This patch series is a rework of the patch originally posted a couple of years ago: https://patchwork.sourceware.org/project/gcc/patch/0e1a740e-46d5-ebfa-36f4-9a069ddf8...@codesourcery.com/
The review comments from that time have been addressed, as have the comments from yesterday's review in the "[RFC] Enabling SVE with offloading to nvptx" thread on the gcc@ list (thanks Jakub). The first patch fixes the motivating issue with the BabelStream benchmark, and the rest fix up some similar cases. I've separated the patches to ease review and because I'm not completely sure if the IFN is overkill. I'm aware that Prathamesh is also working on code in this area. His RFC patch doesn't work for my use-case, and seems to have other issues. This patch conflicts, but hopefully it's not unresolvable. OK for mainline? Andrew Andrew Stubbs (4): openmp: Tune omp_max_vf for offload targets openmp: use offload max_vf for chunk_size openmp: Add IFN_GOMP_MAX_VF openmp: Add testcases for omp_max_vf gcc/gimple-loop-versioning.cc | 2 +- gcc/internal-fn.cc | 8 ++++ gcc/internal-fn.def | 1 + gcc/omp-expand.cc | 64 +++++++++++++++++++++------- gcc/omp-general.cc | 17 +++++++- gcc/omp-general.h | 2 +- gcc/omp-low.cc | 3 +- gcc/omp-offload.cc | 3 ++ gcc/testsuite/gcc.dg/gomp/gomp.exp | 14 ++++++ gcc/testsuite/gcc.dg/gomp/max_vf-1.c | 37 ++++++++++++++++ gcc/testsuite/gcc.dg/gomp/max_vf-2.c | 47 ++++++++++++++++++++ gcc/testsuite/gcc.dg/gomp/max_vf-3.c | 21 +++++++++ 12 files changed, 199 insertions(+), 20 deletions(-) create mode 100644 gcc/testsuite/gcc.dg/gomp/max_vf-1.c create mode 100644 gcc/testsuite/gcc.dg/gomp/max_vf-2.c create mode 100644 gcc/testsuite/gcc.dg/gomp/max_vf-3.c -- 2.46.0
