public inbox for gcc-bugs@sourceware.org
help / color / mirror / Atom feed
* [Bug target/113871] New: psrlq is not used for PERM<a,{0},1,2,3,4>
@ 2024-02-11 5:23 pinskia at gcc dot gnu.org
2024-02-11 5:32 ` [Bug target/113871] " pinskia at gcc dot gnu.org
` (7 more replies)
0 siblings, 8 replies; 9+ messages in thread
From: pinskia at gcc dot gnu.org @ 2024-02-11 5:23 UTC (permalink / raw)
To: gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113871
Bug ID: 113871
Summary: psrlq is not used for PERM<a,{0},1,2,3,4>
Product: gcc
Version: 14.0
Status: UNCONFIRMED
Keywords: missed-optimization
Severity: enhancement
Priority: P3
Component: target
Assignee: unassigned at gcc dot gnu.org
Reporter: pinskia at gcc dot gnu.org
Target Milestone: ---
Target: x86_64
Take:
```
#define vect64 __attribute__((vector_size(8)))
void f(vect64unsigned short *a)
{
*a = __builtin_shufflevector(*a,(vect64 unsigned short){0}, 1,2,3, 4);
}
```
This should just produce:
```
movq (%rdi), %xmm0
psrlq $16, %xmm0, %xmm0
movq %xmm0, (%rdi)
retq
```
But instead we get:
```
movzwl 6(%rdi), %eax
movzwl 4(%rdi), %edx
salq $16, %rax
orq %rdx, %rax
movzwl 2(%rdi), %edx
salq $16, %rax
orq %rdx, %rax
movq %rax, (%rdi)
ret
```
With AVX enabled we get slightly better:
```
f:
.LFB0:
.cfi_startproc
vmovq (%rdi), %xmm0
vpxor %xmm1, %xmm1, %xmm1
vpshufb .LC1(%rip), %xmm1, %xmm1
vpshufb .LC0(%rip), %xmm0, %xmm0
vpor %xmm1, %xmm0, %xmm0
vmovq %xmm0, (%rdi)
ret
```
Note LLVM is able to catch this for x86_64 (for aarch64, GCC is able to use
`ushr d31, d31, 16` while LLVM does not).
I suspect vec_shr_<mode> pattern is missing and once it is added, it will just
work.
^ permalink raw reply [flat|nested] 9+ messages in thread
* [Bug target/113871] psrlq is not used for PERM<a,{0},1,2,3,4>
2024-02-11 5:23 [Bug target/113871] New: psrlq is not used for PERM<a,{0},1,2,3,4> pinskia at gcc dot gnu.org
@ 2024-02-11 5:32 ` pinskia at gcc dot gnu.org
2024-02-11 5:56 ` pinskia at gcc dot gnu.org
` (6 subsequent siblings)
7 siblings, 0 replies; 9+ messages in thread
From: pinskia at gcc dot gnu.org @ 2024-02-11 5:32 UTC (permalink / raw)
To: gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113871
Andrew Pinski <pinskia at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
See Also| |https://github.com/llvm/llv
| |m-project/issues/81393
--- Comment #1 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Note LLVM is not able to handle this for aarch64; file
https://github.com/llvm/llvm-project/issues/81393 for that.
^ permalink raw reply [flat|nested] 9+ messages in thread
* [Bug target/113871] psrlq is not used for PERM<a,{0},1,2,3,4>
2024-02-11 5:23 [Bug target/113871] New: psrlq is not used for PERM<a,{0},1,2,3,4> pinskia at gcc dot gnu.org
2024-02-11 5:32 ` [Bug target/113871] " pinskia at gcc dot gnu.org
@ 2024-02-11 5:56 ` pinskia at gcc dot gnu.org
2024-02-12 8:46 ` rguenth at gcc dot gnu.org
` (5 subsequent siblings)
7 siblings, 0 replies; 9+ messages in thread
From: pinskia at gcc dot gnu.org @ 2024-02-11 5:56 UTC (permalink / raw)
To: gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113871
--- Comment #2 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Note `PERM<{0},a,{1,2,3,4}>` should be handled too, that means defining
`vec_shl_` patterns too.
^ permalink raw reply [flat|nested] 9+ messages in thread
* [Bug target/113871] psrlq is not used for PERM<a,{0},1,2,3,4>
2024-02-11 5:23 [Bug target/113871] New: psrlq is not used for PERM<a,{0},1,2,3,4> pinskia at gcc dot gnu.org
2024-02-11 5:32 ` [Bug target/113871] " pinskia at gcc dot gnu.org
2024-02-11 5:56 ` pinskia at gcc dot gnu.org
@ 2024-02-12 8:46 ` rguenth at gcc dot gnu.org
2024-02-13 18:27 ` ubizjak at gmail dot com
` (4 subsequent siblings)
7 siblings, 0 replies; 9+ messages in thread
From: rguenth at gcc dot gnu.org @ 2024-02-12 8:46 UTC (permalink / raw)
To: gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113871
Richard Biener <rguenth at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Target|x86_64 |x86_64-*-* i?86-*-*
Ever confirmed|0 |1
Status|UNCONFIRMED |NEW
Last reconfirmed| |2024-02-12
--- Comment #3 from Richard Biener <rguenth at gcc dot gnu.org> ---
Confirmed.
