https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106081
Richard Biener changed:
What|Removed |Added
Status|ASSIGNED|RESOLVED
Resolution|---
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106081
--- Comment #16 from CVS Commits ---
The master branch has been updated by Richard Biener :
https://gcc.gnu.org/g:5d09fb683a8abce49dc0992f5102aa0189f8f632
commit r14-2790-g5d09fb683a8abce49dc0992f5102aa0189f8f632
Author: Richard Biener
Date:
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106081
--- Comment #15 from rguenther at suse dot de ---
On Wed, 26 Jul 2023, rsandifo at gcc dot gnu.org wrote:
> https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106081
>
> --- Comment #14 from rsandifo at gcc dot gnu.org gnu.org> ---
> FWIW,
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106081
--- Comment #14 from rsandifo at gcc dot gnu.org
---
FWIW, changing:
if (!STMT_VINFO_GROUPED_ACCESS (dr_stmt))
continue;
to:
if (!STMT_VINFO_GROUPED_ACCESS (dr_stmt))
{
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106081
--- Comment #13 from rsandifo at gcc dot gnu.org
---
(In reply to Richard Biener from comment #12)
> Btw, I see we actually materialize a permute before the splat:
>
> t.c:14:24: note: node 0x5b311c0 (max_nunits=1, refcnt=2) vector(2)
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106081
--- Comment #12 from Richard Biener ---
Btw, I see we actually materialize a permute before the splat:
t.c:14:24: note: node 0x5b311c0 (max_nunits=1, refcnt=2) vector(2) double
t.c:14:24: note: op: VEC_PERM_EXPR
t.c:14:24: note:
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106081
--- Comment #11 from Richard Biener ---
(In reply to rsand...@gcc.gnu.org from comment #10)
> (In reply to Richard Biener from comment #9)
[...]
> > The following elides the unnecessary permutation for this special case
> > (but not the general
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106081
--- Comment #10 from rsandifo at gcc dot gnu.org
---
(In reply to Richard Biener from comment #9)
> So I can adjust change_layout_cost in a bit awkward way, but it seems that
> internal_node_cost would already work out that a permute can be
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106081
Richard Biener changed:
What|Removed |Added
Assignee|unassigned at gcc dot gnu.org |rguenth at gcc dot
gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106081
--- Comment #9 from Richard Biener ---
So I can adjust change_layout_cost in a bit awkward way, but it seems that
internal_node_cost would already work out that a permute can be merged into
an existing permute.
It seems that existing permutes
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106081
--- Comment #8 from Jan Hubicka ---
Imagemagick improved by 17% on zen3 and 11% on altra
https://lnt.opensuse.org/db_default/v4/SPEC/37550
https://lnt.opensuse.org/db_default/v4/SPEC/37543
which is cool :)
The loop is now optimized as:
.L2:
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106081
--- Comment #7 from rsandifo at gcc dot gnu.org
---
I don't think the splat creates a new layout, but instead a
splat should be allowed to change its layout at zero cost.
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106081
Richard Biener changed:
What|Removed |Added
CC||rsandifo at gcc dot gnu.org
---
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106081
Bug 106081 depends on bug 96208, which changed state.
Bug 96208 Summary: non-grouped load can be SLP vectorized for 2-element vectors
case
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=96208
What|Removed
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106081
Richard Biener changed:
What|Removed |Added
Depends on||96208
--- Comment #5 from Richard
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106081
Andrew Pinski changed:
What|Removed |Added
Severity|normal |enhancement
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106081
--- Comment #4 from Jan Hubicka ---
Thanks! It seems that imagemagick has quite few loops that inovlve consuming
shorts and producing doubles. Also it would be nice to do something about
__builtin_convertvector so it does not produce stupid
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106081
Richard Biener changed:
What|Removed |Added
Keywords||missed-optimization
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106081
Martin Liška changed:
What|Removed |Added
Blocks||26163
Ever confirmed|0
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106081
Hongtao.liu changed:
What|Removed |Added
CC||crazylht at gmail dot com
--- Comment #2
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106081
--- Comment #1 from Jan Hubicka ---
This is an attempt to vectorize by hand, but it seems we do not generate
vpmovsxwd for the vector short->double conversion
struct pixels
{
short a __attribute__ ((vector_size(4*2)));
} *pixels;
21 matches
Mail list logo