maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity
llvm-project/clang/test/OpenMP/taskloop_simd_lastprivate_c...

527 lines
29 KiB
C++
Raw Normal View History

[OPENMP 4.5] Codegen for 'lastprivate' clauses in 'taskloop' directives. OpenMP 4.5 adds taskloop/taskloop simd directives. These directives allow to use lastprivate clause. Patch adds codegen for this clause. llvm-svn: 268618
2016-05-05 16:46:22 +08:00
// RUN: %clang_cc1 -verify -fopenmp -x c++ -triple x86_64-apple-darwin10 -emit-llvm %s -o - | FileCheck %s
// RUN: %clang_cc1 -fopenmp -x c++ -std=c++11 -triple x86_64-apple-darwin10 -emit-pch -o %t %s
// RUN: %clang_cc1 -fopenmp -x c++ -triple x86_64-apple-darwin10 -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s
// RUN: %clang_cc1 -verify -fopenmp -x c++ -std=c++11 -DLAMBDA -triple x86_64-apple-darwin10 -emit-llvm %s -o - | FileCheck -check-prefix=LAMBDA %s
// RUN: %clang_cc1 -verify -fopenmp -x c++ -fblocks -DBLOCKS -triple x86_64-apple-darwin10 -emit-llvm %s -o - | FileCheck -check-prefix=BLOCKS %s
// RUN: %clang_cc1 -verify -fopenmp -x c++ -std=c++11 -DARRAY -triple x86_64-apple-darwin10 -emit-llvm %s -o - | FileCheck -check-prefix=ARRAY %s
[OPENMP] Support for -fopenmp-simd option with compilation of simd loops only. Added support for -fopenmp-simd option that allows compilation of simd-based constructs without emission of OpenMP runtime calls. llvm-svn: 321560
2017-12-30 02:07:07 +08:00
// RUN: %clang_cc1 -verify -fopenmp-simd -x c++ -triple x86_64-apple-darwin10 -emit-llvm %s -o - | FileCheck --check-prefix SIMD-ONLY0 %s
// RUN: %clang_cc1 -fopenmp-simd -x c++ -std=c++11 -triple x86_64-apple-darwin10 -emit-pch -o %t %s
// RUN: %clang_cc1 -fopenmp-simd -x c++ -triple x86_64-apple-darwin10 -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck --check-prefix SIMD-ONLY0 %s
// RUN: %clang_cc1 -verify -fopenmp-simd -x c++ -std=c++11 -DLAMBDA -triple x86_64-apple-darwin10 -emit-llvm %s -o - | FileCheck --check-prefix SIMD-ONLY0 %s
// RUN: %clang_cc1 -verify -fopenmp-simd -x c++ -fblocks -DBLOCKS -triple x86_64-apple-darwin10 -emit-llvm %s -o - | FileCheck --check-prefix SIMD-ONLY0 %s
// RUN: %clang_cc1 -verify -fopenmp-simd -x c++ -std=c++11 -DARRAY -triple x86_64-apple-darwin10 -emit-llvm %s -o - | FileCheck --check-prefix SIMD-ONLY0 %s
// SIMD-ONLY0-NOT: {{__kmpc|__tgt}}
[OPENMP 4.5] Codegen for 'lastprivate' clauses in 'taskloop' directives. OpenMP 4.5 adds taskloop/taskloop simd directives. These directives allow to use lastprivate clause. Patch adds codegen for this clause. llvm-svn: 268618
2016-05-05 16:46:22 +08:00
// expected-no-diagnostics
Revert 9 changes from r284793, they still fail on some bots llvm-svn: 284794
2016-10-21 07:30:39 +08:00
// It doesn't pass on win32.
// REQUIRES: shell
[OPENMP 4.5] Codegen for 'lastprivate' clauses in 'taskloop' directives. OpenMP 4.5 adds taskloop/taskloop simd directives. These directives allow to use lastprivate clause. Patch adds codegen for this clause. llvm-svn: 268618
2016-05-05 16:46:22 +08:00
#ifndef ARRAY
#ifndef HEADER
#define HEADER
template <class T>
struct S {
T f;
S(T a) : f(a) {}
S() : f() {}
S(const S &s, T t = T()) : f(s.f + t) {}
operator T() { return T(); }
~S() {}
};
volatile double g;
[OPENMP] Codegen for reduction clauses in 'taskloop' directives. Adds codegen for taskloop-based directives. llvm-svn: 308174
2017-07-17 21:30:36 +08:00
// CHECK-DAG: [[KMP_TASK_T_TY:%.+]] = type { i8*, i32 (i32, i8*)*, i32, %union{{.+}}, %union{{.+}}, i64, i64, i64, i32, i8* }
[OPENMP 4.5] Codegen for 'lastprivate' clauses in 'taskloop' directives. OpenMP 4.5 adds taskloop/taskloop simd directives. These directives allow to use lastprivate clause. Patch adds codegen for this clause. llvm-svn: 268618
2016-05-05 16:46:22 +08:00
// CHECK-DAG: [[S_DOUBLE_TY:%.+]] = type { double }
// CHECK-DAG: [[PRIVATES_MAIN_TY:%.+]] = type {{.?}}{ [2 x [[S_DOUBLE_TY]]], [[S_DOUBLE_TY]], i32, [2 x i32]
// CHECK-DAG: [[CAP_MAIN_TY:%.+]] = type { [2 x i32]*, i32*, [2 x [[S_DOUBLE_TY]]]*, [[S_DOUBLE_TY]]*, i{{[0-9]+}}* }
// CHECK-DAG: [[KMP_TASK_MAIN_TY:%.+]] = type { [[KMP_TASK_T_TY]], [[PRIVATES_MAIN_TY]] }
// CHECK-DAG: [[S_INT_TY:%.+]] = type { i32 }
// CHECK-DAG: [[CAP_TMAIN_TY:%.+]] = type { [2 x i32]*, i32*, [2 x [[S_INT_TY]]]*, [[S_INT_TY]]* }
// CHECK-DAG: [[PRIVATES_TMAIN_TY:%.+]] = type { i32, [2 x i32], [2 x [[S_INT_TY]]], [[S_INT_TY]], [104 x i8] }
// CHECK-DAG: [[KMP_TASK_TMAIN_TY:%.+]] = type { [[KMP_TASK_T_TY]], [{{[0-9]+}} x i8], [[PRIVATES_TMAIN_TY]] }
template <typename T>
T tmain() {
S<T> ttt;
S<T> test;
T t_var __attribute__((aligned(128))) = T();
T vec[] = {1, 2};
S<T> s_arr[] = {1, 2};
S<T> var(3);
#pragma omp taskloop simd lastprivate(t_var, vec, s_arr, s_arr, var, var)
for (int i = 0; i < 10; ++i) {
vec[0] = t_var;
s_arr[0] = var;
}
return T();
}
int main() {
static int sivar;
#ifdef LAMBDA
// LAMBDA: [[G:@.+]] = global double
// LAMBDA: [[SIVAR:@.+]] = internal global i{{[0-9]+}} 0,
// LAMBDA-LABEL: @main
// LAMBDA: call{{( x86_thiscallcc)?}} void [[OUTER_LAMBDA:@.+]](
[&]() {
// LAMBDA: define{{.*}} internal{{.*}} void [[OUTER_LAMBDA]](
[OPENMP] Codegen for reduction clauses in 'taskloop' directives. Adds codegen for taskloop-based directives. llvm-svn: 308174
2017-07-17 21:30:36 +08:00
// LAMBDA: [[RES:%.+]] = call i8* @__kmpc_omp_task_alloc(%{{[^ ]+}} @{{[^,]+}}, i32 %{{[^,]+}}, i32 1, i64 96, i64 16, i32 (i32, i8*)* bitcast (i32 (i32, %{{[^*]+}}*)* [[TASK_ENTRY:@[^ ]+]] to i32 (i32, i8*)*))
[OPENMP 4.5] Codegen for 'lastprivate' clauses in 'taskloop' directives. OpenMP 4.5 adds taskloop/taskloop simd directives. These directives allow to use lastprivate clause. Patch adds codegen for this clause. llvm-svn: 268618
2016-05-05 16:46:22 +08:00
// LAMBDA: [[PRIVATES:%.+]] = getelementptr inbounds %{{.+}}, %{{.+}}* %{{.+}}, i{{.+}} 0, i{{.+}} 1
Do not always request an implicit taskgroup region inside the kmpc_taskloop function Summary: For the following code: ``` int i; #pragma omp taskloop for (i = 0; i < 100; ++i) {} #pragma omp taskloop nogroup for (i = 0; i < 100; ++i) {} ``` Clang emits the following LLVM IR: ``` ... call void @__kmpc_taskgroup(%struct.ident_t* @0, i32 %0) %2 = call i8* @__kmpc_omp_task_alloc(%struct.ident_t* @0, i32 %0, i32 1, i64 80, i64 8, i32 (i32, i8*)* bitcast (i32 (i32, %struct.kmp_task_t_with_privates*)* @.omp_task_entry. to i32 (i32, i8*)*)) ... call void @__kmpc_taskloop(%struct.ident_t* @0, i32 %0, i8* %2, i32 1, i64* %8, i64* %9, i64 %13, i32 0, i32 0, i64 0, i8* null) call void @__kmpc_end_taskgroup(%struct.ident_t* @0, i32 %0) ... %15 = call i8* @__kmpc_omp_task_alloc(%struct.ident_t* @0, i32 %0, i32 1, i64 80, i64 8, i32 (i32, i8*)* bitcast (i32 (i32, %struct.kmp_task_t_with_privates.1*)* @.omp_task_entry..2 to i32 (i32, i8*)*)) ... call void @__kmpc_taskloop(%struct.ident_t* @0, i32 %0, i8* %15, i32 1, i64* %21, i64* %22, i64 %26, i32 0, i32 0, i64 0, i8* null) ``` The first set of instructions corresponds to the first taskloop construct. It is important to note that the implicit taskgroup region associated with the taskloop construct has been materialized in our IR: the `__kmpc_taskloop` occurs inside a taskgroup region. Note also that this taskgroup region does not exist in our second taskloop because we are using the `nogroup` clause. The issue here is the 4th argument of the kmpc_taskloop call, starting from the end, is always a zero. Checking the LLVM OpenMP RT implementation, we see that this argument corresponds to the nogroup parameter: ``` void __kmpc_taskloop(ident_t *loc, int gtid, kmp_task_t *task, int if_val, kmp_uint64 *lb, kmp_uint64 *ub, kmp_int64 st, int nogroup, int sched, kmp_uint64 grainsize, void *task_dup); ``` So basically we always tell to the RT to do another taskgroup region. For the first taskloop, this means that we create two taskgroup regions. For the second example, it means that despite the fact we had a nogroup clause we are going to have a taskgroup region, so we unnecessary wait until all descendant tasks have been executed. Reviewers: ABataev Reviewed By: ABataev Subscribers: rogfer01, cfe-commits Differential Revision: https://reviews.llvm.org/D53636 llvm-svn: 345180
