Imported Upstream version 6.10.0.49

Former-commit-id: 1d6753294b2993e1fbf92de9366bb9544db4189b

external/llvm-project/openmp/runtime/test/ompt/worksharing/for/auto.c

@@ -0,0 +1,7 @@
+// RUN: %libomp-compile-and-run | %sort-threads | FileCheck %S/base.h
+// REQUIRES: ompt
+// GCC doesn't call runtime for auto = static schedule
+// XFAIL: gcc
+
+#define SCHEDULE auto
+#include "base.h"

external/llvm-project/openmp/runtime/test/ompt/worksharing/for/auto_serialized.c

@@ -0,0 +1,7 @@
+// RUN: %libomp-compile-and-run | %sort-threads | FileCheck %S/base_serialized.h
+// REQUIRES: ompt
+// GCC doesn't call runtime for auto = static schedule
+// XFAIL: gcc
+
+#define SCHEDULE auto
+#include "base_serialized.h"

external/llvm-project/openmp/runtime/test/ompt/worksharing/for/auto_split.c

@@ -0,0 +1,8 @@
+// RUN: %libomp-compile-and-run | %sort-threads | FileCheck %S/base_split.h
+// RUN: %libomp-compile-and-run | %sort-threads | FileCheck --check-prefix=CHECK-LOOP %S/base_split.h
+// REQUIRES: ompt
+// GCC doesn't call runtime for auto = static schedule
+// XFAIL: gcc
+
+#define SCHEDULE auto
+#include "base_split.h"

external/llvm-project/openmp/runtime/test/ompt/worksharing/for/base.h

@@ -0,0 +1,43 @@
+#include "callback.h"
+#include <omp.h>
+
+int main()
+{
+  unsigned int i;
+
+  #pragma omp parallel for num_threads(4) schedule(SCHEDULE)
+  for (i = 0; i < 4; i++) {
+  }
+
+  // Check if libomp supports the callbacks for this test.
+  // CHECK-NOT: {{^}}0: Could not register callback 'ompt_callback_parallel_begin'
+  // CHECK-NOT: {{^}}0: Could not register callback 'ompt_callback_parallel_end'
+  // CHECK-NOT: {{^}}0: Could not register callback 'ompt_callback_implicit_task'
+  // CHECK-NOT: {{^}}0: Could not register callback 'ompt_callback_work'
+
+
+  // CHECK: 0: NULL_POINTER=[[NULL:.*$]]
+  // CHECK: {{^}}[[MASTER_ID:[0-9]+]]: ompt_event_parallel_begin: parent_task_id={{[0-9]+}}, parent_task_frame.exit=[[NULL]], parent_task_frame.reenter={{0x[0-f]+}}, parallel_id=[[PARALLEL_ID:[0-9]+]], requested_team_size=4, codeptr_ra=0x{{[0-f]+}}, invoker={{[0-9]+}}
+
+  // CHECK: {{^}}[[MASTER_ID]]: ompt_event_implicit_task_begin: parallel_id=[[PARALLEL_ID]], task_id=[[IMPLICIT_TASK_ID:[0-9]+]]
+  // CHECK: {{^}}[[MASTER_ID]]: ompt_event_loop_begin: parallel_id=[[PARALLEL_ID]], parent_task_id=[[IMPLICIT_TASK_ID]], codeptr_ra=
+  // CHECK: {{^}}[[MASTER_ID]]: ompt_event_loop_end: parallel_id=[[PARALLEL_ID]], task_id=[[IMPLICIT_TASK_ID]]
+  // CHECK: {{^}}[[MASTER_ID]]: ompt_event_implicit_task_end: parallel_id={{[0-9]+}}, task_id=[[IMPLICIT_TASK_ID]]
+
+  // CHECK: {{^}}[[THREAD_ID:[0-9]+]]: ompt_event_implicit_task_begin: parallel_id=[[PARALLEL_ID]], task_id=[[IMPLICIT_TASK_ID:[0-9]+]]
+  // CHECK: {{^}}[[THREAD_ID]]: ompt_event_loop_begin: parallel_id=[[PARALLEL_ID]], parent_task_id=[[IMPLICIT_TASK_ID]], codeptr_ra=
+  // CHECK: {{^}}[[THREAD_ID]]: ompt_event_loop_end: parallel_id=[[PARALLEL_ID]], task_id=[[IMPLICIT_TASK_ID]]
+  // CHECK: {{^}}[[THREAD_ID]]: ompt_event_implicit_task_end: parallel_id={{[0-9]+}}, task_id=[[IMPLICIT_TASK_ID]]
+
+  // CHECK: {{^}}[[THREAD_ID:[0-9]+]]: ompt_event_implicit_task_begin: parallel_id=[[PARALLEL_ID]], task_id=[[IMPLICIT_TASK_ID:[0-9]+]]
+  // CHECK: {{^}}[[THREAD_ID]]: ompt_event_loop_begin: parallel_id=[[PARALLEL_ID]], parent_task_id=[[IMPLICIT_TASK_ID]], codeptr_ra=
+  // CHECK: {{^}}[[THREAD_ID]]: ompt_event_loop_end: parallel_id=[[PARALLEL_ID]], task_id=[[IMPLICIT_TASK_ID]]
+  // CHECK: {{^}}[[THREAD_ID]]: ompt_event_implicit_task_end: parallel_id={{[0-9]+}}, task_id=[[IMPLICIT_TASK_ID]]
+
+  // CHECK: {{^}}[[THREAD_ID:[0-9]+]]: ompt_event_implicit_task_begin: parallel_id=[[PARALLEL_ID]], task_id=[[IMPLICIT_TASK_ID:[0-9]+]]
+  // CHECK: {{^}}[[THREAD_ID]]: ompt_event_loop_begin: parallel_id=[[PARALLEL_ID]], parent_task_id=[[IMPLICIT_TASK_ID]], codeptr_ra=
+  // CHECK: {{^}}[[THREAD_ID]]: ompt_event_loop_end: parallel_id=[[PARALLEL_ID]], task_id=[[IMPLICIT_TASK_ID]]
+  // CHECK: {{^}}[[THREAD_ID]]: ompt_event_implicit_task_end: parallel_id={{[0-9]+}}, task_id=[[IMPLICIT_TASK_ID]]
+
+  return 0;
+}

