Re: bzr commit into mysql-5.1-bugteam branch (alfranio.correia:2862) Bug#43949

[Andrei Elkin]

Alfranio, hello.

>>> @@ -105,7 +105,7 @@ int init_relay_log_info(Relay_log_info* 
>>>    rli->tables_to_lock_count= 0;
>>>  
>>>    fn_format(rli->slave_patternload_file, PREFIX_SQL_LOAD,
> slave_load_tmpdir, "",
>>> -            MY_PACK_FILENAME | MY_UNPACK_FILENAME |
>>> +            MY_SAFE_PATH | MY_UNPACK_FILENAME |
>>>              MY_RETURN_REAL_PATH);
>>>     
>>
>> Good. the "annihilating" line got corrected.
>>
>> Now fn_format() is capable to return NULL due to MY_SAFE_PATH.
>> I think we are better to check the return value.
>> Maybe it's unnecessary to have a dedicate test for the error
>> propagation, but I strongly recommend to simulate failures in all use
>> cases calling init_relay_log_info().
>>   
> Sure. I will check if null is returned.
> If you want I can create a test that will force returning null but
> I think it is not necessary.

I still it's pretty redundant as the case is obvious.
Manual testing is enough.


cheers,

Andrei