I reduced that to:

$ cat pr109711.i
void lspf2lpc();

int interpolate_lpc_q_0;

void
interpolate_lpc(int subframe_num) {
  float weight;
  if (interpolate_lpc_q_0)
    weight = subframe_num;
  else
    weight = 1.0;
  if (weight != 1.0)
    lspf2lpc();
}

void
qcelp_decode_frame() {
  int i;
  for (;; i++)
    interpolate_lpc(i);
}

$ /dev/shm/objdir2/gcc/xgcc -B/dev/shm/objdir2/gcc pr109711.i -c -O2
during IPA pass: inline
pr109711.i:21:1: internal compiler error: tree check: expected class ‘type’, have ‘exceptional’ (error_mark) in verify_range, at value-range.cc:1060
   21 | }
      | ^
0x84dfd7 tree_class_check_failed(tree_node const*, tree_code_class, char const*, int, char const*)
	/home/marxin/Programming/gcc/gcc/tree.cc:8949
0x85baff tree_class_check(tree_node*, tree_code_class, char const*, int, char const*)
	/home/marxin/Programming/gcc/gcc/tree.h:3663
0x85baff irange::verify_range()
	/home/marxin/Programming/gcc/gcc/value-range.cc:1060
0x1279d97 irange::operator=(irange const&)
	/home/marxin/Programming/gcc/gcc/value-range.cc:923
0xcd61a7 int_range<2u>::operator=(int_range<2u> const&)
	/home/marxin/Programming/gcc/gcc/value-range.h:817
0xcd61a7 evaluate_conditions_for_known_args
	/home/marxin/Programming/gcc/gcc/ipa-fnsummary.cc:517
0xcee9c5 do_estimate_edge_size(cgraph_edge*)
	/home/marxin/Programming/gcc/gcc/ipa-inline-analysis.cc:335
0xcf1465 estimate_edge_size(cgraph_edge*)
	/home/marxin/Programming/gcc/gcc/ipa-inline.h:79
0xcf1465 estimate_edge_growth(cgraph_edge*)
	/home/marxin/Programming/gcc/gcc/ipa-inline.h:100
0xcefe15 do_estimate_growth_1
	/home/marxin/Programming/gcc/gcc/ipa-inline-analysis.cc:434
0xcefef6 cgraph_node::call_for_symbol_and_aliases(bool (*)(cgraph_node*, void*), void*, bool)
	/home/marxin/Programming/gcc/gcc/cgraph.h:3412
0xcefef6 estimate_growth(cgraph_node*)
	/home/marxin/Programming/gcc/gcc/ipa-inline-analysis.cc:472
0x1d972c4 inline_small_functions
	/home/marxin/Programming/gcc/gcc/ipa-inline.cc:2042
0x1d972c4 ipa_inline
	/home/marxin/Programming/gcc/gcc/ipa-inline.cc:2811
0x1d972c4 execute
	/home/marxin/Programming/gcc/gcc/ipa-inline.cc:3210
Please submit a full bug report, with preprocessed source (by using -freport-bug).
Please include the complete backtrace with any bug report.
See <https://gcc.gnu.org/bugs/> for instructions.

Started with r14-377-gc92b8be9b52b7e and it's not related to aarch64, can reproduce on x86_64-linux-gnu as well.

Hm, on x86_64-linux-gnu, it started with r13-6616-g2246d576f922ba.

Slightly smaller:

```
void lspf2lpc();

void interpolate_lpc(int subframe_num) {
  float weight = 0.25 * subframe_num + 1;
  if (weight)
    lspf2lpc();
}

void qcelp_decode_frame() {
  int i;
  for (;; i++)
    interpolate_lpc(i);
}
```

(In reply to Martin Liška from comment #3)
> Hm, on x86_64-linux-gnu, it started with r13-6616-g2246d576f922ba.