external/llvm-project/openmp/runtime/test/ompt/worksharing/for/base_serialized.h

@@ -0,0 +1,28 @@
+#include "callback.h"
+#include <omp.h>
+
+int main()
+{
+  unsigned int i;
+
+  #pragma omp parallel for num_threads(1) schedule(SCHEDULE)
+  for (i = 0; i < 1; i++) {
+  }
+  
+  // Check if libomp supports the callbacks for this test.
+  // CHECK-NOT: {{^}}0: Could not register callback 'ompt_callback_parallel_begin'
+  // CHECK-NOT: {{^}}0: Could not register callback 'ompt_callback_parallel_end'
+  // CHECK-NOT: {{^}}0: Could not register callback 'ompt_callback_implicit_task'
+  // CHECK-NOT: {{^}}0: Could not register callback 'ompt_callback_work'
+
+
+  // CHECK: 0: NULL_POINTER=[[NULL:.*$]]
+  // CHECK: {{^}}[[MASTER_ID:[0-9]+]]: ompt_event_parallel_begin: parent_task_id=[[PARENT_TASK_ID:[0-9]+]], parent_task_frame.exit=[[NULL]], parent_task_frame.reenter={{0x[0-f]+}}, parallel_id=[[PARALLEL_ID:[0-9]+]], requested_team_size=1, codeptr_ra=0x{{[0-f]+}}, invoker={{[0-9]+}}
+
+  // CHECK: {{^}}[[MASTER_ID]]: ompt_event_implicit_task_begin: parallel_id=[[PARALLEL_ID:[0-9]+]], task_id=[[IMPLICIT_TASK_ID:[0-9]+]]
+  // CHECK: {{^}}[[MASTER_ID]]: ompt_event_loop_begin: parallel_id=[[PARALLEL_ID]], parent_task_id=[[IMPLICIT_TASK_ID]], codeptr_ra=0x{{[0-f]+}}
+  // CHECK: {{^}}[[MASTER_ID]]: ompt_event_loop_end: parallel_id=[[PARALLEL_ID]], task_id=[[IMPLICIT_TASK_ID]]
+  // CHECK: {{^}}[[MASTER_ID]]: ompt_event_implicit_task_end: parallel_id={{[PARALLEL_ID,0]}}, task_id=[[IMPLICIT_TASK_ID]]
+
+  return 0;
+}