2018-10-25 03:06:37 +08:00
// LAMBDA: call void @__kmpc_taskloop(%{{.+}}* @{{.+}}, i32 %{{.+}}, i8* [[RES]], i32 1, i64* %{{.+}}, i64* %{{.+}}, i64 %{{.+}}, i32 1, i32 0, i64 0, i8* bitcast (void ([[KMP_TASK_MAIN_TY:%[^*]+]]*, [[KMP_TASK_MAIN_TY]]*, i32)* [[MAIN_DUP:@.+]] to i8*))
[OPENMP 4.5] Codegen for 'lastprivate' clauses in 'taskloop' directives. OpenMP 4.5 adds taskloop/taskloop simd directives. These directives allow to use lastprivate clause. Patch adds codegen for this clause. llvm-svn: 268618
2016-05-05 16:46:22 +08:00
// LAMBDA: ret
#pragma omp taskloop simd lastprivate(g, sivar)
for (int i = 0; i < 10; ++i) {
// LAMBDA: define {{.+}} void [[INNER_LAMBDA:@.+]](%{{.+}}* [[ARG_PTR:%.+]])
// LAMBDA: store %{{.+}}* [[ARG_PTR]], %{{.+}}** [[ARG_PTR_REF:%.+]],
// LAMBDA: [[ARG_PTR:%.+]] = load %{{.+}}*, %{{.+}}** [[ARG_PTR_REF]]
// LAMBDA: [[G_PTR_REF:%.+]] = getelementptr inbounds %{{.+}}, %{{.+}}* [[ARG_PTR]], i{{[0-9]+}} 0, i{{[0-9]+}} 0
// LAMBDA: [[G_REF:%.+]] = load double*, double** [[G_PTR_REF]]
// LAMBDA: store double 2.0{{.+}}, double* [[G_REF]]
// LAMBDA: store double* %{{.+}}, double** %{{.+}},
// LAMBDA: define internal i32 [[TASK_ENTRY]](i32, %{{.+}}* noalias)
g = 1;
sivar = 11;
// LAMBDA: store double 1.0{{.+}}, double* %{{.+}},
// LAMBDA: store i{{[0-9]+}} 11, i{{[0-9]+}}* %{{.+}},
// LAMBDA: call void [[INNER_LAMBDA]](%
// LAMBDA: icmp ne i32 %{{.+}}, 0
// LAMBDA: br i1
// LAMBDA: load double, double* %
// LAMBDA: store volatile double %
// LAMBDA: load i32, i32* %
// LAMBDA: store i32 %
// LAMBDA: ret
[&]() {
g = 2;
sivar = 22;
}();
}
}();
return 0;
#elif defined(BLOCKS)
// BLOCKS: [[G:@.+]] = global double
// BLOCKS-LABEL: @main
// BLOCKS: call void {{%.+}}(i8
^{
// BLOCKS: define{{.*}} internal{{.*}} void {{.+}}(i8*
[OPENMP] Codegen for reduction clauses in 'taskloop' directives. Adds codegen for taskloop-based directives. llvm-svn: 308174
2017-07-17 21:30:36 +08:00
// BLOCKS: [[RES:%.+]] = call i8* @__kmpc_omp_task_alloc(%{{[^ ]+}} @{{[^,]+}}, i32 %{{[^,]+}}, i32 1, i64 96, i64 16, i32 (i32, i8*)* bitcast (i32 (i32, %{{[^*]+}}*)* [[TASK_ENTRY:@[^ ]+]] to i32 (i32, i8*)*))
[OPENMP 4.5] Codegen for 'lastprivate' clauses in 'taskloop' directives. OpenMP 4.5 adds taskloop/taskloop simd directives. These directives allow to use lastprivate clause. Patch adds codegen for this clause. llvm-svn: 268618
2016-05-05 16:46:22 +08:00
// BLOCKS: [[PRIVATES:%.+]] = getelementptr inbounds %{{.+}}, %{{.+}}* %{{.+}}, i{{.+}} 0, i{{.+}} 1
Do not always request an implicit taskgroup region inside the kmpc_taskloop function Summary: For the following code: ``` int i; #pragma omp taskloop for (i = 0; i < 100; ++i) {} #pragma omp taskloop nogroup for (i = 0; i < 100; ++i) {} ``` Clang emits the following LLVM IR: ``` ... call void @__kmpc_taskgroup(%struct.ident_t* @0, i32 %0) %2 = call i8* @__kmpc_omp_task_alloc(%struct.ident_t* @0, i32 %0, i32 1, i64 80, i64 8, i32 (i32, i8*)* bitcast (i32 (i32, %struct.kmp_task_t_with_privates*)* @.omp_task_entry. to i32 (i32, i8*)*)) ... call void @__kmpc_taskloop(%struct.ident_t* @0, i32 %0, i8* %2, i32 1, i64* %8, i64* %9, i64 %13, i32 0, i32 0, i64 0, i8* null) call void @__kmpc_end_taskgroup(%struct.ident_t* @0, i32 %0) ... %15 = call i8* @__kmpc_omp_task_alloc(%struct.ident_t* @0, i32 %0, i32 1, i64 80, i64 8, i32 (i32, i8*)* bitcast (i32 (i32, %struct.kmp_task_t_with_privates.1*)* @.omp_task_entry..2 to i32 (i32, i8*)*)) ... call void @__kmpc_taskloop(%struct.ident_t* @0, i32 %0, i8* %15, i32 1, i64* %21, i64* %22, i64 %26, i32 0, i32 0, i64 0, i8* null) ``` The first set of instructions corresponds to the first taskloop construct. It is important to note that the implicit taskgroup region associated with the taskloop construct has been materialized in our IR: the `__kmpc_taskloop` occurs inside a taskgroup region. Note also that this taskgroup region does not exist in our second taskloop because we are using the `nogroup` clause. The issue here is the 4th argument of the kmpc_taskloop call, starting from the end, is always a zero. Checking the LLVM OpenMP RT implementation, we see that this argument corresponds to the nogroup parameter: ``` void __kmpc_taskloop(ident_t *loc, int gtid, kmp_task_t *task, int if_val, kmp_uint64 *lb, kmp_uint64 *ub, kmp_int64 st, int nogroup, int sched, kmp_uint64 grainsize, void *task_dup); ``` So basically we always tell to the RT to do another taskgroup region. For the first taskloop, this means that we create two taskgroup regions. For the second example, it means that despite the fact we had a nogroup clause we are going to have a taskgroup region, so we unnecessary wait until all descendant tasks have been executed. Reviewers: ABataev Reviewed By: ABataev Subscribers: rogfer01, cfe-commits Differential Revision: https://reviews.llvm.org/D53636 llvm-svn: 345180
2018-10-25 03:06:37 +08:00
// BLOCKS: call void @__kmpc_taskloop(%{{.+}}* @{{.+}}, i32 %{{.+}}, i8* [[RES]], i32 1, i64* %{{.+}}, i64* %{{.+}}, i64 %{{.+}}, i32 1, i32 0, i64 0, i8* bitcast (void ([[KMP_TASK_MAIN_TY:%[^*]+]]*, [[KMP_TASK_MAIN_TY]]*, i32)* [[MAIN_DUP:@.+]] to i8*))
[OPENMP 4.5] Codegen for 'lastprivate' clauses in 'taskloop' directives. OpenMP 4.5 adds taskloop/taskloop simd directives. These directives allow to use lastprivate clause. Patch adds codegen for this clause. llvm-svn: 268618
2016-05-05 16:46:22 +08:00
// BLOCKS: ret
#pragma omp taskloop simd lastprivate(g, sivar)
for (int i = 0; i < 10; ++i) {
// BLOCKS: define {{.+}} void {{@.+}}(i8*
// BLOCKS-NOT: [[G]]{{[[^:word:]]}}
// BLOCKS: store double 2.0{{.+}}, double*
// BLOCKS-NOT: [[G]]{{[[^:word:]]}}
// BLOCKS-NOT: [[ISVAR]]{{[[^:word:]]}}
// BLOCKS: store i{{[0-9]+}} 22, i{{[0-9]+}}*
// BLOCKS-NOT: [[SIVAR]]{{[[^:word:]]}}
// BLOCKS: ret
// BLOCKS: store double* %{{.+}}, double** %{{.+}},
// BLOCKS: store i{{[0-9]+}}* %{{.+}}, i{{[0-9]+}}** %{{.+}},
// BLOCKS: define internal i32 [[TASK_ENTRY]](i32, %{{.+}}* noalias)
g = 1;
sivar = 11;
// BLOCKS: store double 1.0{{.+}}, double* %{{.+}},
// BLOCKS-NOT: [[G]]{{[[^:word:]]}}
// BLOCKS: store i{{[0-9]+}} 11, i{{[0-9]+}}* %{{.+}},
// BLOCKS-NOT: [[SIVAR]]{{[[^:word:]]}}
// BLOCKS: call void {{%.+}}(i8
// BLOCKS: icmp ne i32 %{{.+}}, 0
// BLOCKS: br i1
// BLOCKS: load double, double* %
// BLOCKS: store volatile double %
// BLOCKS: load i32, i32* %
// BLOCKS: store i32 %
^{
g = 2;
sivar = 22;
}();
}
}();
return 0;
#else
S<double> ttt;
S<double> test;
int t_var = 0;
int vec[] = {1, 2};
S<double> s_arr[] = {1, 2};
S<double> var(3);
#pragma omp taskloop simd lastprivate(var, t_var, s_arr, vec, s_arr, var, sivar)
for (int i = 0; i < 10; ++i) {
vec[0] = t_var;
s_arr[0] = var;
sivar = 33;
}
return tmain<int>();
#endif
}
// CHECK: [[SIVAR:.+]] = internal global i{{[0-9]+}} 0,
// CHECK: define i{{[0-9]+}} @main()
// CHECK: alloca [[S_DOUBLE_TY]],
// CHECK: [[TEST:%.+]] = alloca [[S_DOUBLE_TY]],
// CHECK: [[T_VAR_ADDR:%.+]] = alloca i32,
// CHECK: [[VEC_ADDR:%.+]] = alloca [2 x i32],
// CHECK: [[S_ARR_ADDR:%.+]] = alloca [2 x [[S_DOUBLE_TY]]],
// CHECK: [[VAR_ADDR:%.+]] = alloca [[S_DOUBLE_TY]],
// CHECK: [[GTID:%.+]] = call i32 @__kmpc_global_thread_num([[LOC:%.+]])
Revert "[CodeGenCXX] Treat 'this' as noalias in constructors" This reverts commit https://reviews.llvm.org/rL344150 which causes MachineOutliner related failures on the ppc64le multistage buildbot. llvm-svn: 344526
2018-10-15 23:43:00 +08:00
// CHECK: call {{.*}} [[S_DOUBLE_TY_CONSTR:@.+]]([[S_DOUBLE_TY]]* [[TEST]])
[OPENMP 4.5] Codegen for 'lastprivate' clauses in 'taskloop' directives. OpenMP 4.5 adds taskloop/taskloop simd directives. These directives allow to use lastprivate clause. Patch adds codegen for this clause. llvm-svn: 268618
2016-05-05 16:46:22 +08:00
// Store original variables in capture struct.
