Hi Sandra!
Commenting on just one single item:
On 2022-09-14T11:32:11-0600, Sandra Loosemore <san...@codesourcery.com> wrote:
> --- a/gcc/omp-simd-clone.cc
> +++ b/gcc/omp-simd-clone.cc
> void
> expand_simd_clones (struct cgraph_node *node)
> {
> - tree attr = lookup_attribute ("omp declare simd",
> - DECL_ATTRIBUTES (node->decl));
> - if (attr == NULL_TREE
> - || node->inlined_to
> + tree attr;
> + bool error_p = true;
> +
> + if (node->inlined_to
> || lookup_attribute ("noclone", DECL_ATTRIBUTES (node->decl)))
> return;
>
> + attr = lookup_attribute ("omp declare simd",
> + DECL_ATTRIBUTES (node->decl));
> +
> + /* See if we can add an "omp declare simd" directive implicitly
> + before giving up. */
> + /* FIXME: OpenACC "#pragma acc routine" translates into
> + "omp declare target", but appears also to have some other effects
> + that conflict with generating SIMD clones, causing ICEs. So don't
> + do this if we've got OpenACC instead of OpenMP. */
Uh, ICEs... (But I suppose this processing is not relevant for OpenACC
'routine's.)
However, OpenACC and OpenMP support may be active at the same time...
> + if (attr == NULL_TREE
> + && flag_openmp_target_simd_clone && !flag_openacc)
..., so '!flag_openacc' is not the right check here. Instead you'd do
'!oacc_get_fn_attrib (DECL_ATTRIBUTES (node->decl))' (untested) or
similar.
> + {
> + attr = mark_auto_simd_clone (node);
> + error_p = false;
> + }
> + if (attr == NULL_TREE)
> + return;
Grüße
Thomas
-----------------
Siemens Electronic Design Automation GmbH; Anschrift: Arnulfstraße 201, 80634
München; Gesellschaft mit beschränkter Haftung; Geschäftsführer: Thomas
Heurung, Frank Thürauf; Sitz der Gesellschaft: München; Registergericht
München, HRB 106955