external/llvm-project/openmp/runtime/test/ompt/worksharing/for/base_split.h

@@ -0,0 +1,66 @@
+#include "callback.h"
+#include <omp.h>
+
+/* With the combined parallel-for construct (base.h), the return-addresses are hard to compare.
+   With the separate parallel and for-nowait construct, the addresses become more predictable,
+   but the begin of the for-loop still generates additional code, so the offset of loop-begin 
+   to the label is >4 Byte.
+*/
+
+int main()
+{
+  unsigned int i;
+
+  #pragma omp parallel num_threads(4) 
+  {
+    print_current_address(0);
+    #pragma omp for schedule(SCHEDULE) nowait
+    for (i = 0; i < 4; i++) {
+      print_fuzzy_address(1);
+    }
+    print_fuzzy_address(2);
+  }
+  print_fuzzy_address(3);
+
+  // Check if libomp supports the callbacks for this test.
+  // CHECK-NOT: {{^}}0: Could not register callback 'ompt_callback_parallel_begin'
+  // CHECK-NOT: {{^}}0: Could not register callback 'ompt_callback_parallel_end'
+  // CHECK-NOT: {{^}}0: Could not register callback 'ompt_callback_implicit_task'
+  // CHECK-NOT: {{^}}0: Could not register callback 'ompt_callback_work'
+
+
+  // CHECK: 0: NULL_POINTER=[[NULL:.*$]]
+  // CHECK: {{^}}[[MASTER_ID:[0-9]+]]: ompt_event_parallel_begin: parent_task_id={{[0-9]+}}, parent_task_frame.exit=[[NULL]], parent_task_frame.reenter={{0x[0-f]+}}, parallel_id=[[PARALLEL_ID:[0-9]+]], requested_team_size=4, codeptr_ra=[[PARALLEL_RETURN_ADDRESS:0x[0-f]+]]{{[0-f][0-f]}}, invoker={{[0-9]+}}
+
+  // CHECK: {{^}}[[MASTER_ID]]: ompt_event_loop_begin: parallel_id=[[PARALLEL_ID]], parent_task_id={{[0-9]+}}, codeptr_ra=[[LOOP_BEGIN_RETURN_ADDRESS:0x[0-f]+]]{{[0-f][0-f]}}
+  // CHECK: {{^}}[[MASTER_ID]]: ompt_event_loop_end: parallel_id=[[PARALLEL_ID]], task_id={{[0-9]+}}, codeptr_ra=[[LOOP_END_RETURN_ADDRESS:0x[0-f]+]]{{[0-f][0-f]}}
+  // CHECK: {{^}}[[MASTER_ID]]: fuzzy_address={{.*}}[[LOOP_END_RETURN_ADDRESS]]
+
+  // CHECK: {{^}}[[MASTER_ID]]: ompt_event_parallel_end: parallel_id=[[PARALLEL_ID]], task_id={{[0-9]+}}, invoker={{[0-9]+}}, codeptr_ra=[[PARALLEL_RETURN_ADDRESS]]
+  // CHECK: {{^}}[[MASTER_ID]]: fuzzy_address={{.*}}[[PARALLEL_RETURN_ADDRESS]]
