================
@@ -146,6 +146,14 @@ class CGOpenMPRuntimeGPU : public CGOpenMPRuntime {
bool isDelayedVariableLengthDecl(CodeGenFunction &CGF,
const VarDecl *VD) const override;
+ /// Check whether the target kernel being emitted is tagged SPMD_NO_LOOP, to
+ /// complete the promotion to a "no-loop" SPMD kernel, mirroring Flang's MLIR
+ /// promotion path.
+ bool canPromoteToNoLoop() const override {
+ return KernelAttrs.ExecFlags ==
----------------
dhruvachak wrote:
I don't like this cache. While today there is no reachable path that gets a
wrong answer, not having a directive passed in as the context is error-prone in
the future.
https://github.com/llvm/llvm-project/pull/224041
_______________________________________________
cfe-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