// CHECK: [[VEC_REF:%.+]] = getelementptr inbounds [[CAP_MAIN_TY]], [[CAP_MAIN_TY]]* %{{.+}}, i{{[0-9]+}} 0, i{{[0-9]+}} 0
// CHECK: store [2 x i32]* [[VEC_ADDR]], [2 x i32]** [[VEC_REF]],
// CHECK: [[T_VAR_REF:%.+]] = getelementptr inbounds [[CAP_MAIN_TY]], [[CAP_MAIN_TY]]* %{{.+}}, i{{[0-9]+}} 0, i{{[0-9]+}} 1
// CHECK: store i32* [[T_VAR_ADDR]], i32** [[T_VAR_REF]],
// CHECK: [[S_ARR_REF:%.+]] = getelementptr inbounds [[CAP_MAIN_TY]], [[CAP_MAIN_TY]]* %{{.+}}, i{{[0-9]+}} 0, i{{[0-9]+}} 2
// CHECK: store [2 x [[S_DOUBLE_TY]]]* [[S_ARR_ADDR]], [2 x [[S_DOUBLE_TY]]]** [[S_ARR_REF]],
// CHECK: [[VAR_REF:%.+]] = getelementptr inbounds [[CAP_MAIN_TY]], [[CAP_MAIN_TY]]* %{{.+}}, i{{[0-9]+}} 0, i{{[0-9]+}} 3
// CHECK: store [[S_DOUBLE_TY]]* [[VAR_ADDR]], [[S_DOUBLE_TY]]** [[VAR_REF]],
// CHECK: [[SIVAR_REF:%.+]] = getelementptr inbounds [[CAP_MAIN_TY]], [[CAP_MAIN_TY]]* %{{.+}}, i{{[0-9]+}} 0, i{{[0-9]+}} 4
// CHECK: store i{{[0-9]+}}* [[SIVAR]], i{{[0-9]+}}** [[SIVAR_REF]],
// Allocate task.
// Returns struct kmp_task_t {
// [[KMP_TASK_T]] task_data;
// [[KMP_TASK_MAIN_TY]] privates;
// };
[OPENMP] Codegen for reduction clauses in 'taskloop' directives. Adds codegen for taskloop-based directives. llvm-svn: 308174
2017-07-17 21:30:36 +08:00
// CHECK: [[RES:%.+]] = call i8* @__kmpc_omp_task_alloc([[LOC]], i32 [[GTID]], i32 9, i64 120, i64 40, i32 (i32, i8*)* bitcast (i32 (i32, [[KMP_TASK_MAIN_TY]]*)* [[TASK_ENTRY:@[^ ]+]] to i32 (i32, i8*)*))
[OPENMP 4.5] Codegen for 'lastprivate' clauses in 'taskloop' directives. OpenMP 4.5 adds taskloop/taskloop simd directives. These directives allow to use lastprivate clause. Patch adds codegen for this clause. llvm-svn: 268618
2016-05-05 16:46:22 +08:00
// CHECK: [[RES_KMP_TASK:%.+]] = bitcast i8* [[RES]] to [[KMP_TASK_MAIN_TY]]*
// Fill kmp_task_t->shareds by copying from original capture argument.
// CHECK: [[TASK:%.+]] = getelementptr inbounds [[KMP_TASK_MAIN_TY]], [[KMP_TASK_MAIN_TY]]* [[RES_KMP_TASK]], i{{[0-9]+}} 0, i{{[0-9]+}} 0
// CHECK: [[SHAREDS_REF_ADDR:%.+]] = getelementptr inbounds [[KMP_TASK_T_TY]], [[KMP_TASK_T_TY]]* [[TASK]], i{{[0-9]+}} 0, i{{[0-9]+}} 0
// CHECK: [[SHAREDS_REF:%.+]] = load i8*, i8** [[SHAREDS_REF_ADDR]],
// CHECK: [[CAPTURES_ADDR:%.+]] = bitcast [[CAP_MAIN_TY]]* %{{.+}} to i8*
Change memcpy/memove/memset to have dest and source alignment attributes (Step 1). Summary: Upstream LLVM is changing the the prototypes of the @llvm.memcpy/memmove/memset intrinsics. This change updates the Clang tests for this change. The @llvm.memcpy/memmove/memset intrinsics currently have an explicit argument which is required to be a constant integer. It represents the alignment of the dest (and source), and so must be the minimum of the actual alignment of the two. This change removes the alignment argument in favour of placing the alignment attribute on the source and destination pointers of the memory intrinsic call. For example, code which used to read: call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* %src, i32 100, i32 4, i1 false) will now read call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 4 %dest, i8* align 4 %src, i32 100, i1 false) At this time the source and destination alignments must be the same (Step 1). Step 2 of the change, to be landed shortly, will relax that contraint and allow the source and destination to have different alignments. llvm-svn: 322964
2018-01-20 01:12:54 +08:00
// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 8 [[SHAREDS_REF]], i8* align 8 [[CAPTURES_ADDR]], i64 40, i1 false)
[OPENMP 4.5] Codegen for 'lastprivate' clauses in 'taskloop' directives. OpenMP 4.5 adds taskloop/taskloop simd directives. These directives allow to use lastprivate clause. Patch adds codegen for this clause. llvm-svn: 268618
2016-05-05 16:46:22 +08:00
// Initialize kmp_task_t->privates with default values (no init for simple types, default constructors for classes).
// Also copy address of private copy to the corresponding shareds reference.
// CHECK: [[PRIVATES:%.+]] = getelementptr inbounds [[KMP_TASK_MAIN_TY]], [[KMP_TASK_MAIN_TY]]* [[RES_KMP_TASK]], i{{[0-9]+}} 0, i{{[0-9]+}} 1
// Constructors for s_arr and var.
// s_arr;
// CHECK: [[PRIVATE_S_ARR_REF:%.+]] = getelementptr inbounds [[PRIVATES_MAIN_TY]], [[PRIVATES_MAIN_TY]]* [[PRIVATES]], i{{[0-9]+}} 0, i{{[0-9]+}} 0
Revert "[CodeGenCXX] Treat 'this' as noalias in constructors" This reverts commit https://reviews.llvm.org/rL344150 which causes MachineOutliner related failures on the ppc64le multistage buildbot. llvm-svn: 344526
2018-10-15 23:43:00 +08:00
// CHECK: call {{.*}} [[S_DOUBLE_TY_CONSTR]]([[S_DOUBLE_TY]]* [[S_ARR_CUR:%[^,]+]])
[OPENMP 4.5] Codegen for 'lastprivate' clauses in 'taskloop' directives. OpenMP 4.5 adds taskloop/taskloop simd directives. These directives allow to use lastprivate clause. Patch adds codegen for this clause. llvm-svn: 268618
2016-05-05 16:46:22 +08:00
// CHECK: getelementptr inbounds [[S_DOUBLE_TY]], [[S_DOUBLE_TY]]* [[S_ARR_CUR]], i{{.+}} 1
// CHECK: icmp eq
// CHECK: br i1
// var;
// CHECK: [[PRIVATE_VAR_REF:%.+]] = getelementptr inbounds [[PRIVATES_MAIN_TY]], [[PRIVATES_MAIN_TY]]* [[PRIVATES]], i{{.+}} 0, i{{.+}} 1
Revert "[CodeGenCXX] Treat 'this' as noalias in constructors" This reverts commit https://reviews.llvm.org/rL344150 which causes MachineOutliner related failures on the ppc64le multistage buildbot. llvm-svn: 344526
2018-10-15 23:43:00 +08:00
// CHECK: call {{.*}} [[S_DOUBLE_TY_CONSTR]]([[S_DOUBLE_TY]]* [[PRIVATE_VAR_REF]])
[OPENMP 4.5] Codegen for 'lastprivate' clauses in 'taskloop' directives. OpenMP 4.5 adds taskloop/taskloop simd directives. These directives allow to use lastprivate clause. Patch adds codegen for this clause. llvm-svn: 268618
2016-05-05 16:46:22 +08:00
// t_var;
// vec;
// sivar;
// Provide pointer to destructor function, which will destroy private variables at the end of the task.
// CHECK: [[DESTRUCTORS_REF:%.+]] = getelementptr inbounds [[KMP_TASK_T_TY]], [[KMP_TASK_T_TY]]* [[TASK]], i{{.+}} 0, i{{.+}} 3
[OPENMP 4.5] Fixed codegen for 'priority' and destructors in task-based directives. 'kmp_task_t' record type added a new field for 'priority' clause and changed the representation of pointer to destructors for privates used within loop-based directives. Old representation: typedef struct kmp_task { /* GEH: Shouldn't this be aligned somehow? */ void *shareds; /**< pointer to block of pointers to shared vars */ kmp_routine_entry_t routine; /**< pointer to routine to call for executing task */ kmp_int32 part_id; /**< part id for the task */ kmp_routine_entry_t destructors; /* pointer to function to invoke deconstructors of firstprivate C++ objects */ /* private vars */ } kmp_task_t; New representation: typedef struct kmp_task { /* GEH: Shouldn't this be aligned somehow? */ void *shareds; /**< pointer to block of pointers to shared vars */ kmp_routine_entry_t routine; /**< pointer to routine to call for executing task */ kmp_int32 part_id; /**< part id for the task */ kmp_cmplrdata_t data1; /* Two known optional additions: destructors and priority */ kmp_cmplrdata_t data2; /* Process destructors first, priority second */ /* future data */ /* private vars */ } kmp_task_t; Also excessive initialization of 'destructors' fields to 'null' was removed from codegen if it is known that no destructors shal be used. Currently a special bit is used in 'kmp_tasking_flags_t' bitfields ('destructors_thunk' bitfield). llvm-svn: 271201
2016-05-30 17:06:50 +08:00
// CHECK: [[DESTRUCTORS_PTR:%.+]] = bitcast %union{{.+}}* [[DESTRUCTORS_REF]] to i32 (i32, i8*)**
// CHECK: store i32 (i32, i8*)* bitcast (i32 (i32, [[KMP_TASK_MAIN_TY]]*)* [[DESTRUCTORS:@.+]] to i32 (i32, i8*)*), i32 (i32, i8*)** [[DESTRUCTORS_PTR]],
[OPENMP 4.5] Codegen for 'lastprivate' clauses in 'taskloop' directives. OpenMP 4.5 adds taskloop/taskloop simd directives. These directives allow to use lastprivate clause. Patch adds codegen for this clause. llvm-svn: 268618
2016-05-05 16:46:22 +08:00
// Start task.
