[clang] [llvm] [clang][OpenMP] Add no-loop SPMD kernel promotion (PR #224041)

[Nicole Aschenbrenner via cfe-commits]

================
@@ -0,0 +1,91 @@
+// clang-format off
+// C counterpart of fortran/target-no-loop.f90.
+
+// RUN: %libomptarget-compile-generic -O3 -fopenmp-assume-threads-oversubscription -fopenmp-assume-teams-oversubscription
+// RUN: env LIBOMPTARGET_INFO=16 OMP_NUM_TEAMS=16 OMP_TEAMS_THREAD_LIMIT=16 %libomptarget-run-generic 2>&1 | %fcheck-generic
----------------
nicebert wrote:

The test is a straight port of the existing Fortran one (fortran/target-no-loop.f90), so I added the run without OMP_NUM_TEAMS/OMP_TEAMS_THREAD_LIMIT to both in a follow-up, #229677. It checks that every kernel still launches in the same mode with the runtime's default grid.


https://github.com/llvm/llvm-project/pull/224041