[OPENMP] Fix for '#pragma omp task' codegen.

Internal task structure must be generated like
typedef struct kmp_task {
  void *              shareds;
  kmp_routine_entry_t routine;
  kmp_int32           part_id;
  kmp_routine_entry_t destructors;
} kmp_task_t;

struct kmp_task_t_with_privates {
  kmp_task_t task_data;
  .kmp_private. privates;
};
to avoid possible additional alignment bytes in first fields (shareds, routine, part_id and destructors). Runtime library is not aware of such kind additional alignment bytes.

llvm-svn: 237561

2 files changed, 42 insertions, 44 deletions

diff --git a/clang/test/OpenMP/task_firstprivate_codegen.cpp b/clang/test/OpenMP/task_firstprivate_codegen.cpp
index 204a852dd8d..7091cc4e081 100644
--- a/clang/test/OpenMP/task_firstprivate_codegen.cpp
+++ b/clang/test/OpenMP/task_firstprivate_codegen.cpp
@@ -22,14 +22,15 @@ struct S {
 
 volatile double g;
 
+// CHECK-DAG: [[KMP_TASK_T_TY:%.+]] = type { i8*, i32 (i32, i8*)*, i32, i32 (i32, i8*)* }
 // CHECK-DAG: [[S_DOUBLE_TY:%.+]] = type { double }
 // CHECK-DAG: [[CAP_MAIN_TY:%.+]] = type { [2 x i32]*, i32*, [2 x [[S_DOUBLE_TY]]]*, [[S_DOUBLE_TY]]* }
 // CHECK-DAG: [[PRIVATES_MAIN_TY:%.+]] = type {{.?}}{ [[S_DOUBLE_TY]], [2 x [[S_DOUBLE_TY]]], i32, [2 x i32]
-// CHECK-DAG: [[KMP_TASK_MAIN_TY:%.+]] = type { i8*, i32 (i32, i8*)*, i32, i32 (i32, i8*)*, [[PRIVATES_MAIN_TY]] }
+// 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]] }
-// CHECK-DAG: [[KMP_TASK_TMAIN_TY:%.+]] = type { i8*, i32 (i32, i8*)*, i32, i32 (i32, i8*)*, [[PRIVATES_TMAIN_TY]] }
+// CHECK-DAG: [[KMP_TASK_TMAIN_TY:%.+]] = type { [[KMP_TASK_T_TY]], [[PRIVATES_TMAIN_TY]] }
 template <typename T>
 T tmain() {
   S<T> ttt;
@@ -54,7 +55,7 @@ int main() {
   [&]() {
   // LAMBDA: define{{.*}} internal{{.*}} void [[OUTER_LAMBDA]](
   // LAMBDA: [[RES:%.+]] = call i8* @__kmpc_omp_task_alloc(%{{[^ ]+}} @{{[^,]+}}, i32 %{{[^,]+}}, i32 1, i64 40, i64 8, i32 (i32, i8*)* bitcast (i32 (i32, %{{[^*]+}}*)* [[TASK_ENTRY:@[^ ]+]] to i32 (i32, i8*)*))
-// LAMBDA: [[PRIVATES:%.+]] = getelementptr inbounds %{{.+}}, %{{.+}}* %{{.+}}, i{{.+}} 0, i{{.+}} 4
+// LAMBDA: [[PRIVATES:%.+]] = getelementptr inbounds %{{.+}}, %{{.+}}* %{{.+}}, i{{.+}} 0, i{{.+}} 1
 // LAMBDA: [[G_PRIVATE_ADDR:%.+]] = getelementptr inbounds %{{.+}}, %{{.+}}* [[PRIVATES]], i{{.+}} 0, i{{.+}} 0
 // LAMBDA: [[G_ADDR_REF:%.+]] = getelementptr inbounds %{{.+}}, %{{.+}}* %{{.+}}, i{{.+}} 0, i{{.+}} 0
 // LAMBDA: [[G_REF:%.+]] = load double*, double** [[G_ADDR_REF]]
@@ -91,7 +92,7 @@ int main() {
   ^{
   // BLOCKS: define{{.*}} internal{{.*}} void {{.+}}(i8*
   // BLOCKS: [[RES:%.+]] = call i8* @__kmpc_omp_task_alloc(%{{[^ ]+}} @{{[^,]+}}, i32 %{{[^,]+}}, i32 1, i64 40, i64 8, i32 (i32, i8*)* bitcast (i32 (i32, %{{[^*]+}}*)* [[TASK_ENTRY:@[^ ]+]] to i32 (i32, i8*)*))
-  // BLOCKS: [[PRIVATES:%.+]] = getelementptr inbounds %{{.+}}, %{{.+}}* %{{.+}}, i{{.+}} 0, i{{.+}} 4
+  // BLOCKS: [[PRIVATES:%.+]] = getelementptr inbounds %{{.+}}, %{{.+}}* %{{.+}}, i{{.+}} 0, i{{.+}} 1
   // BLOCKS: [[G_PRIVATE_ADDR:%.+]] = getelementptr inbounds %{{.+}}, %{{.+}}* [[PRIVATES]], i{{.+}} 0, i{{.+}} 0
   // BLOCKS: [[G_ADDR_REF:%.+]] = getelementptr inbounds %{{.+}}, %{{.+}}* %{{.+}}, i{{.+}} 0, i{{.+}} 0
   // BLOCKS: [[G_REF:%.+]] = load double*, double** [[G_ADDR_REF]]
@@ -159,24 +160,22 @@ int main() {
 
