[PATCH v2 8/8] resolv: Add test case tst-ns_sprintrr (bug 34033, bug 34069)

[Adhemerval Zanella Netto]

On 19/06/26 13:41, Florian Weimer wrote:
> * Adhemerval Zanella Netto:
> 
>> On 19/06/26 11:50, Florian Weimer wrote:
>>
>>>
>>>>> +  bool success = false;
>>>>> +  for (size_t result_size = 1; result_size <= max_result_size; ++result_size)
>>>>> +    {
>>>>> +      char *result_start = ntf_out.buffer + ntf_out.length - result_size;
>>>>> +      memset (result_start, 'X', result_size);
>>>>> +
>>>>> +      /* ns_sprintrr was deprecated in 2.34.  */
>>>>> +      DIAG_PUSH_NEEDS_COMMENT;
>>>>> +      DIAG_IGNORE_NEEDS_COMMENT (4.9, "-Wdeprecated-declarations");
>>>>
>>>> Maybe use 12.0, since it is the current minimal support version
>>>> anyway.
>>>
>>> Not sure what the rules for permanent exemptions are.  Should I use
>>> 16.1?
> 
>> My understanding is the first supported version this warning happens, so 
>> we can easily remove them if/when we raise de minimum supported version.
> 
> But this doesn't really apply here.  It's not a compiler deficiency we
> working around, but an expectation deprecation message triggered by our
> headers.  Leaving it at 4.9 may actually be helpful because it's not
> something we want to delete when we increase the baseline from GCC 12 to
> GCC 13.

Ok, that's make sense then.