[Openmp-commits] [openmp] Reland: [OpenMP][clang] 6.0: num_threads strict (part 1: host runtime) (PR #147532)

[Robert Imschweiler via Openmp-commits]

https://github.com/ro-i created https://github.com/llvm/llvm-project/pull/147532

OpenMP 6.0 12.1.2 specifies the behavior of the strict modifier for the num_threads clause on parallel directives, along with the message and severity clauses. This commit implements necessary host runtime changes.

Reland https://github.com/llvm/llvm-project/pull/146403. After manual testing on a gfx90a machine, I could not reproduce the failing test, which makes it even more likely that the test has just been flaky. (Or at least that it's not an issue related to this patch.)

>From b3e7d2eb642b67ab8fbfeeddd8e9178170b1a3dc Mon Sep 17 00:00:00 2001
From: Robert Imschweiler <robert.imschweiler at amd.com>
Date: Fri, 27 Jun 2025 07:54:07 -0500
Subject: [PATCH] [OpenMP][clang] 6.0: num_threads strict (part 1: host
 runtime)

OpenMP 6.0 12.1.2 specifies the behavior of the strict modifier for the
num_threads clause on parallel directives, along with the message and
severity clauses. This commit implements necessary host runtime changes.
---
 openmp/runtime/src/kmp_runtime.cpp | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/openmp/runtime/src/kmp_runtime.cpp b/openmp/runtime/src/kmp_runtime.cpp
index 417eceb8ebecc..6afea9b994de4 100644
--- a/openmp/runtime/src/kmp_runtime.cpp
+++ b/openmp/runtime/src/kmp_runtime.cpp
@@ -1214,6 +1214,12 @@ void __kmp_serialized_parallel(ident_t *loc, kmp_int32 global_tid) {
   // Reset for next parallel region
   this_thr->th.th_set_proc_bind = proc_bind_default;
 
+  // OpenMP 6.0 12.1.2 requires the num_threads 'strict' modifier to also have
+  // effect when parallel execution is disabled by a corresponding if clause
+  // attached to the parallel directive.
+  if (this_thr->th.th_nt_strict && this_thr->th.th_set_nproc > 1)
+    __kmpc_error(this_thr->th.th_nt_loc, this_thr->th.th_nt_sev,
+                 this_thr->th.th_nt_msg);
   // Reset num_threads for next parallel region
   this_thr->th.th_set_nproc = 0;