 // Allocate task.
 // Returns struct kmp_task_t {
-//         void *              shareds;
-//         kmp_routine_entry_t routine;
-//         kmp_int32           part_id;
-//         kmp_routine_entry_t destructors;
+//         [[KMP_TASK_T]] task_data;
 //         [[KMP_TASK_MAIN_TY]] privates;
 //       };
 // CHECK: [[RES:%.+]] = call i8* @__kmpc_omp_task_alloc([[LOC]], i32 [[GTID]], i32 1, i64 72, i64 32, i32 (i32, i8*)* bitcast (i32 (i32, [[KMP_TASK_MAIN_TY]]*)* [[TASK_ENTRY:@[^ ]+]] to i32 (i32, i8*)*))
 // CHECK: [[RES_KMP_TASK:%.+]] = bitcast i8* [[RES]] to [[KMP_TASK_MAIN_TY]]*
 
 // Fill kmp_task_t->shareds by copying from original capture argument.
-// CHECK: [[SHAREDS_REF_ADDR:%.+]] = getelementptr inbounds [[KMP_TASK_MAIN_TY]], [[KMP_TASK_MAIN_TY]]* [[RES_KMP_TASK]], i{{[0-9]+}} 0, i{{[0-9]+}} 0
+// 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*
 // CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[SHAREDS_REF]], i8* [[CAPTURES_ADDR]], i64 32, i32 8, i1 false)
 
 // 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]+}} 4
+// CHECK: [[PRIVATES:%.+]] = getelementptr inbounds [[KMP_TASK_MAIN_TY]], [[KMP_TASK_MAIN_TY]]* [[RES_KMP_TASK]], i{{[0-9]+}} 0, i{{[0-9]+}} 1
 // CHECK: [[SHAREDS:%.+]] = bitcast i8* [[SHAREDS_REF]] to [[CAP_MAIN_TY]]*
 
 // Constructors for s_arr and var.
@@ -225,7 +224,7 @@ int main() {
 // CHECK: store [2 x i32]* [[PRIVATE_VEC_REF]], [2 x i32]** [[VEC_ADDR_REF]],
 
