https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106322
--- Comment #44 from Kewen Lin <linkw at gcc dot gnu.org> ---
(In reply to Richard Biener from comment #43)
> (In reply to Richard Biener from comment #42)
> > I think this goes wrong in vectorizable_operation which does
> >
> > if (using_emulated_vectors_p
> > && !vect_can_vectorize_without_simd_p (code))
> >
> > to guard this but I'm not sure how this slips through?
>
> Ah, it's an internal function. I think we should simply return false
> during analysis for any vect_emulated_vector_p type in vectorizable_call.
>
> Alternatively pattern recognition could also be made to fail but the above
> is definitely more future proof.
Thanks for the pointer! I think you meant:
diff --git a/gcc/tree-vect-stmts.cc b/gcc/tree-vect-stmts.cc
index c9534ef9b1e..ee10fa3e0fb 100644
--- a/gcc/tree-vect-stmts.cc
+++ b/gcc/tree-vect-stmts.cc
@@ -3388,6 +3388,14 @@ vectorizable_call (vec_info *vinfo,
return false;
}
+ if (vect_emulated_vector_p (vectype_in) || vect_emulated_vector_p
(vectype_out))
+ {
+ if (dump_enabled_p ())
+ dump_printf_loc (MSG_MISSED_OPTIMIZATION, vect_location,
+ "use emulated vector type for call\n");
+ return false;
+ }
+
/* FORNOW */
nunits_in = TYPE_VECTOR_SUBPARTS (vectype_in);
nunits_out = TYPE_VECTOR_SUBPARTS (vectype_out);
Will kick off some testings on x64/aarch64/ppc64{,le} and post it later.