+  
+  // CHECK: {{^}}[[THREAD_ID:[0-9]+]]: ompt_event_loop_begin: parallel_id=[[PARALLEL_ID]], parent_task_id={{[0-9]+}}, codeptr_ra=0x{{[0-f]+}}
+  // CHECK: {{^}}[[THREAD_ID]]: ompt_event_loop_end: parallel_id=[[PARALLEL_ID]], task_id={{[0-9]+}}, codeptr_ra=[[LOOP_END_RETURN_ADDRESS:0x[0-f]+]]{{[0-f][0-f]}}
+  // CHECK: {{^}}[[THREAD_ID]]: fuzzy_address={{.*}}[[LOOP_END_RETURN_ADDRESS]]
+
+  // CHECK: {{^}}[[THREAD_ID:[0-9]+]]: ompt_event_loop_begin: parallel_id=[[PARALLEL_ID]], parent_task_id={{[0-9]+}}, codeptr_ra=0x{{[0-f]+}}
+  // CHECK: {{^}}[[THREAD_ID]]: ompt_event_loop_end: parallel_id=[[PARALLEL_ID]], task_id={{[0-9]+}}, codeptr_ra=[[LOOP_END_RETURN_ADDRESS:0x[0-f]+]]{{[0-f][0-f]}}
+  // CHECK: {{^}}[[THREAD_ID]]: fuzzy_address={{.*}}[[LOOP_END_RETURN_ADDRESS]]
+
+  // CHECK: {{^}}[[THREAD_ID:[0-9]+]]: ompt_event_loop_begin: parallel_id=[[PARALLEL_ID]], parent_task_id={{[0-9]+}}, codeptr_ra=0x{{[0-f]+}}
+  // CHECK: {{^}}[[THREAD_ID]]: ompt_event_loop_end: parallel_id=[[PARALLEL_ID]], task_id={{[0-9]+}}, codeptr_ra=[[LOOP_END_RETURN_ADDRESS:0x[0-f]+]]{{[0-f][0-f]}}
+  // CHECK: {{^}}[[THREAD_ID]]: fuzzy_address={{.*}}[[LOOP_END_RETURN_ADDRESS]]
+
+
+  // CHECK-LOOP: 0: NULL_POINTER=[[NULL:.*$]]
+  // CHECK-LOOP: 0: ompt_event_runtime_shutdown
+  // CHECK-LOOP: {{^}}[[MASTER_ID:[0-9]+]]: ompt_event_parallel_begin: parent_task_id={{[0-9]+}}, parent_task_frame.exit=[[NULL]], parent_task_frame.reenter={{0x[0-f]+}}, parallel_id=[[PARALLEL_ID:[0-9]+]], requested_team_size=4, codeptr_ra={{0x[0-f]+}}, invoker={{[0-9]+}}
+  // CHECK-LOOP: {{^}}[[MASTER_ID]]: ompt_event_loop_begin: parallel_id=[[PARALLEL_ID]], parent_task_id={{[0-9]+}}, codeptr_ra=[[LOOP_BEGIN_RETURN_ADDRESS:0x[0-f]+]]{{[0-f][0-f]}}
+  // CHECK-LOOP: {{^}}{{[0-9]+}}: fuzzy_address={{.*}}[[LOOP_BEGIN_RETURN_ADDRESS]]
+  // CHECK-LOOP: {{^}}{{[0-9]+}}: fuzzy_address={{.*}}[[LOOP_BEGIN_RETURN_ADDRESS]]
+  // CHECK-LOOP: {{^}}{{[0-9]+}}: fuzzy_address={{.*}}[[LOOP_BEGIN_RETURN_ADDRESS]]
+  // CHECK-LOOP: {{^}}{{[0-9]+}}: fuzzy_address={{.*}}[[LOOP_BEGIN_RETURN_ADDRESS]]
+
+
+  return 0;
+}