 // Provide pointer to destructor function, which will destroy private variables at the end of the task.
-// CHECK: [[DESTRUCTORS_REF:%.+]] = getelementptr inbounds [[KMP_TASK_MAIN_TY]], [[KMP_TASK_MAIN_TY]]* [[RES_KMP_TASK]], i{{.+}} 0, i{{.+}} 3
+// CHECK: [[DESTRUCTORS_REF:%.+]] = getelementptr inbounds [[KMP_TASK_T_TY]], [[KMP_TASK_T_TY]]* [[TASK]], i{{.+}} 0, i{{.+}} 3
 // CHECK: store i32 (i32, i8*)* bitcast (i32 (i32, [[KMP_TASK_MAIN_TY]]*)* [[DESTRUCTORS:@.+]] to i32 (i32, i8*)*), i32 (i32, i8*)** [[DESTRUCTORS_REF]],
 
 // Start task.
@@ -243,7 +242,8 @@ int main() {
 //
 // CHECK: define internal i32 [[TASK_ENTRY]](i32, [[KMP_TASK_MAIN_TY]]*)
 
-// CHECK: [[SHAREDS_ADDR_REF:%.+]] = getelementptr inbounds [[KMP_TASK_MAIN_TY]], [[KMP_TASK_MAIN_TY]]* [[RES_KMP_TASK:%.+]], i{{.+}} 0, i{{.+}} 0
+// 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_ADDR_REF:%.+]] = getelementptr inbounds [[KMP_TASK_T_TY]], [[KMP_TASK_T_TY]]* [[TASK]], i{{.+}} 0, i{{.+}} 0
 // CHECK: [[SHAREDS_REF:%.+]] = load i8*, i8** [[SHAREDS_ADDR_REF]],
 // CHECK: [[SHAREDS:%.+]] = bitcast i8* [[SHAREDS_REF]] to [[CAP_MAIN_TY]]*
 
@@ -256,7 +256,7 @@ int main() {
 // CHECK: ret
 
 // CHECK: define internal i32 [[DESTRUCTORS]](i32, [[KMP_TASK_MAIN_TY]]*)
-// CHECK: [[PRIVATES:%.+]] = getelementptr inbounds [[KMP_TASK_MAIN_TY]], [[KMP_TASK_MAIN_TY]]* [[RES_KMP_TASK:%.+]], i{{[0-9]+}} 0, i{{[0-9]+}} 4
+// 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_VAR_REF:%.+]] = getelementptr inbounds [[PRIVATES_MAIN_TY]], [[PRIVATES_MAIN_TY]]* [[PRIVATES]], i{{.+}} 0, i{{.+}} 0
 // CHECK: [[PRIVATE_S_ARR_REF:%.+]] = getelementptr inbounds [[PRIVATES_MAIN_TY]], [[PRIVATES_MAIN_TY]]* [[PRIVATES]], i{{.+}} 0, i{{.+}} 1
 // CHECK: getelementptr inbounds [2 x [[S_DOUBLE_TY]]], [2 x [[S_DOUBLE_TY]]]* [[PRIVATE_S_ARR_REF]], i{{.+}} 0, i{{.+}} 0
@@ -291,23 +291,21 @@ int main() {
 
 // Allocate task.
 // Returns struct kmp_task_t {
-//         void *              shareds;
-//         kmp_routine_entry_t routine;
-//         kmp_int32           part_id;
-//         kmp_routine_entry_t destructors;
+//         [[KMP_TASK_T_TY]] task_data;
 //         [[KMP_TASK_TMAIN_TY]] privates;
 //       };
 // CHECK: [[RES:%.+]] = call i8* @__kmpc_omp_task_alloc([[LOC]], i32 [[GTID]], i32 1, i64 56, i64 32, i32 (i32, i8*)* bitcast (i32 (i32, [[KMP_TASK_TMAIN_TY]]*)* [[TASK_ENTRY:@[^ ]+]] to i32 (i32, i8*)*))
 // CHECK: [[RES_KMP_TASK:%.+]] = bitcast i8* [[RES]] to [[KMP_TASK_TMAIN_TY]]*
 