Do not always request an implicit taskgroup region inside the kmpc_taskloop function Summary: For the following code: ``` int i; #pragma omp taskloop for (i = 0; i < 100; ++i) {} #pragma omp taskloop nogroup for (i = 0; i < 100; ++i) {} ``` Clang emits the following LLVM IR: ``` ... call void @__kmpc_taskgroup(%struct.ident_t* @0, i32 %0) %2 = call i8* @__kmpc_omp_task_alloc(%struct.ident_t* @0, i32 %0, i32 1, i64 80, i64 8, i32 (i32, i8*)* bitcast (i32 (i32, %struct.kmp_task_t_with_privates*)* @.omp_task_entry. to i32 (i32, i8*)*)) ... call void @__kmpc_taskloop(%struct.ident_t* @0, i32 %0, i8* %2, i32 1, i64* %8, i64* %9, i64 %13, i32 0, i32 0, i64 0, i8* null) call void @__kmpc_end_taskgroup(%struct.ident_t* @0, i32 %0) ... %15 = call i8* @__kmpc_omp_task_alloc(%struct.ident_t* @0, i32 %0, i32 1, i64 80, i64 8, i32 (i32, i8*)* bitcast (i32 (i32, %struct.kmp_task_t_with_privates.1*)* @.omp_task_entry..2 to i32 (i32, i8*)*)) ... call void @__kmpc_taskloop(%struct.ident_t* @0, i32 %0, i8* %15, i32 1, i64* %21, i64* %22, i64 %26, i32 0, i32 0, i64 0, i8* null) ``` The first set of instructions corresponds to the first taskloop construct. It is important to note that the implicit taskgroup region associated with the taskloop construct has been materialized in our IR: the `__kmpc_taskloop` occurs inside a taskgroup region. Note also that this taskgroup region does not exist in our second taskloop because we are using the `nogroup` clause. The issue here is the 4th argument of the kmpc_taskloop call, starting from the end, is always a zero. Checking the LLVM OpenMP RT implementation, we see that this argument corresponds to the nogroup parameter: ``` void __kmpc_taskloop(ident_t *loc, int gtid, kmp_task_t *task, int if_val, kmp_uint64 *lb, kmp_uint64 *ub, kmp_int64 st, int nogroup, int sched, kmp_uint64 grainsize, void *task_dup); ``` So basically we always tell to the RT to do another taskgroup region. For the first taskloop, this means that we create two taskgroup regions. For the second example, it means that despite the fact we had a nogroup clause we are going to have a taskgroup region, so we unnecessary wait until all descendant tasks have been executed. Reviewers: ABataev Reviewed By: ABataev Subscribers: rogfer01, cfe-commits Differential Revision: https://reviews.llvm.org/D53636 llvm-svn: 345180
2018-10-25 03:06:37 +08:00
// CHECK: call void @__kmpc_taskloop([[LOC]], i32 [[GTID]], i8* [[RES]], i32 1, i64* %{{.+}}, i64* %{{.+}}, i64 %{{.+}}, i32 1, i32 0, i64 0, i8* bitcast (void ([[KMP_TASK_MAIN_TY]]*, [[KMP_TASK_MAIN_TY]]*, i32)* [[MAIN_DUP:@.+]] to i8*))
[OPENMP 4.5] Codegen for 'lastprivate' clauses in 'taskloop' directives. OpenMP 4.5 adds taskloop/taskloop simd directives. These directives allow to use lastprivate clause. Patch adds codegen for this clause. llvm-svn: 268618
2016-05-05 16:46:22 +08:00
// CHECK: = call i{{.+}} [[TMAIN_INT:@.+]]()
// No destructors must be called for private copies of s_arr and var.
// CHECK-NOT: getelementptr inbounds [[PRIVATES_MAIN_TY]], [[PRIVATES_MAIN_TY]]* [[PRIVATES]], i{{.+}} 0, i{{.+}} 2
// CHECK-NOT: getelementptr inbounds [[PRIVATES_MAIN_TY]], [[PRIVATES_MAIN_TY]]* [[PRIVATES]], i{{.+}} 0, i{{.+}} 3
// CHECK: call void [[S_DOUBLE_TY_DESTR:@.+]]([[S_DOUBLE_TY]]*
// CHECK-NOT: getelementptr inbounds [[PRIVATES_MAIN_TY]], [[PRIVATES_MAIN_TY]]* [[PRIVATES]], i{{.+}} 0, i{{.+}} 2
// CHECK-NOT: getelementptr inbounds [[PRIVATES_MAIN_TY]], [[PRIVATES_MAIN_TY]]* [[PRIVATES]], i{{.+}} 0, i{{.+}} 3
// CHECK: ret
//
// CHECK: define internal void [[PRIVATES_MAP_FN:@.+]]([[PRIVATES_MAIN_TY]]* noalias, [[S_DOUBLE_TY]]** noalias, i32** noalias, [2 x [[S_DOUBLE_TY]]]** noalias, [2 x i32]** noalias, i32** noalias)
// CHECK: [[PRIVATES:%.+]] = load [[PRIVATES_MAIN_TY]]*, [[PRIVATES_MAIN_TY]]**
// CHECK: [[PRIV_S_VAR:%.+]] = getelementptr inbounds [[PRIVATES_MAIN_TY]], [[PRIVATES_MAIN_TY]]* [[PRIVATES]], i32 0, i32 0
// CHECK: [[ARG3:%.+]] = load [2 x [[S_DOUBLE_TY]]]**, [2 x [[S_DOUBLE_TY]]]*** %{{.+}},
// CHECK: store [2 x [[S_DOUBLE_TY]]]* [[PRIV_S_VAR]], [2 x [[S_DOUBLE_TY]]]** [[ARG3]],
// CHECK: [[PRIV_VAR:%.+]] = getelementptr inbounds [[PRIVATES_MAIN_TY]], [[PRIVATES_MAIN_TY]]* [[PRIVATES]], i32 0, i32 1
// CHECK: [[ARG1:%.+]] = load [[S_DOUBLE_TY]]**, [[S_DOUBLE_TY]]*** {{.+}},
// CHECK: store [[S_DOUBLE_TY]]* [[PRIV_VAR]], [[S_DOUBLE_TY]]** [[ARG1]],
// CHECK: [[PRIV_T_VAR:%.+]] = getelementptr inbounds [[PRIVATES_MAIN_TY]], [[PRIVATES_MAIN_TY]]* [[PRIVATES]], i32 0, i32 2
// CHECK: [[ARG2:%.+]] = load i32**, i32*** %{{.+}},
// CHECK: store i32* [[PRIV_T_VAR]], i32** [[ARG2]],
// CHECK: [[PRIV_VEC:%.+]] = getelementptr inbounds [[PRIVATES_MAIN_TY]], [[PRIVATES_MAIN_TY]]* [[PRIVATES]], i32 0, i32 3
// CHECK: [[ARG4:%.+]] = load [2 x i32]**, [2 x i32]*** %{{.+}},
// CHECK: store [2 x i32]* [[PRIV_VEC]], [2 x i32]** [[ARG4]],
// CHECK: [[PRIV_SIVAR:%.+]] = getelementptr inbounds [[PRIVATES_MAIN_TY]], [[PRIVATES_MAIN_TY]]* [[PRIVATES]], i32 0, i32 4
// CHECK: [[ARG5:%.+]] = load i{{[0-9]+}}**, i{{[0-9]+}}*** %{{.+}},
// CHECK: store i{{[0-9]+}}* [[PRIV_SIVAR]], i{{[0-9]+}}** [[ARG5]],
// CHECK: ret void
// CHECK: define internal i32 [[TASK_ENTRY]](i32, [[KMP_TASK_MAIN_TY]]* noalias)
// CHECK: [[PRIV_VAR_ADDR:%.+]] = alloca [[S_DOUBLE_TY]]*,
// CHECK: [[PRIV_T_VAR_ADDR:%.+]] = alloca i32*,
// CHECK: [[PRIV_S_ARR_ADDR:%.+]] = alloca [2 x [[S_DOUBLE_TY]]]*,
// CHECK: [[PRIV_VEC_ADDR:%.+]] = alloca [2 x i32]*,
// CHECK: [[PRIV_SIVAR_ADDR:%.+]] = alloca i32*,
// CHECK: store void (i8*, ...)* bitcast (void ([[PRIVATES_MAIN_TY]]*, [[S_DOUBLE_TY]]**, i32**, [2 x [[S_DOUBLE_TY]]]**, [2 x i32]**, i32**)* [[PRIVATES_MAP_FN]] to void (i8*, ...)*), void (i8*, ...)** [[MAP_FN_ADDR:%.+]],
// CHECK: [[MAP_FN:%.+]] = load void (i8*, ...)*, void (i8*, ...)** [[MAP_FN_ADDR]],
// CHECK: call void (i8*, ...) [[MAP_FN]](i8* %{{.+}}, [[S_DOUBLE_TY]]** [[PRIV_VAR_ADDR]], i32** [[PRIV_T_VAR_ADDR]], [2 x [[S_DOUBLE_TY]]]** [[PRIV_S_ARR_ADDR]], [2 x i32]** [[PRIV_VEC_ADDR]], i32** [[PRIV_SIVAR_ADDR]])
// CHECK: [[PRIV_VAR:%.+]] = load [[S_DOUBLE_TY]]*, [[S_DOUBLE_TY]]** [[PRIV_VAR_ADDR]],
// CHECK: [[PRIV_T_VAR:%.+]] = load i32*, i32** [[PRIV_T_VAR_ADDR]],
// CHECK: [[PRIV_S_ARR:%.+]] = load [2 x [[S_DOUBLE_TY]]]*, [2 x [[S_DOUBLE_TY]]]** [[PRIV_S_ARR_ADDR]],
// CHECK: [[PRIV_VEC:%.+]] = load [2 x i32]*, [2 x i32]** [[PRIV_VEC_ADDR]],
// CHECK: [[PRIV_SIVAR:%.+]] = load i32*, i32** [[PRIV_SIVAR_ADDR]],
// Privates actually are used.
