https://gcc.gnu.org/bugzilla/show_bug.cgi?id=124649
Andrew Pinski <pinskia at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|UNCONFIRMED |NEW
Known to fail| |16.0
Ever confirmed|0 |1
Known to work| |15.2.0
Summary|ICE during RTL |[16 Regression] ICE during
|cprop_hardreg |RTL cprop_hardreg
Component|target |rtl-optimization
Target Milestone|--- |16.0
Last reconfirmed| |2026-03-26
--- Comment #2 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
#2 0x0000000001736ae3 in partial_subreg_p (outermode=E_DImode,
innermode=E_VNx4QImode) at ../../gcc/rtl.h:3252
#4 0x0000000001e4bd0c in maybe_mode_change (orig_mode=E_DImode,
copy_mode=E_DImode, new_mode=E_VNx4QImode, regno=2, copy_regno=63) at
../../gcc/regcprop.cc:452
```
(insn 91 36 94 3 (set (reg:DI 63 v31 [orig:103 _5 ] [103])
(reg:DI 2 x2 [161])) "t6.c":13:5 105 {*movdi_aarch64}
(expr_list:REG_DEAD (reg:DI 2 x2 [161])
(nil)))
(insn 94 91 40 3 (set (reg:VNx4QI 62 v30 [orig:147 vect__10.18_30 ] [147])
(reg:VNx4QI 63 v31 [orig:103 _5 ] [103])) "t6.c":16:12 discrim 33792
5627 {*aarch64_sve_movvnx4qi_no_ldr_str}
(nil))
(insn:TI 40 94 39 3 (set (mem:SI (reg/f:DI 1 x1 [138]) [1 (*c.4_21)[0]+0 S4
A32])
(reg:SI 63 v31 [orig:103 _5 ] [103])) "t6.c":14:8 104 {*movsi_aarch64}
(expr_list:REG_DEAD (reg/f:DI 1 x1 [138])
(nil)))
```
This sequence does not show up in GCC 15.2.0 so it was ok 15.2.0 at least.
