https://gcc.gnu.org/bugzilla/show_bug.cgi?id=119210
--- Comment #7 from xiezhiheng at huawei dot com ---
For other information,
https://godbolt.org/z/xdPYGsjYd
LLVM seems always dominate block .LBB0_14
.LBB0_11:
add x23, x23, #1
msr TPIDR2_EL0, xzr
cmp x23, #6
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=119210
--- Comment #5 from xiezhiheng at huawei dot com ---
(In reply to Andrew Pinski from comment #3)
> So:
> mrs x16, tpidr2_el0
> cbnzx16, .L22 <== it will branch to .L22, and miss 'smstart za'
> mov x0, x3
>
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=119210
--- Comment #4 from xiezhiheng at huawei dot com ---
(In reply to Andrew Pinski from comment #3)
> So:
> mrs x16, tpidr2_el0
> cbnzx16, .L22 <== it will branch to .L22, and miss 'smstart za'
> mov x0, x3
>
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=119210
--- Comment #2 from xiezhiheng at huawei dot com ---
(In reply to Andrew Pinski from comment #1)
> Created attachment 60705 [details]
> testcase
>
> -O2-march=armv9-a+sve+sve2+sme-f64f64
>
>
> Next time please attach the testcase (the att
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=119210
Bug ID: 119210
Summary: [SME] 'smstart za' seems not to dominate the block
that uses za register
Product: gcc
Version: 15.0
Status: UNCONFIRMED
Severity: norma
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94442
--- Comment #9 from xiezhiheng at huawei dot com ---
(In reply to Jakub Jelinek from comment #8)
> So, is this fixed by any of the
> r11-2190-gbf592b2ff776aef71c91924cdb5e0d10488496cf
> r11-2448-g072a8b8fb6e861d8ac2db847bcc81dbcb1ef1b35
> r11-2554