
//
//
// CHECK11-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZN2S12r1Ei_l121.omp_outlined
// CHECK11-SAME: (ptr noalias noundef [[DOTGLOBAL_TID_:%.*]], ptr noalias noundef [[DOTBOUND_TID_:%.*]], ptr noundef [[THIS:%.*]], i32 noundef [[B:%.*]]) #[[ATTR1]] {
// CHECK11-NEXT: entry:
// CHECK11-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca ptr, align 4
// CHECK11-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca ptr, align 4
// CHECK11-NEXT: [[THIS_ADDR:%.*]] = alloca ptr, align 4
// CHECK11-NEXT: [[B_ADDR:%.*]] = alloca i32, align 4
// CHECK11-NEXT: store ptr [[DOTGLOBAL_TID_]], ptr [[DOTGLOBAL_TID__ADDR]], align 4
// CHECK11-NEXT: store ptr [[DOTBOUND_TID_]], ptr [[DOTBOUND_TID__ADDR]], align 4
// CHECK11-NEXT: store ptr [[THIS]], ptr [[THIS_ADDR]], align 4
// CHECK11-NEXT: store i32 [[B]], ptr [[B_ADDR]], align 4
// CHECK11-NEXT: [[TMP0:%.*]] = load ptr, ptr [[THIS_ADDR]], align 4
// CHECK11-NEXT: [[TMP1:%.*]] = load i32, ptr [[B_ADDR]], align 4
// CHECK11-NEXT: [[CONV:%.*]] = sitofp i32 [[TMP1]] to double
// CHECK11-NEXT: [[ADD:%.*]] = fadd double [[CONV]], 1.500000e+00
// CHECK11-NEXT: [[A:%.*]] = getelementptr inbounds [[STRUCT_S1:%.*]], ptr [[TMP0]], i32 0, i32 0
// CHECK11-NEXT: store double [[ADD]], ptr [[A]], align 4
// CHECK11-NEXT: ret void
//
//
// CHECK11-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZN2S12r1Ei_l126
// CHECK11-SAME: (ptr noundef [[THIS:%.*]]) #[[ATTR0]] {
// CHECK11-NEXT: entry:
// CHECK11-NEXT: [[THIS_ADDR:%.*]] = alloca ptr, align 4
// CHECK11-NEXT: [[TMP0:%.*]] = call i32 @__kmpc_global_thread_num(ptr @[[GLOB1]])
// CHECK11-NEXT: store ptr [[THIS]], ptr [[THIS_ADDR]], align 4
// CHECK11-NEXT: [[TMP1:%.*]] = load ptr, ptr [[THIS_ADDR]], align 4
// CHECK11-NEXT: call void @__kmpc_push_num_threads(ptr @[[GLOB1]], i32 [[TMP0]], i32 1024)
// CHECK11-NEXT: call void (ptr, i32, ptr, ...) @__kmpc_fork_call(ptr @[[GLOB1]], i32 1, ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZN2S12r1Ei_l126.omp_outlined, ptr [[TMP1]])
// CHECK11-NEXT: ret void
//
//
// CHECK11-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__ZN2S12r1Ei_l126.omp_outlined
// CHECK11-SAME: (ptr noalias noundef [[DOTGLOBAL_TID_:%.*]], ptr noalias noundef [[DOTBOUND_TID_:%.*]], ptr noundef [[THIS:%.*]]) #[[ATTR1]] {
// CHECK11-NEXT: entry:
// CHECK11-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca ptr, align 4
// CHECK11-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca ptr, align 4
// CHECK11-NEXT: [[THIS_ADDR:%.*]] = alloca ptr, align 4
// CHECK11-NEXT: store ptr [[DOTGLOBAL_TID_]], ptr [[DOTGLOBAL_TID__ADDR]], align 4
// CHECK11-NEXT: store ptr [[DOTBOUND_TID_]], ptr [[DOTBOUND_TID__ADDR]], align 4
// CHECK11-NEXT: store ptr [[THIS]], ptr [[THIS_ADDR]], align 4
// CHECK11-NEXT: [[TMP0:%.*]] = load ptr, ptr [[THIS_ADDR]], align 4
// CHECK11-NEXT: [[A:%.*]] = getelementptr inbounds [[STRUCT_S1:%.*]], ptr [[TMP0]], i32 0, i32 0
// CHECK11-NEXT: store double 2.500000e+00, ptr [[A]], align 4
// CHECK11-NEXT: ret void
//
//
// CHECK11-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIiET_i_l88
// CHECK11-SAME: () #[[ATTR0]] {
// CHECK11-NEXT: entry:
// CHECK11-NEXT: [[TMP0:%.*]] = call i32 @__kmpc_global_thread_num(ptr @[[GLOB1]])
// CHECK11-NEXT: call void @__kmpc_push_num_threads(ptr @[[GLOB1]], i32 [[TMP0]], i32 20)
// CHECK11-NEXT: call void (ptr, i32, ptr, ...) @__kmpc_fork_call(ptr @[[GLOB1]], i32 0, ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIiET_i_l88.omp_outlined)