// CHECK-DAG: [[PRIV_VAR]]
// CHECK-DAG: [[PRIV_T_VAR]]
// CHECK-DAG: [[PRIV_S_ARR]]
// CHECK-DAG: [[PRIV_VEC]]
// CHECK-DAG: [[PRIV_SIVAR]]
// CHECK: icmp ne i32 %{{.+}}, 0
// CHECK-NEXT: br i1
// CHECK: bitcast [[S_DOUBLE_TY]]* %{{.+}} to i8*
// CHECK: bitcast [[S_DOUBLE_TY]]* %{{.+}} to i8*
Change memcpy/memove/memset to have dest and source alignment attributes (Step 1). Summary: Upstream LLVM is changing the the prototypes of the @llvm.memcpy/memmove/memset intrinsics. This change updates the Clang tests for this change. The @llvm.memcpy/memmove/memset intrinsics currently have an explicit argument which is required to be a constant integer. It represents the alignment of the dest (and source), and so must be the minimum of the actual alignment of the two. This change removes the alignment argument in favour of placing the alignment attribute on the source and destination pointers of the memory intrinsic call. For example, code which used to read: call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* %src, i32 100, i32 4, i1 false) will now read call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 4 %dest, i8* align 4 %src, i32 100, i1 false) At this time the source and destination alignments must be the same (Step 1). Step 2 of the change, to be landed shortly, will relax that contraint and allow the source and destination to have different alignments. llvm-svn: 322964
2018-01-20 01:12:54 +08:00
// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align {{[0-9]+}} %
[OPENMP 4.5] Codegen for 'lastprivate' clauses in 'taskloop' directives. OpenMP 4.5 adds taskloop/taskloop simd directives. These directives allow to use lastprivate clause. Patch adds codegen for this clause. llvm-svn: 268618
2016-05-05 16:46:22 +08:00
// CHECK: load i32, i32* %
// CHECK: store i32 %{{.+}}, i32* %
// CHECK: getelementptr inbounds [2 x [[S_DOUBLE_TY]]], [2 x [[S_DOUBLE_TY]]]* %
// CHECK: phi [[S_DOUBLE_TY]]*
// CHECK: phi [[S_DOUBLE_TY]]*
Change memcpy/memove/memset to have dest and source alignment attributes (Step 1). Summary: Upstream LLVM is changing the the prototypes of the @llvm.memcpy/memmove/memset intrinsics. This change updates the Clang tests for this change. The @llvm.memcpy/memmove/memset intrinsics currently have an explicit argument which is required to be a constant integer. It represents the alignment of the dest (and source), and so must be the minimum of the actual alignment of the two. This change removes the alignment argument in favour of placing the alignment attribute on the source and destination pointers of the memory intrinsic call. For example, code which used to read: call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* %src, i32 100, i32 4, i1 false) will now read call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 4 %dest, i8* align 4 %src, i32 100, i1 false) At this time the source and destination alignments must be the same (Step 1). Step 2 of the change, to be landed shortly, will relax that contraint and allow the source and destination to have different alignments. llvm-svn: 322964
2018-01-20 01:12:54 +08:00
// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align {{[0-9]+}} %
[OPENMP 4.5] Codegen for 'lastprivate' clauses in 'taskloop' directives. OpenMP 4.5 adds taskloop/taskloop simd directives. These directives allow to use lastprivate clause. Patch adds codegen for this clause. llvm-svn: 268618
2016-05-05 16:46:22 +08:00
// CHECK: icmp eq [[S_DOUBLE_TY]]* %
// CHECK-NEXT: br i1
// CHECK: bitcast [2 x i32]* %{{.+}} to i8*
// CHECK: bitcast [2 x i32]* %{{.+}} to i8*
Change memcpy/memove/memset to have dest and source alignment attributes (Step 1). Summary: Upstream LLVM is changing the the prototypes of the @llvm.memcpy/memmove/memset intrinsics. This change updates the Clang tests for this change. The @llvm.memcpy/memmove/memset intrinsics currently have an explicit argument which is required to be a constant integer. It represents the alignment of the dest (and source), and so must be the minimum of the actual alignment of the two. This change removes the alignment argument in favour of placing the alignment attribute on the source and destination pointers of the memory intrinsic call. For example, code which used to read: call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* %src, i32 100, i32 4, i1 false) will now read call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 4 %dest, i8* align 4 %src, i32 100, i1 false) At this time the source and destination alignments must be the same (Step 1). Step 2 of the change, to be landed shortly, will relax that contraint and allow the source and destination to have different alignments. llvm-svn: 322964
2018-01-20 01:12:54 +08:00
// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align {{[0-9]+}} %
[OPENMP 4.5] Codegen for 'lastprivate' clauses in 'taskloop' directives. OpenMP 4.5 adds taskloop/taskloop simd directives. These directives allow to use lastprivate clause. Patch adds codegen for this clause. llvm-svn: 268618
2016-05-05 16:46:22 +08:00
// CHECK: load i32, i32* %
// CHECK: store i32 %{{.+}}, i32* %
// CHECK: br label
// CHECK: ret
// CHECK: define internal void [[MAIN_DUP]]([[KMP_TASK_MAIN_TY]]*, [[KMP_TASK_MAIN_TY]]*, i32)
// CHECK: getelementptr inbounds [[KMP_TASK_MAIN_TY]], [[KMP_TASK_MAIN_TY]]* %{{.+}}, i32 0, i32 0
[OPENMP 4.5] Fixed codegen for 'priority' and destructors in task-based directives. 'kmp_task_t' record type added a new field for 'priority' clause and changed the representation of pointer to destructors for privates used within loop-based directives. Old representation: typedef struct kmp_task { /* GEH: Shouldn't this be aligned somehow? */ void *shareds; /**< pointer to block of pointers to shared vars */ kmp_routine_entry_t routine; /**< pointer to routine to call for executing task */ kmp_int32 part_id; /**< part id for the task */ kmp_routine_entry_t destructors; /* pointer to function to invoke deconstructors of firstprivate C++ objects */ /* private vars */ } kmp_task_t; New representation: typedef struct kmp_task { /* GEH: Shouldn't this be aligned somehow? */ void *shareds; /**< pointer to block of pointers to shared vars */ kmp_routine_entry_t routine; /**< pointer to routine to call for executing task */ kmp_int32 part_id; /**< part id for the task */ kmp_cmplrdata_t data1; /* Two known optional additions: destructors and priority */ kmp_cmplrdata_t data2; /* Process destructors first, priority second */ /* future data */ /* private vars */ } kmp_task_t; Also excessive initialization of 'destructors' fields to 'null' was removed from codegen if it is known that no destructors shal be used. Currently a special bit is used in 'kmp_tasking_flags_t' bitfields ('destructors_thunk' bitfield). llvm-svn: 271201
2016-05-30 17:06:50 +08:00
// CHECK: getelementptr inbounds [[KMP_TASK_T_TY]], [[KMP_TASK_T_TY]]* %{{.+}}, i32 0, i32 8
[OPENMP 4.5] Codegen for 'lastprivate' clauses in 'taskloop' directives. OpenMP 4.5 adds taskloop/taskloop simd directives. These directives allow to use lastprivate clause. Patch adds codegen for this clause. llvm-svn: 268618
2016-05-05 16:46:22 +08:00
// CHECK: load i32, i32* %
// CHECK: store i32 %{{.+}}, i32* %
// CHECK: getelementptr inbounds [[KMP_TASK_MAIN_TY]], [[KMP_TASK_MAIN_TY]]* %{{.+}}, i32 0, i32 1
// CHECK: getelementptr inbounds [[PRIVATES_MAIN_TY]], [[PRIVATES_MAIN_TY]]* %{{.+}}, i32 0, i32 0
// CHECK: getelementptr inbounds [2 x [[S_DOUBLE_TY]]], [2 x [[S_DOUBLE_TY]]]* %{{.+}}, i32 0, i32 0
// CHECK: getelementptr inbounds [[S_DOUBLE_TY]], [[S_DOUBLE_TY]]* %{{.+}}, i64 2
// CHECK: br label %
// CHECK: phi [[S_DOUBLE_TY]]*
// CHECK: call {{.*}} [[S_DOUBLE_TY_CONSTR]]([[S_DOUBLE_TY]]*
// CHECK: getelementptr inbounds [[S_DOUBLE_TY]], [[S_DOUBLE_TY]]* %{{.+}}, i64 1
// CHECK: icmp eq [[S_DOUBLE_TY]]* %
// CHECK: br i1 %
// CHECK: getelementptr inbounds [[PRIVATES_MAIN_TY]], [[PRIVATES_MAIN_TY]]* %{{.+}}, i32 0, i32 1
// CHECK: call {{.*}} [[S_DOUBLE_TY_CONSTR]]([[S_DOUBLE_TY]]*
// CHECK: ret void
// CHECK: define internal i32 [[DESTRUCTORS]](i32, [[KMP_TASK_MAIN_TY]]* noalias)
// CHECK: [[PRIVATES:%.+]] = getelementptr inbounds [[KMP_TASK_MAIN_TY]], [[KMP_TASK_MAIN_TY]]* [[RES_KMP_TASK:%.+]], i{{[0-9]+}} 0, i{{[0-9]+}} 1
// CHECK: [[PRIVATE_S_ARR_REF:%.+]] = getelementptr inbounds [[PRIVATES_MAIN_TY]], [[PRIVATES_MAIN_TY]]* [[PRIVATES]], i{{.+}} 0, i{{.+}} 0
// CHECK: [[PRIVATE_VAR_REF:%.+]] = getelementptr inbounds [[PRIVATES_MAIN_TY]], [[PRIVATES_MAIN_TY]]* [[PRIVATES]], i{{.+}} 0, i{{.+}} 1
// CHECK: call {{.*}} [[S_DOUBLE_TY_DESTR]]([[S_DOUBLE_TY]]* [[PRIVATE_VAR_REF]])
// CHECK: getelementptr inbounds [2 x [[S_DOUBLE_TY]]], [2 x [[S_DOUBLE_TY]]]* [[PRIVATE_S_ARR_REF]], i{{.+}} 0, i{{.+}} 0
// CHECK: getelementptr inbounds [[S_DOUBLE_TY]], [[S_DOUBLE_TY]]* %{{.+}}, i{{.+}} 2
// CHECK: [[PRIVATE_S_ARR_ELEM_REF:%.+]] = getelementptr inbounds [[S_DOUBLE_TY]], [[S_DOUBLE_TY]]* %{{.+}}, i{{.+}} -1
// CHECK: call {{.*}} [[S_DOUBLE_TY_DESTR]]([[S_DOUBLE_TY]]* [[PRIVATE_S_ARR_ELEM_REF]])
// CHECK: icmp eq
// CHECK: br i1
// CHECK: ret i32
// CHECK: define {{.*}} i{{[0-9]+}} [[TMAIN_INT]]()
// CHECK: alloca [[S_INT_TY]],
// CHECK: [[TEST:%.+]] = alloca [[S_INT_TY]],
// CHECK: [[T_VAR_ADDR:%.+]] = alloca i32, align 128
// CHECK: [[VEC_ADDR:%.+]] = alloca [2 x i32],
// CHECK: [[S_ARR_ADDR:%.+]] = alloca [2 x [[S_INT_TY]]],
// CHECK: [[VAR_ADDR:%.+]] = alloca [[S_INT_TY]],
// CHECK: [[GTID:%.+]] = call i32 @__kmpc_global_thread_num([[LOC:%.+]])
Revert "[CodeGenCXX] Treat 'this' as noalias in constructors" This reverts commit https://reviews.llvm.org/rL344150 which causes MachineOutliner related failures on the ppc64le multistage buildbot. llvm-svn: 344526
2018-10-15 23:43:00 +08:00
// CHECK: call {{.*}} [[S_INT_TY_CONSTR:@.+]]([[S_INT_TY]]* [[TEST]])
[OPENMP 4.5] Codegen for 'lastprivate' clauses in 'taskloop' directives. OpenMP 4.5 adds taskloop/taskloop simd directives. These directives allow to use lastprivate clause. Patch adds codegen for this clause. llvm-svn: 268618
2016-05-05 16:46:22 +08:00
// Store original variables in capture struct.
