[Bug tree-optimization/91201] [7/8/9/10 Regression] SIMD not generated for horizontal sum of bytes in array

[glisse at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=91201

--- Comment #9 from Marc Glisse <glisse at gcc dot gnu.org> ---
(In reply to Jakub Jelinek from comment #7)
> Untested patch to add the reduc_plus_scal_v{16,32,64}qi expanders.
> Wonder if we don't need also reduc_plus_scal_v8qi expander for
> TARGET_MMX_WITH_SSE.

I was even kind of hoping that reduc_plus_scal_v8qi would be the only one
needed, with v16qi reducing to it the same way you reduce v32qi and v64qi to
v16qi, but that's probably too optimistic because of MMX.