// CHECK11-NEXT: ret void
//
//
// CHECK11-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIiET_i_l88.omp_outlined
// CHECK11-SAME: (ptr noalias noundef [[DOTGLOBAL_TID_:%.*]], ptr noalias noundef [[DOTBOUND_TID_:%.*]]) #[[ATTR1]] {
// CHECK11-NEXT: entry:
// CHECK11-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca ptr, align 4
// CHECK11-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca ptr, align 4
// CHECK11-NEXT: store ptr [[DOTGLOBAL_TID_]], ptr [[DOTGLOBAL_TID__ADDR]], align 4
// CHECK11-NEXT: store ptr [[DOTBOUND_TID_]], ptr [[DOTBOUND_TID__ADDR]], align 4
// CHECK11-NEXT: ret void
//
//
// CHECK11-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIiET_i_l93
// CHECK11-SAME: (i32 noundef [[A:%.*]], i32 noundef [[B:%.*]], i32 noundef [[DOTCAPTURE_EXPR_:%.*]]) #[[ATTR0]] {
// CHECK11-NEXT: entry:
// CHECK11-NEXT: [[A_ADDR:%.*]] = alloca i32, align 4
// CHECK11-NEXT: [[B_ADDR:%.*]] = alloca i32, align 4
// CHECK11-NEXT: [[DOTCAPTURE_EXPR__ADDR:%.*]] = alloca i32, align 4
// CHECK11-NEXT: [[A_CASTED:%.*]] = alloca i32, align 4
// CHECK11-NEXT: [[B_CASTED:%.*]] = alloca i32, align 4
// CHECK11-NEXT: [[TMP0:%.*]] = call i32 @__kmpc_global_thread_num(ptr @[[GLOB1]])
// CHECK11-NEXT: store i32 [[A]], ptr [[A_ADDR]], align 4
// CHECK11-NEXT: store i32 [[B]], ptr [[B_ADDR]], align 4
// CHECK11-NEXT: store i32 [[DOTCAPTURE_EXPR_]], ptr [[DOTCAPTURE_EXPR__ADDR]], align 4
// CHECK11-NEXT: [[TMP1:%.*]] = load i16, ptr [[DOTCAPTURE_EXPR__ADDR]], align 2
// CHECK11-NEXT: [[TMP2:%.*]] = sext i16 [[TMP1]] to i32
// CHECK11-NEXT: call void @__kmpc_push_num_threads(ptr @[[GLOB1]], i32 [[TMP0]], i32 [[TMP2]])
// CHECK11-NEXT: [[TMP3:%.*]] = load i32, ptr [[A_ADDR]], align 4
// CHECK11-NEXT: store i32 [[TMP3]], ptr [[A_CASTED]], align 4
// CHECK11-NEXT: [[TMP4:%.*]] = load i32, ptr [[A_CASTED]], align 4
// CHECK11-NEXT: [[TMP5:%.*]] = load i16, ptr [[B_ADDR]], align 2
// CHECK11-NEXT: store i16 [[TMP5]], ptr [[B_CASTED]], align 2
// CHECK11-NEXT: [[TMP6:%.*]] = load i32, ptr [[B_CASTED]], align 4
// CHECK11-NEXT: call void (ptr, i32, ptr, ...) @__kmpc_fork_call(ptr @[[GLOB1]], i32 2, ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIiET_i_l93.omp_outlined, i32 [[TMP4]], i32 [[TMP6]])
// CHECK11-NEXT: ret void
//
//
// CHECK11-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z9ftemplateIiET_i_l93.omp_outlined
// CHECK11-SAME: (ptr noalias noundef [[DOTGLOBAL_TID_:%.*]], ptr noalias noundef [[DOTBOUND_TID_:%.*]], i32 noundef [[A:%.*]], i32 noundef [[B:%.*]]) #[[ATTR1]] {
// CHECK11-NEXT: entry:
// CHECK11-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca ptr, align 4
// CHECK11-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca ptr, align 4
// CHECK11-NEXT: [[A_ADDR:%.*]] = alloca i32, align 4
// CHECK11-NEXT: [[B_ADDR:%.*]] = alloca i32, align 4
// CHECK11-NEXT: store ptr [[DOTGLOBAL_TID_]], ptr [[DOTGLOBAL_TID__ADDR]], align 4
// CHECK11-NEXT: store ptr [[DOTBOUND_TID_]], ptr [[DOTBOUND_TID__ADDR]], align 4
// CHECK11-NEXT: store i32 [[A]], ptr [[A_ADDR]], align 4
// CHECK11-NEXT: store i32 [[B]], ptr [[B_ADDR]], align 4
// CHECK11-NEXT: [[TMP0:%.*]] = load i16, ptr [[B_ADDR]], align 2
// CHECK11-NEXT: [[CONV:%.*]] = sext i16 [[TMP0]] to i32
// CHECK11-NEXT: [[TMP1:%.*]] = load i32, ptr [[A_ADDR]], align 4
// CHECK11-NEXT: [[ADD:%.*]] = add nsw i32 [[TMP1]], [[CONV]]
// CHECK11-NEXT: store i32 [[ADD]], ptr [[A_ADDR]], align 4
// CHECK11-NEXT: ret void
//
Reference in New Issue View Git Blame Copy Permalink