external/llvm-project/openmp/runtime/test/ompt/worksharing/for/dynamic.c

@@ -0,0 +1,5 @@
+// RUN: %libomp-compile-and-run | %sort-threads | FileCheck %S/base.h
+// REQUIRES: ompt
+
+#define SCHEDULE dynamic
+#include "base.h"

external/llvm-project/openmp/runtime/test/ompt/worksharing/for/dynamic_serialized.c

@@ -0,0 +1,5 @@
+// RUN: %libomp-compile-and-run | %sort-threads | FileCheck %S/base_serialized.h
+// REQUIRES: ompt
+
+#define SCHEDULE dynamic
+#include "base_serialized.h"

external/llvm-project/openmp/runtime/test/ompt/worksharing/for/dynamic_split.c

@@ -0,0 +1,7 @@
+// RUN: %libomp-compile-and-run | %sort-threads | FileCheck %S/base_split.h
+// RUN: %libomp-compile-and-run | %sort-threads | FileCheck --check-prefix=CHECK-LOOP %S/base_split.h
+// REQUIRES: ompt
+// UNSUPPORTED: gcc-4, gcc-5, gcc-6, gcc-7
+
+#define SCHEDULE dynamic
+#include "base_split.h"

external/llvm-project/openmp/runtime/test/ompt/worksharing/for/guided.c

@@ -0,0 +1,5 @@
+// RUN: %libomp-compile-and-run | %sort-threads | FileCheck %S/base.h
+// REQUIRES: ompt
+
+#define SCHEDULE guided
+#include "base.h"

external/llvm-project/openmp/runtime/test/ompt/worksharing/for/guided_serialized.c

@@ -0,0 +1,5 @@
+// RUN: %libomp-compile-and-run | %sort-threads | FileCheck %S/base_serialized.h
+// REQUIRES: ompt
+
+#define SCHEDULE guided
+#include "base_serialized.h"

external/llvm-project/openmp/runtime/test/ompt/worksharing/for/guided_split.c

@@ -0,0 +1,7 @@
+// RUN: %libomp-compile-and-run | %sort-threads | FileCheck %S/base_split.h
+// RUN: %libomp-compile-and-run | %sort-threads | FileCheck --check-prefix=CHECK-LOOP %S/base_split.h
+// REQUIRES: ompt
+// UNSUPPORTED: gcc-4, gcc-5, gcc-6, gcc-7
+
+#define SCHEDULE guided
+#include "base_split.h"

external/llvm-project/openmp/runtime/test/ompt/worksharing/for/runtime.c

@@ -0,0 +1,5 @@
+// RUN: %libomp-compile-and-run | %sort-threads | FileCheck %S/base.h
+// REQUIRES: ompt
+
+#define SCHEDULE runtime
+#include "base.h"

external/llvm-project/openmp/runtime/test/ompt/worksharing/for/runtime_serialized.c

@@ -0,0 +1,5 @@
+// RUN: %libomp-compile-and-run | %sort-threads | FileCheck %S/base_serialized.h
+// REQUIRES: ompt
+
+#define SCHEDULE runtime
+#include "base_serialized.h"

external/llvm-project/openmp/runtime/test/ompt/worksharing/for/runtime_split.c

@@ -0,0 +1,7 @@
+// RUN: %libomp-compile-and-run | %sort-threads | FileCheck %S/base_split.h
+// RUN: %libomp-compile-and-run | %sort-threads | FileCheck --check-prefix=CHECK-LOOP %S/base_split.h
+// REQUIRES: ompt
+// UNSUPPORTED: gcc-4, gcc-5, gcc-6, gcc-7
+
+#define SCHEDULE runtime
+#include "base_split.h"

external/llvm-project/openmp/runtime/test/ompt/worksharing/for/static.c

@@ -0,0 +1,7 @@
+// RUN: %libomp-compile-and-run | %sort-threads | FileCheck %S/base.h
+// REQUIRES: ompt
+// GCC doesn't call runtime for static schedule
+// XFAIL: gcc
+
+#define SCHEDULE static
+#include "base.h"

external/llvm-project/openmp/runtime/test/ompt/worksharing/for/static_serialized.c

@@ -0,0 +1,7 @@
+// RUN: %libomp-compile-and-run | %sort-threads | FileCheck %S/base_serialized.h
+// REQUIRES: ompt
+// GCC doesn't call runtime for static schedule
+// XFAIL: gcc
+
+#define SCHEDULE static
+#include "base_serialized.h"

external/llvm-project/openmp/runtime/test/ompt/worksharing/for/static_split.c

@@ -0,0 +1,8 @@
+// RUN: %libomp-compile-and-run | %sort-threads | FileCheck %S/base_split.h
+// RUN: %libomp-compile-and-run | %sort-threads | FileCheck --check-prefix=CHECK-LOOP %S/base_split.h
+// REQUIRES: ompt
+// GCC doesn't call runtime for static schedule
+// XFAIL: gcc
+
+#define SCHEDULE static
+#include "base_split.h"