 // Fill kmp_task_t->shareds by copying from original capture argument.
-// CHECK: [[SHAREDS_REF_ADDR:%.+]] = getelementptr inbounds [[KMP_TASK_TMAIN_TY]], [[KMP_TASK_TMAIN_TY]]* [[RES_KMP_TASK]], i{{[0-9]+}} 0, i{{[0-9]+}} 0
+// 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*
 // CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[SHAREDS_REF]], i8* [[CAPTURES_ADDR]], i64 32, i32 8, i1 false)
 
 // 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]+}} 4
+// CHECK: [[PRIVATES:%.+]] = getelementptr inbounds [[KMP_TASK_TMAIN_TY]], [[KMP_TASK_TMAIN_TY]]* [[RES_KMP_TASK]], i{{[0-9]+}} 0, i{{[0-9]+}} 1
 // CHECK: [[SHAREDS:%.+]] = bitcast i8* [[SHAREDS_REF]] to [[CAP_TMAIN_TY]]*
 
 // t_var;
@@ -355,7 +353,7 @@ int main() {
 // CHECK: store [[S_INT_TY]]* [[PRIVATE_VAR_REF]], [[S_INT_TY]]** [[VAR_ADDR_REF]],
 
 // Provide pointer to destructor function, which will destroy private variables at the end of the task.
-// CHECK: [[DESTRUCTORS_REF:%.+]] = getelementptr inbounds [[KMP_TASK_TMAIN_TY]], [[KMP_TASK_TMAIN_TY]]* [[RES_KMP_TASK]], i{{.+}} 0, i{{.+}} 3
+// CHECK: [[DESTRUCTORS_REF:%.+]] = getelementptr inbounds [[KMP_TASK_T_TY]], [[KMP_TASK_T_TY]]* [[TASK]], i{{.+}} 0, i{{.+}} 3
 // CHECK: store i32 (i32, i8*)* bitcast (i32 (i32, [[KMP_TASK_TMAIN_TY]]*)* [[DESTRUCTORS:@.+]] to i32 (i32, i8*)*), i32 (i32, i8*)** [[DESTRUCTORS_REF]],
 
 // Start task.
@@ -372,7 +370,8 @@ int main() {
 // CHECK: define internal i32 [[TASK_ENTRY]](i32, [[KMP_TASK_TMAIN_TY]]*)
 
 // Substitute addresses of shared variables in capture struct by address of private copies from kmp_task_t.
-// CHECK: [[SHAREDS_ADDR_REF:%.+]] = getelementptr inbounds [[KMP_TASK_TMAIN_TY]], [[KMP_TASK_TMAIN_TY]]* [[RES_KMP_TASK:%.+]], i{{.+}} 0, i{{.+}} 0
+// 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_ADDR_REF:%.+]] = getelementptr inbounds [[KMP_TASK_T_TY]], [[KMP_TASK_T_TY]]* [[TASK]], i{{.+}} 0, i{{.+}} 0
 // CHECK: [[SHAREDS_REF:%.+]] = load i8*, i8** [[SHAREDS_ADDR_REF]],
 // CHECK: [[SHAREDS:%.+]] = bitcast i8* [[SHAREDS_REF]] to [[CAP_TMAIN_TY]]*
 