^ permalink raw reply [flat|nested] 9+ messages in thread
* [Bug target/113871] psrlq is not used for PERM<a,{0},1,2,3,4>
2024-02-11 5:23 [Bug target/113871] New: psrlq is not used for PERM<a,{0},1,2,3,4> pinskia at gcc dot gnu.org
` (2 preceding siblings ...)
2024-02-12 8:46 ` rguenth at gcc dot gnu.org
@ 2024-02-13 18:27 ` ubizjak at gmail dot com
2024-02-14 8:16 ` ubizjak at gmail dot com
` (3 subsequent siblings)
7 siblings, 0 replies; 9+ messages in thread
From: ubizjak at gmail dot com @ 2024-02-13 18:27 UTC (permalink / raw)
To: gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113871
Uroš Bizjak <ubizjak at gmail dot com> changed:
What |Removed |Added
----------------------------------------------------------------------------
Assignee|unassigned at gcc dot gnu.org |ubizjak at gmail dot com
Status|NEW |ASSIGNED
--- Comment #4 from Uroš Bizjak <ubizjak at gmail dot com> ---
Created attachment 57417
--> https://gcc.gnu.org/bugzilla/attachment.cgi?id=57417&action=edit
Proposed patch
Patch in testing.
^ permalink raw reply [flat|nested] 9+ messages in thread
* [Bug target/113871] psrlq is not used for PERM<a,{0},1,2,3,4>
2024-02-11 5:23 [Bug target/113871] New: psrlq is not used for PERM<a,{0},1,2,3,4> pinskia at gcc dot gnu.org
` (3 preceding siblings ...)
2024-02-13 18:27 ` ubizjak at gmail dot com
@ 2024-02-14 8:16 ` ubizjak at gmail dot com
2024-02-14 19:44 ` cvs-commit at gcc dot gnu.org
` (2 subsequent siblings)
7 siblings, 0 replies; 9+ messages in thread
From: ubizjak at gmail dot com @ 2024-02-14 8:16 UTC (permalink / raw)
To: gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113871
Uroš Bizjak <ubizjak at gmail dot com> changed:
What |Removed |Added
----------------------------------------------------------------------------
Attachment #57417|0 |1
is obsolete| |
--- Comment #5 from Uroš Bizjak <ubizjak at gmail dot com> ---
Created attachment 57419
--> https://gcc.gnu.org/bugzilla/attachment.cgi?id=57419&action=edit
Proposed v2 patch
New version in testing, also handles 32-bit vectors.
^ permalink raw reply [flat|nested] 9+ messages in thread
* [Bug target/113871] psrlq is not used for PERM<a,{0},1,2,3,4>
2024-02-11 5:23 [Bug target/113871] New: psrlq is not used for PERM<a,{0},1,2,3,4> pinskia at gcc dot gnu.org
` (4 preceding siblings ...)
2024-02-14 8:16 ` ubizjak at gmail dot com
@ 2024-02-14 19:44 ` cvs-commit at gcc dot gnu.org
2024-02-14 19:48 ` ubizjak at gmail dot com
2024-02-27 17:42 ` cvs-commit at gcc dot gnu.org
7 siblings, 0 replies; 9+ messages in thread
From: cvs-commit at gcc dot gnu.org @ 2024-02-14 19:44 UTC (permalink / raw)
To: gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113871
--- Comment #6 from GCC Commits <cvs-commit at gcc dot gnu.org> ---
The master branch has been updated by Uros Bizjak <uros@gcc.gnu.org>:
https://gcc.gnu.org/g:2c2f57e4158924467afbf4c2fd3938e507287dab
commit r14-8989-g2c2f57e4158924467afbf4c2fd3938e507287dab
Author: Uros Bizjak <ubizjak@gmail.com>
Date: Wed Feb 14 20:41:42 2024 +0100
i386: psrlq is not used for PERM<a,{0},1,2,3,4> [PR113871]
Introduce vec_shl_<mode> and vec_shr_<mode> expanders to improve
'*a = __builtin_shufflevector(*a, (vect64){0}, 1, 2, 3, 4);'
and
'*a = __builtin_shufflevector((vect64){0}, *a, 3, 4, 5, 6);'
shuffles. The generated code improves from:
movzwl 6(%rdi), %eax
movzwl 4(%rdi), %edx
salq $16, %rax
orq %rdx, %rax
movzwl 2(%rdi), %edx
salq $16, %rax
orq %rdx, %rax
movq %rax, (%rdi)
to:
movq (%rdi), %xmm0
psrlq $16, %xmm0
movq %xmm0, (%rdi)
and to:
movq (%rdi), %xmm0
psllq $16, %xmm0
movq %xmm0, (%rdi)
in the second case.