$ cat prtest2.c
void lspf2lpc();

int interpolate_lpc_q_0;

void
interpolate_lpc(int subframe_num) {
  float weight;
  if (interpolate_lpc_q_0)
    weight = subframe_num;
  else
    weight = 1.0;
  if (weight != 1.0)
    lspf2lpc();
}

void
qcelp_decode_frame() {
  int i;
  for (;; i++)
    interpolate_lpc(i);
}

$ ./install/bin/gcc --version
gcc (GCC) 13.0.1 20230312 (experimental)

$ git log -1
commit 2246d576f922bae3629da0fe1dbfcc6ff06769ad (HEAD)
Author: Tamar Christina <tamar.christina@arm.com>
Date:   Sun Mar 12 18:39:33 2023 +0000

    middle-end: Revert can_special_div_by_const changes [PR108583]

    This reverts the changes for the CAN_SPECIAL_DIV_BY_CONST hook.

    gcc/ChangeLog:

            PR target/108583
            * doc/tm.texi (TARGET_VECTORIZE_CAN_SPECIAL_DIV_BY_CONST): Remove.
            * doc/tm.texi.in: Likewise.
            * explow.cc (round_push, align_dynamic_address): Revert previous patch.
            * expmed.cc (expand_divmod): Likewise.
            * expmed.h (expand_divmod): Likewise.
            * expr.cc (force_operand, expand_expr_divmod): Likewise.
            * optabs.cc (expand_doubleword_mod, expand_doubleword_divmod): Likewise.
            * target.def (can_special_div_by_const): Remove.
            * target.h: Remove tree-core.h include
            * targhooks.cc (default_can_special_div_by_const): Remove.
            * targhooks.h (default_can_special_div_by_const): Remove.
            * tree-vect-generic.cc (expand_vector_operation): Remove hook.
            * tree-vect-patterns.cc (vect_recog_divmod_pattern): Remove hook.
            * tree-vect-stmts.cc (vectorizable_operation): Remove hook.

$ ./install/bin/gcc -O2 -S -o - prtest2.c
        .file   "prtest2.c"
        .text
        .p2align 4
        .globl  interpolate_lpc
        .type   interpolate_lpc, @function
interpolate_lpc:
.LFB0:
        .cfi_startproc
        movl    interpolate_lpc_q_0(%rip), %eax
        testl   %eax, %eax
        je      .L1
        pxor    %xmm0, %xmm0
        cvtsi2ssl       %edi, %xmm0
        ucomiss .LC0(%rip), %xmm0
        jp      .L4
        jne     .L4
.L1:
        ret
        .p2align 4,,10
        .p2align 3
...

Also that commit doesn't build because I forgot to cp tm.texi to the source directory after the revert.

So I think the bisect probably didn't find it in that range.

https://godbolt.org/z/r44xGzarY indicates GCC 13.1 is fine.  So I don't think this one is mine.

my own bisect does indeed end up at r14-377-gc92b8be9b52b7e and cannot reproduce it on GCC 13.

It looks like IPA is trying to set a range for a float even though value_range does not support floats.  This worked before because legacy silently set a range of error_mark_node.

Created attachment 54980 [details]
untested

This may fix it.

Created attachment 54982 [details]
patch in testing

I think it's easiest to restore the legacy behavior for now.  A proper conversion of IPA should follow.

The master branch has been updated by Aldy Hernandez <aldyh@gcc.gnu.org>:

https://gcc.gnu.org/g:2b8a27634f5d28e3e7c4a08bf065f2daada7aed2

commit r14-458-g2b8a27634f5d28e3e7c4a08bf065f2daada7aed2
Author: Aldy Hernandez <aldyh@redhat.com>
Date:   Wed May 3 17:29:24 2023 +0200

    Allow varying ranges of unknown types in irange::verify_range [PR109711]
    
    The old legacy code allowed building ranges of unknown types so passes
    like IPA could build and propagate VARYING.  For now it's easiest to
    allow the old behavior, it's not like you can do anything with these
    ranges except build them and copy them.
    
    Eventually we should convert all users of set_varying() to use
    supported types.  I will address this in my upcoming IPA work.
    
            PR tree-optimization/109711
    
    gcc/ChangeLog:
    
            * value-range.cc (irange::verify_range): Allow types of
            error_mark_node.

fixed

*** Bug 109696 has been marked as a duplicate of this bug. ***