@@ -385,7 +384,7 @@ int main() {
 // CHECK: ret
 
 // CHECK: define internal i32 [[DESTRUCTORS]](i32, [[KMP_TASK_TMAIN_TY]]*)
-// CHECK: [[PRIVATES:%.+]] = getelementptr inbounds [[KMP_TASK_TMAIN_TY]], [[KMP_TASK_TMAIN_TY]]* [[RES_KMP_TASK:%.+]], i{{[0-9]+}} 0, i{{[0-9]+}} 4
+// CHECK: [[PRIVATES:%.+]] = getelementptr inbounds [[KMP_TASK_TMAIN_TY]], [[KMP_TASK_TMAIN_TY]]* [[RES_KMP_TASK:%.+]], i{{[0-9]+}} 0, i{{[0-9]+}} 1
 // 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]])
diff --git a/clang/test/OpenMP/task_private_codegen.cpp b/clang/test/OpenMP/task_private_codegen.cpp
index 2202d5f2548..197e8a8fa0d 100644
--- a/clang/test/OpenMP/task_private_codegen.cpp
+++ b/clang/test/OpenMP/task_private_codegen.cpp
@@ -22,14 +22,15 @@ struct S {
 
 volatile double g;
 
+// CHECK-DAG: [[KMP_TASK_T_TY:%.+]] = type { i8*, i32 (i32, i8*)*, i32, i32 (i32, i8*)* }
 // CHECK-DAG: [[S_DOUBLE_TY:%.+]] = type { double }
 // CHECK-DAG: [[CAP_MAIN_TY:%.+]] = type { [2 x i32]*, i32*, [2 x [[S_DOUBLE_TY]]]*, [[S_DOUBLE_TY]]* }
 // CHECK-DAG: [[PRIVATES_MAIN_TY:%.+]] = type {{.?}}{ [[S_DOUBLE_TY]], [2 x [[S_DOUBLE_TY]]], i32, [2 x i32]
-// CHECK-DAG: [[KMP_TASK_MAIN_TY:%.+]] = type { i8*, i32 (i32, i8*)*, i32, i32 (i32, i8*)*, [[PRIVATES_MAIN_TY]] }
+// 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]] }
-// CHECK-DAG: [[KMP_TASK_TMAIN_TY:%.+]] = type { i8*, i32 (i32, i8*)*, i32, i32 (i32, i8*)*, [[PRIVATES_TMAIN_TY]] }
+// CHECK-DAG: [[KMP_TASK_TMAIN_TY:%.+]] = type { [[KMP_TASK_T_TY]], [[PRIVATES_TMAIN_TY]] }
 template <typename T>
 T tmain() {
   S<T> test;
@@ -53,7 +54,7 @@ int main() {
   [&]() {
   // LAMBDA: define{{.*}} internal{{.*}} void [[OUTER_LAMBDA]](
   // LAMBDA: [[RES:%.+]] = call i8* @__kmpc_omp_task_alloc(%{{[^ ]+}} @{{[^,]+}}, i32 %{{[^,]+}}, i32 1, i64 40, i64 8, i32 (i32, i8*)* bitcast (i32 (i32, %{{[^*]+}}*)* [[TASK_ENTRY:@[^ ]+]] to i32 (i32, i8*)*))
-// LAMBDA: [[PRIVATES:%.+]] = getelementptr inbounds %{{.+}}, %{{.+}}* %{{.+}}, i{{.+}} 0, i{{.+}} 4
+// LAMBDA: [[PRIVATES:%.+]] = getelementptr inbounds %{{.+}}, %{{.+}}* %{{.+}}, i{{.+}} 0, i{{.+}} 1
 // LAMBDA: [[G_PRIVATE_ADDR:%.+]] = getelementptr inbounds %{{.+}}, %{{.+}}* [[PRIVATES]], i{{.+}} 0, i{{.+}} 0
 // LAMBDA: [[G_ADDR_REF:%.+]] = getelementptr inbounds %{{.+}}, %{{.+}}* %{{.+}}, i{{.+}} 0, i{{.+}} 0
 // LAMBDA: store double* [[G_PRIVATE_ADDR]], double** [[G_ADDR_REF]],
@@ -86,7 +87,7 @@ int main() {
   ^{
   // BLOCKS: define{{.*}} internal{{.*}} void {{.+}}(i8*
   // BLOCKS: [[RES:%.+]] = call i8* @__kmpc_omp_task_alloc(%{{[^ ]+}} @{{[^,]+}}, i32 %{{[^,]+}}, i32 1, i64 40, i64 8, i32 (i32, i8*)* bitcast (i32 (i32, %{{[^*]+}}*)* [[TASK_ENTRY:@[^ ]+]] to i32 (i32, i8*)*))
-  // BLOCKS: [[PRIVATES:%.+]] = getelementptr inbounds %{{.+}}, %{{.+}}* %{{.+}}, i{{.+}} 0, i{{.+}} 4
+  // BLOCKS: [[PRIVATES:%.+]] = getelementptr inbounds %{{.+}}, %{{.+}}* %{{.+}}, i{{.+}} 0, i{{.+}} 1
   // BLOCKS: [[G_PRIVATE_ADDR:%.+]] = getelementptr inbounds %{{.+}}, %{{.+}}* [[PRIVATES]], i{{.+}} 0, i{{.+}} 0
   // BLOCKS: [[G_ADDR_REF:%.+]] = getelementptr inbounds %{{.+}}, %{{.+}}* %{{.+}}, i{{.+}} 0, i{{.+}} 0
   // BLOCKS: store double* [[G_PRIVATE_ADDR]], double** [[G_ADDR_REF]],
@@ -148,24 +149,22 @@ int main() {
 
