[Frontend][OpenMP] Allow implicit clauses to fail to apply

[kparzysz]

The linear(x) clause implies firstprivate(x) on the compound construct if x is not an induction variable. With more construct combinations coming in OpenMP 6.0, the firstprivate clause may not be possible to apply, e.g. in "masked simd".
An additional benefit from this change is that it allows treating leaf constructs as combined constructs with a single constituent. Otherwise, a linear clause on a lone simd construct could imply a firstprivate clause that can't be applied.

[llvmbot]

@llvm/pr-subscribers-flang-openmp

Author: Krzysztof Parzyszek (kparzysz)

Changes

The linear(x) clause implies firstprivate(x) on the compound construct if x is not an induction variable. With more construct combinations coming in OpenMP 6.0, the firstprivate clause may not be possible to apply, e.g. in "masked simd".
An additional benefit from this change is that it allows treating leaf constructs as combined constructs with a single constituent. Otherwise, a linear clause on a simd construct could imply a firstprivate clause that can't be applied.

---

Full diff: https://github.com/llvm/llvm-project/pull/100460.diff

1 Files Affected:

• (modified) llvm/include/llvm/Frontend/OpenMP/ConstructDecompositionT.h (+8-2)

diff --git a/llvm/include/llvm/Frontend/OpenMP/ConstructDecompositionT.h b/llvm/include/llvm/Frontend/OpenMP/ConstructDecompositionT.h
index 349d862135d8c..b93bc594a82bf 100644
--- a/llvm/include/llvm/Frontend/OpenMP/ConstructDecompositionT.h
+++ b/llvm/include/llvm/Frontend/OpenMP/ConstructDecompositionT.h
@@ -1114,6 +1114,11 @@ bool ConstructDecompositionT<C, H>::applyClause(
 template <typename C, typename H> bool ConstructDecompositionT<C, H>::split() {
   bool success = true;
 
+  auto isImplicit = [this](const ClauseTy *node) {
+    return llvm::any_of(
+        implicit, [node](const ClauseTy &clause) { return &clause == node; });
+  };
+
   for (llvm::omp::Directive leaf :
        llvm::omp::getLeafConstructsOrSelf(construct))
     leafs.push_back(LeafReprInternal{leaf, /*clauses=*/{}});
@@ -1153,9 +1158,10 @@ template <typename C, typename H> bool ConstructDecompositionT<C, H>::split() {
   for (const ClauseTy *node : nodes) {
     if (skip(node))
       continue;
-    success =
-        success &&
+    bool result =
         std::visit([&](auto &&s) { return applyClause(s, node); }, node->u);
+    if (!isImplicit(node))
+      success = success && result;
   }
 
   // Apply "allocate".

[skatrak]

It makes sense to me, but I think we should have a test to check it before merging.

[skatrak]

Thank you, LGTM.