openmp: Add support for strict modifier on grainsize/num_tasks clauses

With strict: modifier on these clauses, the standard is explicit about
how many iterations (and which) each generated task of taskloop directive
should contain.  For num_tasks it actually matches what we were already
implementing, but for grainsize it does not (and even violates the old
rule - without strict it requires that the number of iterations (unspecified
which exactly) handled by each generated task is >= grainsize argument and
< 2 * grainsize argument, with strict: it requires that each generated
task handles exactly == grainsize argument iterations, except for the
generated task handling the last iteration which can handles <= grainsize
iterations).

The following patch implements it for C and C++.

2021-08-23  Jakub Jelinek  <jakub@redhat.com>

gcc/
	* tree.h (OMP_CLAUSE_GRAINSIZE_STRICT): Define.
	(OMP_CLAUSE_NUM_TASKS_STRICT): Define.
	* tree-pretty-print.c (dump_omp_clause) <case OMP_CLAUSE_GRAINSIZE,
	case OMP_CLAUSE_NUM_TASKS>: Print strict: modifier.
	* omp-expand.c (expand_task_call): Use GOMP_TASK_FLAG_STRICT in iflags
	if either grainsize or num_tasks clause has the strict modifier.
gcc/c/
	* c-parser.c (c_parser_omp_clause_num_tasks,
	c_parser_omp_clause_grainsize): Parse the optional strict: modifier.
gcc/cp/
	* parser.c (cp_parser_omp_clause_num_tasks,
	cp_parser_omp_clause_grainsize): Parse the optional strict: modifier.
include/
	* gomp-constants.h (GOMP_TASK_FLAG_STRICT): Define.
libgomp/
	* taskloop.c (GOMP_taskloop): Handle GOMP_TASK_FLAG_STRICT.
	* testsuite/libgomp.c-c++-common/taskloop-4.c (main): Fix up comment.
	* testsuite/libgomp.c-c++-common/taskloop-5.c: New test.

1 files changed, 23 insertions, 4 deletions

diff --git a/libgomp/taskloop.c b/libgomp/taskloop.c
index 791178a..9d27dd0 100644
--- a/libgomp/taskloop.c
+++ b/libgomp/taskloop.c
@@ -97,6 +97,7 @@ GOMP_taskloop (void (*fn) (void *), void *data, void (*cpyfn) (void *, void *),
 #endif
 
   TYPE task_step = step;
+  TYPE nfirst_task_step = step;
   unsigned long nfirst = n;
   if (flags & GOMP_TASK_FLAG_GRAINSIZE)
     {
@@ -109,7 +110,22 @@ GOMP_taskloop (void (*fn) (void *), void *data, void (*cpyfn) (void *, void *),
       if (num_tasks != ndiv)
 	num_tasks = ~0UL;
 #endif
-      if (num_tasks <= 1)
+      if ((flags & GOMP_TASK_FLAG_STRICT)
+	  && num_tasks != ~0ULL)
+	{
+	  UTYPE mod = n % grainsize;
+	  task_step = (TYPE) grainsize * step;
+	  if (mod)
+	    {
+	      num_tasks++;
+	      nfirst_task_step = (TYPE) mod * step;
+	      if (num_tasks == 1)
+		task_step = nfirst_task_step;
+	      else
+		nfirst = num_tasks - 2;
+	    }
+	}
+      else if (num_tasks <= 1)
 	{
 	  num_tasks = 1;
 	  task_step = end - start;
@@ -124,6 +140,7 @@ GOMP_taskloop (void (*fn) (void *), void *data, void (*cpyfn) (void *, void *),
 	  task_step = (TYPE) grainsize * step;
 	  if (mul != n)
 	    {
+	      nfirst_task_step = task_step;
 	      task_step += step;
 	      nfirst = n - mul - 1;
 	    }
@@ -135,6 +152,7 @@ GOMP_taskloop (void (*fn) (void *), void *data, void (*cpyfn) (void *, void *),
 	  task_step = (TYPE) div * step;
 	  if (mod)
 	    {
+	      nfirst_task_step = task_step;
 	      task_step += step;
 	      nfirst = mod - 1;
 	    }
@@ -153,6 +171,7 @@ GOMP_taskloop (void (*fn) (void *), void *data, void (*cpyfn) (void *, void *),
 	  task_step = (TYPE) div * step;
 	  if (mod)
 	    {
+	      nfirst_task_step = task_step;
 	      task_step += step;
 	      nfirst = mod - 1;
 	    }
@@ -225,7 +244,7 @@ GOMP_taskloop (void (*fn) (void *), void *data, void (*cpyfn) (void *, void *),
 	      start += task_step;
 	      ((TYPE *)arg)[1] = start;
 	      if (i == nfirst)
-		task_step -= step;
+		task_step = nfirst_task_step;
 	      fn (arg);
 	      arg += arg_size;
 	      if (!priority_queue_empty_p (&task[i].children_queue,
@@ -258,7 +277,7 @@ GOMP_taskloop (void (*fn) (void *), void *data, void (*cpyfn) (void *, void *),
 	    start += task_step;
 	    ((TYPE *)data)[1] = start;
 	    if (i == nfirst)
-	      task_step -= step;
+	      task_step = nfirst_task_step;
 	    fn (data);
 	    if (!priority_queue_empty_p (&task.children_queue,
 					 MEMMODEL_RELAXED))
@@ -303,7 +322,7 @@ GOMP_taskloop (void (*fn) (void *), void *data, void (*cpyfn) (void *, void *),
 	  start += task_step;
 	  ((TYPE *)arg)[1] = start;
 	  if (i == nfirst)
-	    task_step -= step;
+	    task_step = nfirst_task_step;
 	  thr->task = parent;
 	  task->kind = GOMP_TASK_WAITING;
 	  task->fn = fn;