 // Allocate task.
 // Returns struct kmp_task_t {
-//         void *              shareds;
-//         kmp_routine_entry_t routine;
-//         kmp_int32           part_id;
-//         kmp_routine_entry_t destructors;
+//         [[KMP_TASK_T_TY]] task_data;
 //         [[KMP_TASK_MAIN_TY]] privates;
 //       };
 // CHECK: [[RES:%.+]] = call i8* @__kmpc_omp_task_alloc([[LOC]], i32 [[GTID]], i32 1, i64 72, i64 32, i32 (i32, i8*)* bitcast (i32 (i32, [[KMP_TASK_MAIN_TY]]*)* [[TASK_ENTRY:@[^ ]+]] to i32 (i32, i8*)*))
 // CHECK: [[RES_KMP_TASK:%.+]] = bitcast i8* [[RES]] to [[KMP_TASK_MAIN_TY]]*
 
 // Fill kmp_task_t->shareds by copying from original capture argument.
-// CHECK: [[SHAREDS_REF_ADDR:%.+]] = getelementptr inbounds [[KMP_TASK_MAIN_TY]], [[KMP_TASK_MAIN_TY]]* [[RES_KMP_TASK]], i{{[0-9]+}} 0, i{{[0-9]+}} 0
+// 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*
 // CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[SHAREDS_REF]], i8* [[CAPTURES_ADDR]], i64 32, i32 8, i1 false)
 
 // 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]+}} 4
+// CHECK: [[PRIVATES:%.+]] = getelementptr inbounds [[KMP_TASK_MAIN_TY]], [[KMP_TASK_MAIN_TY]]* [[RES_KMP_TASK]], i{{[0-9]+}} 0, i{{[0-9]+}} 1
 // CHECK: [[SHAREDS:%.+]] = bitcast i8* [[SHAREDS_REF]] to [[CAP_MAIN_TY]]*
 
 // Constructors for s_arr and var.
@@ -201,7 +200,7 @@ int main() {
 // CHECK: store [2 x i32]* [[PRIVATE_VEC_REF]], [2 x i32]** [[VEC_ADDR_REF]],
 
 // Provide pointer to destructor function, which will destroy private variables at the end of the task.
-// CHECK: [[DESTRUCTORS_REF:%.+]] = getelementptr inbounds [[KMP_TASK_MAIN_TY]], [[KMP_TASK_MAIN_TY]]* [[RES_KMP_TASK]], i{{.+}} 0, i{{.+}} 3
+// CHECK: [[DESTRUCTORS_REF:%.+]] = getelementptr inbounds [[KMP_TASK_T_TY]], [[KMP_TASK_T_TY]]* [[TASK]], i{{.+}} 0, i{{.+}} 3
 // CHECK: store i32 (i32, i8*)* bitcast (i32 (i32, [[KMP_TASK_MAIN_TY]]*)* [[DESTRUCTORS:@.+]] to i32 (i32, i8*)*), i32 (i32, i8*)** [[DESTRUCTORS_REF]],
 
