Richard Henderson <[email protected]> writes:

> On 12/3/18 4:43 AM, Alex Bennée wrote:
>>>      case R_AARCH64_CONDBR19:
>>> -        reloc_pc19(code_ptr, (tcg_insn_unit *)value);
>>> -        break;
>>> +        return reloc_pc19(code_ptr, (tcg_insn_unit *)value);
>>>      default:
>>>          tcg_abort();
>>>      }
>>> -    return true;
>>
>> nit: the default leg could return false for the same effect
>
> Would it be clearer changed to g_assert_not_reached()?
> Because I'm not intending "unknown relocation" to have
> the same effect as "out of range relocation".

g_assert_not_reached would probably be better then.

Is there any particular reason tcg has tcg_abort(), is it just for the
slightly different report string?

>
>
> r~


--
Alex Bennée

Reply via email to