The patch handles 32-bit vectors as well and improves generated code from:
movd (%rdi), %xmm0
pxor %xmm1, %xmm1
punpcklwd %xmm1, %xmm0
pshuflw $230, %xmm0, %xmm0
movd %xmm0, (%rdi)
to:
movd (%rdi), %xmm0
psrld $16, %xmm0
movd %xmm0, (%rdi)
and to:
movd (%rdi), %xmm0
pslld $16, %xmm0
movd %xmm0, (%rdi)
PR target/113871
gcc/ChangeLog:
* config/i386/mmx.md (V248FI): New mode iterator.
(V24FI_32): DItto.
(vec_shl_<V248FI:mode>): New expander.
(vec_shl_<V24FI_32:mode>): Ditto.
(vec_shr_<V248FI:mode>): Ditto.
(vec_shr_<V24FI_32:mode>): Ditto.
* config/i386/sse.md (vec_shl_<V_128:mode>): Simplify expander.
(vec_shr_<V248FI:mode>): Ditto.
gcc/testsuite/ChangeLog:
* gcc.target/i386/pr113871-1a.c: New test.
* gcc.target/i386/pr113871-1b.c: New test.
* gcc.target/i386/pr113871-2a.c: New test.
* gcc.target/i386/pr113871-2b.c: New test.
* gcc.target/i386/pr113871-3a.c: New test.
* gcc.target/i386/pr113871-3b.c: New test.
* gcc.target/i386/pr113871-4a.c: New test.
^ permalink raw reply [flat|nested] 9+ messages in thread
* [Bug target/113871] psrlq is not used for PERM<a,{0},1,2,3,4>
2024-02-11 5:23 [Bug target/113871] New: psrlq is not used for PERM<a,{0},1,2,3,4> pinskia at gcc dot gnu.org
` (5 preceding siblings ...)
2024-02-14 19:44 ` cvs-commit at gcc dot gnu.org
@ 2024-02-14 19:48 ` ubizjak at gmail dot com
2024-02-27 17:42 ` cvs-commit at gcc dot gnu.org
7 siblings, 0 replies; 9+ messages in thread
From: ubizjak at gmail dot com @ 2024-02-14 19:48 UTC (permalink / raw)
To: gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113871
Uroš Bizjak <ubizjak at gmail dot com> changed:
What |Removed |Added
----------------------------------------------------------------------------
Resolution|--- |FIXED
Target Milestone|--- |14.0
Status|ASSIGNED |RESOLVED
--- Comment #7 from Uroš Bizjak <ubizjak at gmail dot com> ---
Implemented for gcc-14.
^ permalink raw reply [flat|nested] 9+ messages in thread
* [Bug target/113871] psrlq is not used for PERM<a,{0},1,2,3,4>
2024-02-11 5:23 [Bug target/113871] New: psrlq is not used for PERM<a,{0},1,2,3,4> pinskia at gcc dot gnu.org
` (6 preceding siblings ...)
2024-02-14 19:48 ` ubizjak at gmail dot com
@ 2024-02-27 17:42 ` cvs-commit at gcc dot gnu.org
7 siblings, 0 replies; 9+ messages in thread
From: cvs-commit at gcc dot gnu.org @ 2024-02-27 17:42 UTC (permalink / raw)
To: gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113871
--- Comment #8 from GCC Commits <cvs-commit at gcc dot gnu.org> ---
The master branch has been updated by Uros Bizjak <uros@gcc.gnu.org>:
https://gcc.gnu.org/g:15d1dae0d4d1be88d28ad7578a60fd3e36de36d8
commit r14-9198-g15d1dae0d4d1be88d28ad7578a60fd3e36de36d8
Author: Uros Bizjak <ubizjak@gmail.com>
Date: Tue Feb 27 18:41:24 2024 +0100
i386: psrlq is not used for PERM<a,{0},1,2,3,4> [PR113871]
Also handle V2BF mode.
PR target/113871
gcc/ChangeLog:
* config/i386/mmx.md (V248FI): Add V2BF mode.
(V24FI_32): Ditto.
gcc/testsuite/ChangeLog:
* gcc.target/i386/pr113871-5a.c: New test.
* gcc.target/i386/pr113871-5b.c: New test.
^ permalink raw reply [flat|nested] 9+ messages in thread
end of thread, other threads:[~2024-02-27 17:42 UTC | newest]
Thread overview: 9+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2024-02-11 5:23 [Bug target/113871] New: psrlq is not used for PERM<a,{0},1,2,3,4> pinskia at gcc dot gnu.org
2024-02-11 5:32 ` [Bug target/113871] " pinskia at gcc dot gnu.org
2024-02-11 5:56 ` pinskia at gcc dot gnu.org
2024-02-12 8:46 ` rguenth at gcc dot gnu.org
2024-02-13 18:27 ` ubizjak at gmail dot com
2024-02-14 8:16 ` ubizjak at gmail dot com
2024-02-14 19:44 ` cvs-commit at gcc dot gnu.org
2024-02-14 19:48 ` ubizjak at gmail dot com
2024-02-27 17:42 ` cvs-commit at gcc dot gnu.org
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).