public inbox for gcc-bugs@sourceware.org
help / color / mirror / Atom feed
* [Bug c/111765] New: RISC-V: Faild to vectorize gen-vect-34.c
@ 2023-10-10 22:38 juzhe.zhong at rivai dot ai
2023-10-11 2:31 ` [Bug tree-optimization/111765] " pinskia at gcc dot gnu.org
` (2 more replies)
0 siblings, 3 replies; 4+ messages in thread
From: juzhe.zhong at rivai dot ai @ 2023-10-10 22:38 UTC (permalink / raw)
To: gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=111765
Bug ID: 111765
Summary: RISC-V: Faild to vectorize gen-vect-34.c
Product: gcc
Version: 14.0
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: c
Assignee: unassigned at gcc dot gnu.org
Reporter: juzhe.zhong at rivai dot ai
Target Milestone: ---
https://godbolt.org/z/17qYYKWPM
float summul(int n, float *arg1, float *arg2)
{
int i;
float res1 = 1.0;
for(i = 0; i < n; i++) {
if(arg2[i])
res1 *= arg1[i];
}
return res1;
}
I found both ARM SVE and RVV failed to vectorize this case wheras X86 can
vectorize it with -mavx2.
It seems that we need some tricks on both SVE and RVV backend ?
^ permalink raw reply [flat|nested] 4+ messages in thread
* [Bug tree-optimization/111765] RISC-V: Faild to vectorize gen-vect-34.c
2023-10-10 22:38 [Bug c/111765] New: RISC-V: Faild to vectorize gen-vect-34.c juzhe.zhong at rivai dot ai
@ 2023-10-11 2:31 ` pinskia at gcc dot gnu.org
2023-10-11 2:33 ` juzhe.zhong at rivai dot ai
2023-10-11 2:35 ` juzhe.zhong at rivai dot ai
2 siblings, 0 replies; 4+ messages in thread
From: pinskia at gcc dot gnu.org @ 2023-10-11 2:31 UTC (permalink / raw)
To: gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=111765
--- Comment #1 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
For SVE:
/app/example.cpp:6:18: missed: missing target support for reduction on
variable-length vectors.
I assume it is the same issue for RVV.
^ permalink raw reply [flat|nested] 4+ messages in thread
* [Bug tree-optimization/111765] RISC-V: Faild to vectorize gen-vect-34.c
2023-10-10 22:38 [Bug c/111765] New: RISC-V: Faild to vectorize gen-vect-34.c juzhe.zhong at rivai dot ai
2023-10-11 2:31 ` [Bug tree-optimization/111765] " pinskia at gcc dot gnu.org
@ 2023-10-11 2:33 ` juzhe.zhong at rivai dot ai
2023-10-11 2:35 ` juzhe.zhong at rivai dot ai
2 siblings, 0 replies; 4+ messages in thread
From: juzhe.zhong at rivai dot ai @ 2023-10-11 2:33 UTC (permalink / raw)
To: gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=111765
--- Comment #2 from JuzheZhong <juzhe.zhong at rivai dot ai> ---
(In reply to Andrew Pinski from comment #1)
> For SVE:
> /app/example.cpp:6:18: missed: missing target support for reduction on
> variable-length vectors.
>
> I assume it is the same issue for RVV.
Yeah.
So may be we should adjust the dump check:
/* { dg-final { scan-tree-dump-times "vectorized 1 loops" 1 "vect" { target
vect_masked_load && !vect_variable_length } } } */
?
^ permalink raw reply [flat|nested] 4+ messages in thread
* [Bug tree-optimization/111765] RISC-V: Faild to vectorize gen-vect-34.c
2023-10-10 22:38 [Bug c/111765] New: RISC-V: Faild to vectorize gen-vect-34.c juzhe.zhong at rivai dot ai
2023-10-11 2:31 ` [Bug tree-optimization/111765] " pinskia at gcc dot gnu.org
2023-10-11 2:33 ` juzhe.zhong at rivai dot ai
@ 2023-10-11 2:35 ` juzhe.zhong at rivai dot ai
2 siblings, 0 replies; 4+ messages in thread
From: juzhe.zhong at rivai dot ai @ 2023-10-11 2:35 UTC (permalink / raw)
To: gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=111765
--- Comment #3 from JuzheZhong <juzhe.zhong at rivai dot ai> ---
If we specify the vector length to RVV and SVE:
https://godbolt.org/z/njvsqYWhn
Both can vectorize.
ARM SVE: -msve-vector-bits=128
RVV: --param=riscv-autovec-preference=fixed-vlmax
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2023-10-11 2:35 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2023-10-10 22:38 [Bug c/111765] New: RISC-V: Faild to vectorize gen-vect-34.c juzhe.zhong at rivai dot ai
2023-10-11 2:31 ` [Bug tree-optimization/111765] " pinskia at gcc dot gnu.org
2023-10-11 2:33 ` juzhe.zhong at rivai dot ai
2023-10-11 2:35 ` juzhe.zhong at rivai dot ai
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).