https://gcc.gnu.org/bugzilla/show_bug.cgi?id=101495
Bug ID: 101495
Summary: Unnecessary vzeroupper
Product: gcc
Version: 11.1.1
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: target
Assignee: unassigned at gcc dot gnu.org
Reporter: hjl.tools at gmail dot com
CC: crazylht at gmail dot com
Target Milestone: ---
Target: i386,x86-64
[hjl@gnu-cfl-2 pr101456]$ cat y.c
#include <x86intrin.h>
extern __m256d x;
extern __m256d bar (void);
__m256d
foo (void)
{
x = _mm256_setzero_pd ();
return bar ();
}
[hjl@gnu-cfl-2 pr101456]$ gcc -S -O2 -mavx2 y.c
[hjl@gnu-cfl-2 pr101456]$ cat y.s
.file "y.c"
.text
.p2align 4
.globl foo
.type foo, @function
foo:
.LFB5667:
.cfi_startproc
vxorpd %xmm0, %xmm0, %xmm0
vmovapd %ymm0, x(%rip)
vzeroupper
jmp bar
.cfi_endproc
.LFE5667:
.size foo, .-foo
.ident "GCC: (GNU) 11.1.1 20210531 (Red Hat 11.1.1-3)"
.section .note.GNU-stack,"",@progbits
[hjl@gnu-cfl-2 pr101456]$
Since bar returns __m256d, it must be compiled with AVX and there is no
need for vzeroupper.