external/llvm-project/openmp/runtime/test/ompt/worksharing/sections.c

@@ -0,0 +1,36 @@
+// RUN: %libomp-compile-and-run | %sort-threads | FileCheck %s
+// REQUIRES: ompt
+// Some compilers generate code that does not distinguish between sections and loops
+// XFAIL: gcc, clang-3, clang-4, clang-5, icc-16, icc-17
+// UNSUPPORTED: icc-18
+
+#include "callback.h"
+#include <omp.h>
+
+int main()
+{
+  #pragma omp parallel sections num_threads(2)
+  {
+    #pragma omp section
+    {
+      printf("%lu: section 1\n", ompt_get_thread_data()->value);
+    }
+    #pragma omp section
+    {
+      printf("%lu: section 2\n", ompt_get_thread_data()->value);
+    }
+  }
+
+  // Check if libomp supports the callbacks for this test.
+  // CHECK-NOT: {{^}}0: Could not register callback 'ompt_callback_work'
+
+  // CHECK: 0: NULL_POINTER=[[NULL:.*$]]
+
+  // CHECK: {{^}}[[MASTER_ID:[0-9]+]]: ompt_event_sections_begin: parallel_id=[[PARALLEL_ID:[0-9]+]], parent_task_id=[[TASK_ID:[0-9]+]], codeptr_ra=[[SECT_BEGIN:0x[0-f]+]], count=2
+  // CHECK: {{^}}[[MASTER_ID]]: ompt_event_sections_end: parallel_id=[[PARALLEL_ID]], task_id={{[0-9]+}}, codeptr_ra=[[SECT_END:0x[0-f]+]]
+
+  // CHECK: {{^}}[[THREAD_ID:[0-9]+]]: ompt_event_sections_begin: parallel_id=[[PARALLEL_ID]], parent_task_id=[[TASK_ID:[0-9]+]], codeptr_ra=[[SECT_BEGIN]], count=2
+  // CHECK: {{^}}[[THREAD_ID]]: ompt_event_sections_end: parallel_id=[[PARALLEL_ID]], task_id={{[0-9]+}}, codeptr_ra=[[SECT_END]]
+
+  return 0;
+}

external/llvm-project/openmp/runtime/test/ompt/worksharing/single.c

@@ -0,0 +1,36 @@
+// RUN: %libomp-compile-and-run | %sort-threads | FileCheck %s
+// REQUIRES: ompt
+// GCC generates code that does not call the runtime for the single construct
+// XFAIL: gcc
+
+#include "callback.h"
+#include <omp.h>
+
+int main()
+{
+  int x = 0;
+  #pragma omp parallel num_threads(2)
+  {
+    #pragma omp single
+    {
+      printf("%" PRIu64 ": in single\n", ompt_get_thread_data()->value);
+      x++;
+    }
+  }
+
+  printf("x=%d\n", x);
+
+  // Check if libomp supports the callbacks for this test.
+  // CHECK-NOT: {{^}}0: Could not register callback 'ompt_callback_work'
+
+  // CHECK: 0: NULL_POINTER=[[NULL:.*$]]
+
+  // CHECK-DAG: {{^}}[[THREAD_ID_1:[0-9]+]]: ompt_event_single_in_block_begin: parallel_id=[[PARALLEL_ID:[0-9]+]], parent_task_id=[[TASK_ID:[0-9]+]], codeptr_ra={{0x[0-f]+}}, count=1
+  // CHECK-DAG: {{^}}[[THREAD_ID_1]]: in single
+  // CHECK-DAG: {{^}}[[THREAD_ID_1]]: ompt_event_single_in_block_end: parallel_id=[[PARALLEL_ID]], task_id=[[TASK_ID]], codeptr_ra={{0x[0-f]+}}, count=1
+
+  // CHECK-DAG: {{^}}[[THREAD_ID_2:[0-9]+]]: ompt_event_single_others_begin: parallel_id=[[PARALLEL_ID:[0-9]+]], task_id=[[TASK_ID:[0-9]+]], codeptr_ra={{0x[0-f]+}}, count=1
+  // CHECK-DAG: {{^}}[[THREAD_ID_2]]: ompt_event_single_others_end: parallel_id=[[PARALLEL_ID]], task_id=[[TASK_ID]], codeptr_ra={{0x[0-f]+}}, count=1
+
+  return 0;
+}