// CHECK: [[VEC_REF:%.+]] = getelementptr inbounds [[CAP_TMAIN_TY]], [[CAP_TMAIN_TY]]* %{{.+}}, i{{[0-9]+}} 0, i{{[0-9]+}} 0
// CHECK: store [2 x i32]* [[VEC_ADDR]], [2 x i32]** [[VEC_REF]],
// CHECK: [[T_VAR_REF:%.+]] = getelementptr inbounds [[CAP_TMAIN_TY]], [[CAP_TMAIN_TY]]* %{{.+}}, i{{[0-9]+}} 0, i{{[0-9]+}} 1
// CHECK: store i32* [[T_VAR_ADDR]], i32** [[T_VAR_REF]],
// CHECK: [[S_ARR_REF:%.+]] = getelementptr inbounds [[CAP_TMAIN_TY]], [[CAP_TMAIN_TY]]* %{{.+}}, i{{[0-9]+}} 0, i{{[0-9]+}} 2
// CHECK: store [2 x [[S_INT_TY]]]* [[S_ARR_ADDR]], [2 x [[S_INT_TY]]]** [[S_ARR_REF]],
// CHECK: [[VAR_REF:%.+]] = getelementptr inbounds [[CAP_TMAIN_TY]], [[CAP_TMAIN_TY]]* %{{.+}}, i{{[0-9]+}} 0, i{{[0-9]+}} 3
// CHECK: store [[S_INT_TY]]* [[VAR_ADDR]], [[S_INT_TY]]** [[VAR_REF]],
// Allocate task.
// Returns struct kmp_task_t {
// [[KMP_TASK_T_TY]] task_data;
// [[KMP_TASK_TMAIN_TY]] privates;
// };
[OPENMP 4.0] Fixed codegen for destructors in task-based directives. If private variables require destructors call at the deletion of the task, additional flag in task flags must be set. Patch fixes this problem. llvm-svn: 269039
2016-05-10 18:36:51 +08:00
// CHECK: [[RES:%.+]] = call i8* @__kmpc_omp_task_alloc([[LOC]], i32 [[GTID]], i32 9, i64 256, i64 32, i32 (i32, i8*)* bitcast (i32 (i32, [[KMP_TASK_TMAIN_TY]]*)* [[TASK_ENTRY:@[^ ]+]] to i32 (i32, i8*)*))
[OPENMP 4.5] Codegen for 'lastprivate' clauses in 'taskloop' directives. OpenMP 4.5 adds taskloop/taskloop simd directives. These directives allow to use lastprivate clause. Patch adds codegen for this clause. llvm-svn: 268618
2016-05-05 16:46:22 +08:00
// CHECK: [[RES_KMP_TASK:%.+]] = bitcast i8* [[RES]] to [[KMP_TASK_TMAIN_TY]]*
// Fill kmp_task_t->shareds by copying from original capture argument.
// CHECK: [[TASK:%.+]] = getelementptr inbounds [[KMP_TASK_TMAIN_TY]], [[KMP_TASK_TMAIN_TY]]* [[RES_KMP_TASK]], i{{[0-9]+}} 0, i{{[0-9]+}} 0
// CHECK: [[SHAREDS_REF_ADDR:%.+]] = getelementptr inbounds [[KMP_TASK_T_TY]], [[KMP_TASK_T_TY]]* [[TASK]], i{{[0-9]+}} 0, i{{[0-9]+}} 0
// CHECK: [[SHAREDS_REF:%.+]] = load i8*, i8** [[SHAREDS_REF_ADDR]],
// CHECK: [[CAPTURES_ADDR:%.+]] = bitcast [[CAP_TMAIN_TY]]* %{{.+}} to i8*
Change memcpy/memove/memset to have dest and source alignment attributes (Step 1). Summary: Upstream LLVM is changing the the prototypes of the @llvm.memcpy/memmove/memset intrinsics. This change updates the Clang tests for this change. The @llvm.memcpy/memmove/memset intrinsics currently have an explicit argument which is required to be a constant integer. It represents the alignment of the dest (and source), and so must be the minimum of the actual alignment of the two. This change removes the alignment argument in favour of placing the alignment attribute on the source and destination pointers of the memory intrinsic call. For example, code which used to read: call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* %src, i32 100, i32 4, i1 false) will now read call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 4 %dest, i8* align 4 %src, i32 100, i1 false) At this time the source and destination alignments must be the same (Step 1). Step 2 of the change, to be landed shortly, will relax that contraint and allow the source and destination to have different alignments. llvm-svn: 322964
2018-01-20 01:12:54 +08:00
// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 8 [[SHAREDS_REF]], i8* align 8 [[CAPTURES_ADDR]], i64 32, i1 false)
[OPENMP 4.5] Codegen for 'lastprivate' clauses in 'taskloop' directives. OpenMP 4.5 adds taskloop/taskloop simd directives. These directives allow to use lastprivate clause. Patch adds codegen for this clause. llvm-svn: 268618
2016-05-05 16:46:22 +08:00
// Initialize kmp_task_t->privates with default values (no init for simple types, default constructors for classes).
// CHECK: [[PRIVATES:%.+]] = getelementptr inbounds [[KMP_TASK_TMAIN_TY]], [[KMP_TASK_TMAIN_TY]]* [[RES_KMP_TASK]], i{{[0-9]+}} 0, i{{[0-9]+}} 2
// t_var;
// vec;
// Constructors for s_arr and var.
// a_arr;
// CHECK: [[PRIVATE_S_ARR_REF:%.+]] = getelementptr inbounds [[PRIVATES_TMAIN_TY]], [[PRIVATES_TMAIN_TY]]* [[PRIVATES]], i{{[0-9]+}} 0, i{{[0-9]+}} 2
// CHECK: getelementptr inbounds [2 x [[S_INT_TY]]], [2 x [[S_INT_TY]]]* [[PRIVATE_S_ARR_REF]], i{{.+}} 0, i{{.+}} 0
// CHECK: getelementptr inbounds [[S_INT_TY]], [[S_INT_TY]]* %{{.+}}, i{{.+}} 2
Revert "[CodeGenCXX] Treat 'this' as noalias in constructors" This reverts commit https://reviews.llvm.org/rL344150 which causes MachineOutliner related failures on the ppc64le multistage buildbot. llvm-svn: 344526
2018-10-15 23:43:00 +08:00
// CHECK: call {{.*}} [[S_INT_TY_CONSTR]]([[S_INT_TY]]* [[S_ARR_CUR:%[^,]+]])
[OPENMP 4.5] Codegen for 'lastprivate' clauses in 'taskloop' directives. OpenMP 4.5 adds taskloop/taskloop simd directives. These directives allow to use lastprivate clause. Patch adds codegen for this clause. llvm-svn: 268618
2016-05-05 16:46:22 +08:00
// CHECK: getelementptr inbounds [[S_INT_TY]], [[S_INT_TY]]* [[S_ARR_CUR]], i{{.+}} 1
// CHECK: icmp eq
// CHECK: br i1
// var;
// CHECK: [[PRIVATE_VAR_REF:%.+]] = getelementptr inbounds [[PRIVATES_TMAIN_TY]], [[PRIVATES_TMAIN_TY]]* [[PRIVATES]], i{{.+}} 0, i{{.+}} 3
Revert "[CodeGenCXX] Treat 'this' as noalias in constructors" This reverts commit https://reviews.llvm.org/rL344150 which causes MachineOutliner related failures on the ppc64le multistage buildbot. llvm-svn: 344526
2018-10-15 23:43:00 +08:00
// CHECK: call {{.*}} [[S_INT_TY_CONSTR]]([[S_INT_TY]]* [[PRIVATE_VAR_REF]])
[OPENMP 4.5] Codegen for 'lastprivate' clauses in 'taskloop' directives. OpenMP 4.5 adds taskloop/taskloop simd directives. These directives allow to use lastprivate clause. Patch adds codegen for this clause. llvm-svn: 268618
2016-05-05 16:46:22 +08:00
// Provide pointer to destructor function, which will destroy private variables at the end of the task.
// CHECK: [[DESTRUCTORS_REF:%.+]] = getelementptr inbounds [[KMP_TASK_T_TY]], [[KMP_TASK_T_TY]]* [[TASK]], i{{.+}} 0, i{{.+}} 3
[OPENMP 4.5] Fixed codegen for 'priority' and destructors in task-based directives. 'kmp_task_t' record type added a new field for 'priority' clause and changed the representation of pointer to destructors for privates used within loop-based directives. Old representation: typedef struct kmp_task { /* GEH: Shouldn't this be aligned somehow? */ void *shareds; /**< pointer to block of pointers to shared vars */ kmp_routine_entry_t routine; /**< pointer to routine to call for executing task */ kmp_int32 part_id; /**< part id for the task */ kmp_routine_entry_t destructors; /* pointer to function to invoke deconstructors of firstprivate C++ objects */ /* private vars */ } kmp_task_t; New representation: typedef struct kmp_task { /* GEH: Shouldn't this be aligned somehow? */ void *shareds; /**< pointer to block of pointers to shared vars */ kmp_routine_entry_t routine; /**< pointer to routine to call for executing task */ kmp_int32 part_id; /**< part id for the task */ kmp_cmplrdata_t data1; /* Two known optional additions: destructors and priority */ kmp_cmplrdata_t data2; /* Process destructors first, priority second */ /* future data */ /* private vars */ } kmp_task_t; Also excessive initialization of 'destructors' fields to 'null' was removed from codegen if it is known that no destructors shal be used. Currently a special bit is used in 'kmp_tasking_flags_t' bitfields ('destructors_thunk' bitfield). llvm-svn: 271201
2016-05-30 17:06:50 +08:00
// CHECK: [[DESTRUCTORS_PTR:%.+]] = bitcast %union{{.+}}* [[DESTRUCTORS_REF]] to i32 (i32, i8*)**
// CHECK: store i32 (i32, i8*)* bitcast (i32 (i32, [[KMP_TASK_TMAIN_TY]]*)* [[DESTRUCTORS:@.+]] to i32 (i32, i8*)*), i32 (i32, i8*)** [[DESTRUCTORS_PTR]],
[OPENMP 4.5] Codegen for 'lastprivate' clauses in 'taskloop' directives. OpenMP 4.5 adds taskloop/taskloop simd directives. These directives allow to use lastprivate clause. Patch adds codegen for this clause. llvm-svn: 268618
2016-05-05 16:46:22 +08:00
// Start task.
