This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: V2 [PATCH] i386: Use scalar operand in SF/DF/SI/DI vec_dup patterns
- From: Uros Bizjak <ubizjak at gmail dot com>
- To: "H. J. Lu" <hjl dot tools at gmail dot com>
- Cc: "gcc-patches at gcc dot gnu dot org" <gcc-patches at gcc dot gnu dot org>
- Date: Thu, 1 Nov 2018 18:55:06 +0100
- Subject: Re: V2 [PATCH] i386: Use scalar operand in SF/DF/SI/DI vec_dup patterns
- References: <20181027055846.1682-1-hjl.tools@gmail.com> <CAFULd4ZQGB07pKBBMzRzVmNZBOxU7JDEOcST0qj25K_=jDf7oQ@mail.gmail.com> <CAMe9rOqPNWNZMM4J6S1cKDj0Do0Ahq6SsKRuiRivTkibDN_6xw@mail.gmail.com>
On Tue, Oct 30, 2018 at 9:36 PM H.J. Lu <hjl.tools@gmail.com> wrote:
> > Why not at the expand time? Rewrite vector constant as vec_duplicate
> > from memory and combine will do the stuff for you. We do have _bcst
> > instruction patterns.
> >
>
> Here is the updated patch to do that. OK for trunk?
Looks mostly good, however IMO we should move the expansion out of
ix86_expand_args_builtin. We have an example of
ix86_expand_unop_vec_merge_builtin and VEC_MERGE type in
i386-builtin-types.def on how to handle special builtins.
Uros.
> Thanks.
>
>
> --
> H.J.