 // Start task.
@@ -219,7 +218,8 @@ int main() {
 //
 // CHECK: define internal i32 [[TASK_ENTRY]](i32, [[KMP_TASK_MAIN_TY]]*)
 
-// CHECK: [[SHAREDS_ADDR_REF:%.+]] = getelementptr inbounds [[KMP_TASK_MAIN_TY]], [[KMP_TASK_MAIN_TY]]* [[RES_KMP_TASK:%.+]], i{{.+}} 0, i{{.+}} 0
+// 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_ADDR_REF:%.+]] = getelementptr inbounds [[KMP_TASK_T_TY]], [[KMP_TASK_T_TY]]* [[TASK]], i{{.+}} 0, i{{.+}} 0
 // CHECK: [[SHAREDS_REF:%.+]] = load i8*, i8** [[SHAREDS_ADDR_REF]],
 // CHECK: [[SHAREDS:%.+]] = bitcast i8* [[SHAREDS_REF]] to [[CAP_MAIN_TY]]*
 
@@ -232,7 +232,7 @@ int main() {
 // CHECK: ret
 
 // CHECK: define internal i32 [[DESTRUCTORS]](i32, [[KMP_TASK_MAIN_TY]]*)
-// CHECK: [[PRIVATES:%.+]] = getelementptr inbounds [[KMP_TASK_MAIN_TY]], [[KMP_TASK_MAIN_TY]]* [[RES_KMP_TASK:%.+]], i{{[0-9]+}} 0, i{{[0-9]+}} 4
+// 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_VAR_REF:%.+]] = getelementptr inbounds [[PRIVATES_MAIN_TY]], [[PRIVATES_MAIN_TY]]* [[PRIVATES]], i{{.+}} 0, i{{.+}} 0
 // CHECK: [[PRIVATE_S_ARR_REF:%.+]] = getelementptr inbounds [[PRIVATES_MAIN_TY]], [[PRIVATES_MAIN_TY]]* [[PRIVATES]], i{{.+}} 0, i{{.+}} 1
 // CHECK: getelementptr inbounds [2 x [[S_DOUBLE_TY]]], [2 x [[S_DOUBLE_TY]]]* [[PRIVATE_S_ARR_REF]], i{{.+}} 0, i{{.+}} 0
@@ -266,23 +266,21 @@ int main() {
 
 // Allocate task.
 // Returns struct kmp_task_t {
-//         void *              shareds;
-//         kmp_routine_entry_t routine;
-//         kmp_int32           part_id;
-//         kmp_routine_entry_t destructors;
+//         [[KMP_TASK_T_TY]] task_data;
 //         [[KMP_TASK_TMAIN_TY]] privates;
 //       };
 // CHECK: [[RES:%.+]] = call i8* @__kmpc_omp_task_alloc([[LOC]], i32 [[GTID]], i32 1, i64 56, i64 32, i32 (i32, i8*)* bitcast (i32 (i32, [[KMP_TASK_TMAIN_TY]]*)* [[TASK_ENTRY:@[^ ]+]] to i32 (i32, i8*)*))
 // CHECK: [[RES_KMP_TASK:%.+]] = bitcast i8* [[RES]] to [[KMP_TASK_TMAIN_TY]]*
 
