[PATCH] D38417: [test-suite] Adding HACCKernels app

Hal Finkel via Phabricator via llvm-commits llvm-commits at lists.llvm.org
Mon Oct 2 16:54:16 PDT 2017


hfinkel added inline comments.


================
Comment at: MultiSource/Benchmarks/DOE-ProxyApps-C++/HACCKernels/GravityForceKernel.cpp:101
+#pragma omp simd reduction(+:lax,lay,laz)
+#endif
+  for (int i = 0; i < n; ++i) {
----------------
As Brian and I discussed offline, I've suggested that we update this to read:

#if _OPENMP >= 201307
#pragma omp simd reduction(+:lax,lay,laz)
#elsif defined(__clang__)
#pragma clang loop vectorize(assume_safety)
#endif

So that we'll vectorize the loop. Looking at this again today, it seems like the problem blocking loop vectorization here is just that we don't self report, by default, a new enough OpenMP version. I suspect that bumping that version is blocked on unrelated things.


https://reviews.llvm.org/D38417





More information about the llvm-commits mailing list