[Bug middle-end/124108] [16 Regression] ICE: in expand_expr_real_2, at expr.cc:10776 with bitint rotate at -O

[rguenth at gcc dot gnu.org via Gcc-bugs]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=124108

Richard Biener <rguenth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2026-02-16
           Priority|P3                          |P1
     Ever confirmed|0                           |1
      Known to work|                            |15.2.1

--- Comment #1 from Richard Biener <rguenth at gcc dot gnu.org> ---
Confirmed.

10774       case LROTATE_EXPR:
10775       case RROTATE_EXPR:
10776         gcc_assert (VECTOR_MODE_P (TYPE_MODE (type))
10777                     || type_has_mode_precision_p (type));

gdb) p debug_tree (type)
 <bitint_type 0x7ffff69c72a0 public unsigned QI
    size <integer_cst 0x7ffff681f060 type <integer_type 0x7ffff681d0a8
bitsizetype> constant 8>
    unit-size <integer_cst 0x7ffff681f078 type <integer_type 0x7ffff681d000
sizetype> constant 1>
    align:8 warn_if_not_align:0 symtab:0 alias-set 1 canonical-type
0x7ffff69c72a0 precision:4 min <integer_cst 0x7ffff69db450 0> max <integer_cst
0x7ffff69db468 15>
    pointer_to_this <pointer_type 0x7ffff69c7150>>

I'll note this isn't a restriction we enforce during GIMPLE type checking.