 // Fill kmp_task_t->shareds by copying from original capture argument.
-// CHECK: [[SHAREDS_REF_ADDR:%.+]] = getelementptr inbounds [[KMP_TASK_TMAIN_TY]], [[KMP_TASK_TMAIN_TY]]* [[RES_KMP_TASK]], i{{[0-9]+}} 0, i{{[0-9]+}} 0
+// 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*
 // CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[SHAREDS_REF]], i8* [[CAPTURES_ADDR]], i64 32, i32 8, i1 false)
 
 // 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]+}} 4
+// CHECK: [[PRIVATES:%.+]] = getelementptr inbounds [[KMP_TASK_TMAIN_TY]], [[KMP_TASK_TMAIN_TY]]* [[RES_KMP_TASK]], i{{[0-9]+}} 0, i{{[0-9]+}} 1
 // CHECK: [[SHAREDS:%.+]] = bitcast i8* [[SHAREDS_REF]] to [[CAP_TMAIN_TY]]*
 
 // shareds->t_var_addr = &kmp_task_t->privates.t_var;
@@ -318,7 +316,7 @@ int main() {
 // CHECK: store [[S_INT_TY]]* [[PRIVATE_VAR_REF]], [[S_INT_TY]]** [[VAR_ADDR_REF]],
 
 // Provide pointer to destructor function, which will destroy private variables at the end of the task.
-// CHECK: [[DESTRUCTORS_REF:%.+]] = getelementptr inbounds [[KMP_TASK_TMAIN_TY]], [[KMP_TASK_TMAIN_TY]]* [[RES_KMP_TASK]], i{{.+}} 0, i{{.+}} 3
+// CHECK: [[DESTRUCTORS_REF:%.+]] = getelementptr inbounds [[KMP_TASK_T_TY]], [[KMP_TASK_T_TY]]* [[TASK]], i{{.+}} 0, i{{.+}} 3
 // CHECK: store i32 (i32, i8*)* bitcast (i32 (i32, [[KMP_TASK_TMAIN_TY]]*)* [[DESTRUCTORS:@.+]] to i32 (i32, i8*)*), i32 (i32, i8*)** [[DESTRUCTORS_REF]],
 
 // Start task.
@@ -335,7 +333,8 @@ int main() {
 // CHECK: define internal i32 [[TASK_ENTRY]](i32, [[KMP_TASK_TMAIN_TY]]*)
 
 // Substitute addresses of shared variables in capture struct by address of private copies from kmp_task_t.
-// CHECK: [[SHAREDS_ADDR_REF:%.+]] = getelementptr inbounds [[KMP_TASK_TMAIN_TY]], [[KMP_TASK_TMAIN_TY]]* [[RES_KMP_TASK:%.+]], i{{.+}} 0, i{{.+}} 0
+// 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_ADDR_REF:%.+]] = getelementptr inbounds [[KMP_TASK_T_TY]], [[KMP_TASK_T_TY]]* [[TASK]], i{{.+}} 0, i{{.+}} 0
 // CHECK: [[SHAREDS_REF:%.+]] = load i8*, i8** [[SHAREDS_ADDR_REF]],
 // CHECK: [[SHAREDS:%.+]] = bitcast i8* [[SHAREDS_REF]] to [[CAP_TMAIN_TY]]*
 
@@ -348,7 +347,7 @@ int main() {
 // CHECK: ret
 
 // CHECK: define internal i32 [[DESTRUCTORS]](i32, [[KMP_TASK_TMAIN_TY]]*)
-// CHECK: [[PRIVATES:%.+]] = getelementptr inbounds [[KMP_TASK_TMAIN_TY]], [[KMP_TASK_TMAIN_TY]]* [[RES_KMP_TASK:%.+]], i{{[0-9]+}} 0, i{{[0-9]+}} 4
+// CHECK: [[PRIVATES:%.+]] = getelementptr inbounds [[KMP_TASK_TMAIN_TY]], [[KMP_TASK_TMAIN_TY]]* [[RES_KMP_TASK:%.+]], i{{[0-9]+}} 0, i{{[0-9]+}} 1
 // 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]])