Do not always request an implicit taskgroup region inside the kmpc_taskloop function Summary: For the following code: ``` int i; #pragma omp taskloop for (i = 0; i < 100; ++i) {} #pragma omp taskloop nogroup for (i = 0; i < 100; ++i) {} ``` Clang emits the following LLVM IR: ``` ... call void @__kmpc_taskgroup(%struct.ident_t* @0, i32 %0) %2 = call i8* @__kmpc_omp_task_alloc(%struct.ident_t* @0, i32 %0, i32 1, i64 80, i64 8, i32 (i32, i8*)* bitcast (i32 (i32, %struct.kmp_task_t_with_privates*)* @.omp_task_entry. to i32 (i32, i8*)*)) ... call void @__kmpc_taskloop(%struct.ident_t* @0, i32 %0, i8* %2, i32 1, i64* %8, i64* %9, i64 %13, i32 0, i32 0, i64 0, i8* null) call void @__kmpc_end_taskgroup(%struct.ident_t* @0, i32 %0) ... %15 = call i8* @__kmpc_omp_task_alloc(%struct.ident_t* @0, i32 %0, i32 1, i64 80, i64 8, i32 (i32, i8*)* bitcast (i32 (i32, %struct.kmp_task_t_with_privates.1*)* @.omp_task_entry..2 to i32 (i32, i8*)*)) ... call void @__kmpc_taskloop(%struct.ident_t* @0, i32 %0, i8* %15, i32 1, i64* %21, i64* %22, i64 %26, i32 0, i32 0, i64 0, i8* null) ``` The first set of instructions corresponds to the first taskloop construct. It is important to note that the implicit taskgroup region associated with the taskloop construct has been materialized in our IR: the `__kmpc_taskloop` occurs inside a taskgroup region. Note also that this taskgroup region does not exist in our second taskloop because we are using the `nogroup` clause. The issue here is the 4th argument of the kmpc_taskloop call, starting from the end, is always a zero. Checking the LLVM OpenMP RT implementation, we see that this argument corresponds to the nogroup parameter: ``` void __kmpc_taskloop(ident_t *loc, int gtid, kmp_task_t *task, int if_val, kmp_uint64 *lb, kmp_uint64 *ub, kmp_int64 st, int nogroup, int sched, kmp_uint64 grainsize, void *task_dup); ``` So basically we always tell to the RT to do another taskgroup region. For the first taskloop, this means that we create two taskgroup regions. For the second example, it means that despite the fact we had a nogroup clause we are going to have a taskgroup region, so we unnecessary wait until all descendant tasks have been executed. Reviewers: ABataev Reviewed By: ABataev Subscribers: rogfer01, cfe-commits Differential Revision: https://reviews.llvm.org/D53636 llvm-svn: 345180
2018-10-25 03:06:37 +08:00
// CHECK: call void @__kmpc_taskloop([[LOC]], i32 [[GTID]], i8* [[RES]], i32 1, i64* %{{.+}}, i64* %{{.+}}, i64 %{{.+}}, i32 1, i32 0, i64 0, i8* bitcast (void ([[KMP_TASK_TMAIN_TY]]*, [[KMP_TASK_TMAIN_TY]]*, i32)* [[TMAIN_DUP:@.+]] to i8*))
[OPENMP 4.5] Codegen for 'lastprivate' clauses in 'taskloop' directives. OpenMP 4.5 adds taskloop/taskloop simd directives. These directives allow to use lastprivate clause. Patch adds codegen for this clause. llvm-svn: 268618
2016-05-05 16:46:22 +08:00
// No destructors must be called for private copies of s_arr and var.
// CHECK-NOT: getelementptr inbounds [[PRIVATES_TMAIN_TY]], [[PRIVATES_TMAIN_TY]]* [[PRIVATES]], i{{.+}} 0, i{{.+}} 2
// CHECK-NOT: getelementptr inbounds [[PRIVATES_TMAIN_TY]], [[PRIVATES_TMAIN_TY]]* [[PRIVATES]], i{{.+}} 0, i{{.+}} 3
// CHECK: call void [[S_INT_TY_DESTR:@.+]]([[S_INT_TY]]*
// CHECK-NOT: getelementptr inbounds [[PRIVATES_TMAIN_TY]], [[PRIVATES_TMAIN_TY]]* [[PRIVATES]], i{{.+}} 0, i{{.+}} 2
// CHECK-NOT: getelementptr inbounds [[PRIVATES_TMAIN_TY]], [[PRIVATES_TMAIN_TY]]* [[PRIVATES]], i{{.+}} 0, i{{.+}} 3
// CHECK: ret
//
// CHECK: define internal void [[PRIVATES_MAP_FN:@.+]]([[PRIVATES_TMAIN_TY]]* noalias, i32** noalias, [2 x i32]** noalias, [2 x [[S_INT_TY]]]** noalias, [[S_INT_TY]]** noalias)
// CHECK: [[PRIVATES:%.+]] = load [[PRIVATES_TMAIN_TY]]*, [[PRIVATES_TMAIN_TY]]**
// CHECK: [[PRIV_T_VAR:%.+]] = getelementptr inbounds [[PRIVATES_TMAIN_TY]], [[PRIVATES_TMAIN_TY]]* [[PRIVATES]], i32 0, i32 0
// CHECK: [[ARG1:%.+]] = load i32**, i32*** %{{.+}},
// CHECK: store i32* [[PRIV_T_VAR]], i32** [[ARG1]],
// CHECK: [[PRIV_VEC:%.+]] = getelementptr inbounds [[PRIVATES_TMAIN_TY]], [[PRIVATES_TMAIN_TY]]* [[PRIVATES]], i32 0, i32 1
// CHECK: [[ARG2:%.+]] = load [2 x i32]**, [2 x i32]*** %{{.+}},
// CHECK: store [2 x i32]* [[PRIV_VEC]], [2 x i32]** [[ARG2]],
// CHECK: [[PRIV_S_VAR:%.+]] = getelementptr inbounds [[PRIVATES_TMAIN_TY]], [[PRIVATES_TMAIN_TY]]* [[PRIVATES]], i32 0, i32 2
// CHECK: [[ARG3:%.+]] = load [2 x [[S_INT_TY]]]**, [2 x [[S_INT_TY]]]*** %{{.+}},
// CHECK: store [2 x [[S_INT_TY]]]* [[PRIV_S_VAR]], [2 x [[S_INT_TY]]]** [[ARG3]],
// CHECK: [[PRIV_VAR:%.+]] = getelementptr inbounds [[PRIVATES_TMAIN_TY]], [[PRIVATES_TMAIN_TY]]* [[PRIVATES]], i32 0, i32 3
// CHECK: [[ARG4:%.+]] = load [[S_INT_TY]]**, [[S_INT_TY]]*** {{.+}},
// CHECK: store [[S_INT_TY]]* [[PRIV_VAR]], [[S_INT_TY]]** [[ARG4]],
// CHECK: ret void
// CHECK: define internal i32 [[TASK_ENTRY]](i32, [[KMP_TASK_TMAIN_TY]]* noalias)
// CHECK: alloca i32*,
// CHECK-DAG: [[PRIV_T_VAR_ADDR:%.+]] = alloca i32*,
// CHECK-DAG: [[PRIV_VEC_ADDR:%.+]] = alloca [2 x i32]*,
// CHECK-DAG: [[PRIV_S_ARR_ADDR:%.+]] = alloca [2 x [[S_INT_TY]]]*,
// CHECK-DAG: [[PRIV_VAR_ADDR:%.+]] = alloca [[S_INT_TY]]*,
// CHECK: store void (i8*, ...)* bitcast (void ([[PRIVATES_TMAIN_TY]]*, i32**, [2 x i32]**, [2 x [[S_INT_TY]]]**, [[S_INT_TY]]**)* [[PRIVATES_MAP_FN]] to void (i8*, ...)*), void (i8*, ...)** [[MAP_FN_ADDR:%.+]],
// CHECK: [[MAP_FN:%.+]] = load void (i8*, ...)*, void (i8*, ...)** [[MAP_FN_ADDR]],
// CHECK: call void (i8*, ...) [[MAP_FN]](i8* %{{.+}}, i32** [[PRIV_T_VAR_ADDR]], [2 x i32]** [[PRIV_VEC_ADDR]], [2 x [[S_INT_TY]]]** [[PRIV_S_ARR_ADDR]], [[S_INT_TY]]** [[PRIV_VAR_ADDR]])
// CHECK: [[PRIV_T_VAR:%.+]] = load i32*, i32** [[PRIV_T_VAR_ADDR]],
// CHECK: [[PRIV_VEC:%.+]] = load [2 x i32]*, [2 x i32]** [[PRIV_VEC_ADDR]],
// CHECK: [[PRIV_S_ARR:%.+]] = load [2 x [[S_INT_TY]]]*, [2 x [[S_INT_TY]]]** [[PRIV_S_ARR_ADDR]],
// CHECK: [[PRIV_VAR:%.+]] = load [[S_INT_TY]]*, [[S_INT_TY]]** [[PRIV_VAR_ADDR]],
// Privates actually are used.
