https://gcc.gnu.org/bugzilla/show_bug.cgi?id=121668
--- Comment #3 from Sam James <sjames at gcc dot gnu.org> ---
(In reply to Sam James from comment #1)
> -g is important. Is it complaining about a debug insn and the assert should
> just be tweaked?
(gdb) s
NEXT_INSN (insn=0x7fffe93f0578) at
/usr/src/debug/sys-devel/gcc-16.0.9999/gcc-16.0.9999/gcc/rtl.h:1491
1491 rtx next = XEXP (insn, 1);
(gdb) call debug(insn)
(note 2 3 5 2 NOTE_INSN_FUNCTION_BEG)
(gdb) s
1492 return safe_as_a <rtx_insn *> (next);
(gdb) call debug(next)
(debug_insn 5 2 16 2 (debug_marker) "major_gc.bpic.i":6:3 -1
(nil))
(gdb) s
268 gcc_checking_assert (is_a <T> (p));
(gdb) p p
$3 = (rtx_def *) 0x7fffe9216440
(gdb) p debug(p)
(debug_insn 5 2 16 2 (debug_marker) "major_gc.bpic.i":6:3 -1
(nil))
$4 = void
(gdb) s
fancy_abort (file=0x5555586b7a38
"/var/tmp/portage/sys-devel/gcc-16.0.9999/work/gcc-16.0.9999/gcc/config/i386/i386-features.cc",
line=3810, function=0x555558696283 "ix86_emit_tls_call")
at
/usr/src/debug/sys-devel/gcc-16.0.9999/gcc-16.0.9999/gcc/diagnostics/context.cc:1654
1654 {