// CHECK-DAG: [[PRIV_VAR]]
// CHECK-DAG: [[PRIV_T_VAR]]
// CHECK-DAG: [[PRIV_S_ARR]]
// CHECK-DAG: [[PRIV_VEC]]
// CHECK: icmp ne i32 %{{.+}}, 0
// CHECK-NEXT: br i1
// CHECK: load i32, i32* %
// CHECK: store i32 %{{.+}}, i32* %
// CHECK: bitcast [2 x i32]* %{{.+}} to i8*
// CHECK: bitcast [2 x i32]* %{{.+}} to i8*
Change memcpy/memove/memset to have dest and source alignment attributes (Step 1). Summary: Upstream LLVM is changing the the prototypes of the @llvm.memcpy/memmove/memset intrinsics. This change updates the Clang tests for this change. The @llvm.memcpy/memmove/memset intrinsics currently have an explicit argument which is required to be a constant integer. It represents the alignment of the dest (and source), and so must be the minimum of the actual alignment of the two. This change removes the alignment argument in favour of placing the alignment attribute on the source and destination pointers of the memory intrinsic call. For example, code which used to read: call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* %src, i32 100, i32 4, i1 false) will now read call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 4 %dest, i8* align 4 %src, i32 100, i1 false) At this time the source and destination alignments must be the same (Step 1). Step 2 of the change, to be landed shortly, will relax that contraint and allow the source and destination to have different alignments. llvm-svn: 322964
2018-01-20 01:12:54 +08:00
// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align {{[0-9]+}} %
[OPENMP 4.5] Codegen for 'lastprivate' clauses in 'taskloop' directives. OpenMP 4.5 adds taskloop/taskloop simd directives. These directives allow to use lastprivate clause. Patch adds codegen for this clause. llvm-svn: 268618
2016-05-05 16:46:22 +08:00
// CHECK: getelementptr inbounds [2 x [[S_INT_TY]]], [2 x [[S_INT_TY]]]* %
// CHECK: phi [[S_INT_TY]]*
// CHECK: phi [[S_INT_TY]]*
Change memcpy/memove/memset to have dest and source alignment attributes (Step 1). Summary: Upstream LLVM is changing the the prototypes of the @llvm.memcpy/memmove/memset intrinsics. This change updates the Clang tests for this change. The @llvm.memcpy/memmove/memset intrinsics currently have an explicit argument which is required to be a constant integer. It represents the alignment of the dest (and source), and so must be the minimum of the actual alignment of the two. This change removes the alignment argument in favour of placing the alignment attribute on the source and destination pointers of the memory intrinsic call. For example, code which used to read: call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* %src, i32 100, i32 4, i1 false) will now read call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 4 %dest, i8* align 4 %src, i32 100, i1 false) At this time the source and destination alignments must be the same (Step 1). Step 2 of the change, to be landed shortly, will relax that contraint and allow the source and destination to have different alignments. llvm-svn: 322964
2018-01-20 01:12:54 +08:00
// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align {{[0-9]+}} %
[OPENMP 4.5] Codegen for 'lastprivate' clauses in 'taskloop' directives. OpenMP 4.5 adds taskloop/taskloop simd directives. These directives allow to use lastprivate clause. Patch adds codegen for this clause. llvm-svn: 268618
2016-05-05 16:46:22 +08:00
// CHECK: icmp eq [[S_INT_TY]]* %
// CHECK-NEXT: br i1
// CHECK: bitcast [[S_INT_TY]]* %{{.+}} to i8*
// CHECK: bitcast [[S_INT_TY]]* %{{.+}} to i8*
Change memcpy/memove/memset to have dest and source alignment attributes (Step 1). Summary: Upstream LLVM is changing the the prototypes of the @llvm.memcpy/memmove/memset intrinsics. This change updates the Clang tests for this change. The @llvm.memcpy/memmove/memset intrinsics currently have an explicit argument which is required to be a constant integer. It represents the alignment of the dest (and source), and so must be the minimum of the actual alignment of the two. This change removes the alignment argument in favour of placing the alignment attribute on the source and destination pointers of the memory intrinsic call. For example, code which used to read: call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* %src, i32 100, i32 4, i1 false) will now read call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 4 %dest, i8* align 4 %src, i32 100, i1 false) At this time the source and destination alignments must be the same (Step 1). Step 2 of the change, to be landed shortly, will relax that contraint and allow the source and destination to have different alignments. llvm-svn: 322964
2018-01-20 01:12:54 +08:00
// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align {{[0-9]+}} %
[OPENMP 4.5] Codegen for 'lastprivate' clauses in 'taskloop' directives. OpenMP 4.5 adds taskloop/taskloop simd directives. These directives allow to use lastprivate clause. Patch adds codegen for this clause. llvm-svn: 268618
2016-05-05 16:46:22 +08:00
// CHECK: br label
// CHECK: ret
// CHECK: define internal void [[TMAIN_DUP]]([[KMP_TASK_TMAIN_TY]]*, [[KMP_TASK_TMAIN_TY]]*, i32)
// CHECK: getelementptr inbounds [[KMP_TASK_TMAIN_TY]], [[KMP_TASK_TMAIN_TY]]* %{{.+}}, i32 0, i32 0
[OPENMP 4.5] Fixed codegen for 'priority' and destructors in task-based directives. 'kmp_task_t' record type added a new field for 'priority' clause and changed the representation of pointer to destructors for privates used within loop-based directives. Old representation: typedef struct kmp_task { /* GEH: Shouldn't this be aligned somehow? */ void *shareds; /**< pointer to block of pointers to shared vars */ kmp_routine_entry_t routine; /**< pointer to routine to call for executing task */ kmp_int32 part_id; /**< part id for the task */ kmp_routine_entry_t destructors; /* pointer to function to invoke deconstructors of firstprivate C++ objects */ /* private vars */ } kmp_task_t; New representation: typedef struct kmp_task { /* GEH: Shouldn't this be aligned somehow? */ void *shareds; /**< pointer to block of pointers to shared vars */ kmp_routine_entry_t routine; /**< pointer to routine to call for executing task */ kmp_int32 part_id; /**< part id for the task */ kmp_cmplrdata_t data1; /* Two known optional additions: destructors and priority */ kmp_cmplrdata_t data2; /* Process destructors first, priority second */ /* future data */ /* private vars */ } kmp_task_t; Also excessive initialization of 'destructors' fields to 'null' was removed from codegen if it is known that no destructors shal be used. Currently a special bit is used in 'kmp_tasking_flags_t' bitfields ('destructors_thunk' bitfield). llvm-svn: 271201
2016-05-30 17:06:50 +08:00
// CHECK: getelementptr inbounds [[KMP_TASK_T_TY]], [[KMP_TASK_T_TY]]* %{{.+}}, i32 0, i32 8
[OPENMP 4.5] Codegen for 'lastprivate' clauses in 'taskloop' directives. OpenMP 4.5 adds taskloop/taskloop simd directives. These directives allow to use lastprivate clause. Patch adds codegen for this clause. llvm-svn: 268618
2016-05-05 16:46:22 +08:00
// CHECK: load i32, i32* %
// CHECK: store i32 %{{.+}}, i32* %
// CHECK: getelementptr inbounds [[KMP_TASK_TMAIN_TY]], [[KMP_TASK_TMAIN_TY]]* %{{.+}}, i32 0, i32 2
// CHECK: getelementptr inbounds [[PRIVATES_TMAIN_TY]], [[PRIVATES_TMAIN_TY]]* %{{.+}}, i32 0, i32 2
// CHECK: getelementptr inbounds [2 x [[S_INT_TY]]], [2 x [[S_INT_TY]]]* %{{.+}}, i32 0, i32 0
// CHECK: getelementptr inbounds [[S_INT_TY]], [[S_INT_TY]]* %{{.+}}, i64 2
// CHECK: br label %
// CHECK: phi [[S_INT_TY]]*
// CHECK: call {{.*}} [[S_INT_TY_CONSTR]]([[S_INT_TY]]*
// CHECK: getelementptr inbounds [[S_INT_TY]], [[S_INT_TY]]* %{{.+}}, i64 1
// CHECK: icmp eq [[S_INT_TY]]* %
// CHECK: br i1 %
// CHECK: getelementptr inbounds [[PRIVATES_TMAIN_TY]], [[PRIVATES_TMAIN_TY]]* %{{.+}}, i32 0, i32 3
// CHECK: call {{.*}} [[S_INT_TY_CONSTR]]([[S_INT_TY]]*
// CHECK: ret void
// CHECK: define internal i32 [[DESTRUCTORS]](i32, [[KMP_TASK_TMAIN_TY]]* noalias)
// CHECK: [[PRIVATES:%.+]] = getelementptr inbounds [[KMP_TASK_TMAIN_TY]], [[KMP_TASK_TMAIN_TY]]* [[RES_KMP_TASK:%.+]], i{{[0-9]+}} 0, i{{[0-9]+}} 2
// CHECK: [[PRIVATE_S_ARR_REF:%.+]] = getelementptr inbounds [[PRIVATES_TMAIN_TY]], [[PRIVATES_TMAIN_TY]]* [[PRIVATES]], i{{.+}} 0, i{{.+}} 2
// CHECK: [[PRIVATE_VAR_REF:%.+]] = getelementptr inbounds [[PRIVATES_TMAIN_TY]], [[PRIVATES_TMAIN_TY]]* [[PRIVATES]], i{{.+}} 0, i{{.+}} 3
// CHECK: call void [[S_INT_TY_DESTR]]([[S_INT_TY]]* [[PRIVATE_VAR_REF]])
// CHECK: getelementptr inbounds [2 x [[S_INT_TY]]], [2 x [[S_INT_TY]]]* [[PRIVATE_S_ARR_REF]], i{{.+}} 0, i{{.+}} 0
// CHECK: getelementptr inbounds [[S_INT_TY]], [[S_INT_TY]]* %{{.+}}, i{{.+}} 2
// CHECK: [[PRIVATE_S_ARR_ELEM_REF:%.+]] = getelementptr inbounds [[S_INT_TY]], [[S_INT_TY]]* %{{.+}}, i{{.+}} -1
// CHECK: call void [[S_INT_TY_DESTR]]([[S_INT_TY]]* [[PRIVATE_S_ARR_ELEM_REF]])
// CHECK: icmp eq
// CHECK: br i1
// CHECK: ret i32
#endif
#else
// ARRAY-LABEL: array_func
struct St {
int a, b;
St() : a(0), b(0) {}
St(const St &) {}
~St() {}
};
void array_func(int n, float a[n], St s[2]) {
// ARRAY: call i8* @__kmpc_omp_task_alloc(
// ARRAY: call void @__kmpc_taskloop(
// ARRAY: store float** %{{.+}}, float*** %{{.+}},
// ARRAY: store %struct.St** %{{.+}}, %struct.St*** %{{.+}},
// ARRAY: icmp ne i32 %{{.+}}, 0
// ARRAY: store float* %{{.+}}, float** %{{.+}},
// ARRAY: store %struct.St* %{{.+}}, %struct.St** %{{.+}},
#pragma omp taskloop simd lastprivate(a, s)
for (int i = 0; i < 10; ++i)
;
}
#endif