List:Commits« Previous MessageNext Message »
From:Ingo Strüwing Date:January 11 2008 3:09pm
Subject:Re: bk commit into 5.1 tree (istruewing:1.2641) BUG#33222
View as plain text  
Hi Ramil,

thanks for the review.

Ramil Kalimullin, 11.01.2008 13:12:
...
> Ingo Struewing wrote:
...
>>     ALTERing a table with long char columns warned about corruptions
>>   and left the altered table empty.
> 
> Could you please describe the problem a bit more accurately:

Ok.

...
>> +        to[0]= (uchar) ((new_length & 127)+128);
> as you're changing the line, please add spaces around '+'

Ok.

...
>> -      else if (*to++ != (char) new_length)
>> +      else if (*to++ != (uchar) new_length)
> do we really need the type cast above (both are unsigned)?

Sergey asked the same initially. But he agreed to my reply that some
compilers might warn about "data might be truncated as the source
variable can carry a bigger range of values" (uint assigned to uchar).
Do you also agree?

Regards
Ingo
-- 
Ingo Strüwing, Senior Software Developer
MySQL GmbH, Dachauer Str. 37, D-80335 München
Geschäftsführer: Kaj Arnö - HRB München 162140
Thread
bk commit into 5.1 tree (istruewing:1.2641) BUG#33222Ingo Struewing18 Dec
  • Re: bk commit into 5.1 tree (istruewing:1.2641) BUG#33222Ramil Kalimullin11 Jan
    • Re: bk commit into 5.1 tree (istruewing:1.2641) BUG#33222Ingo Strüwing11 Jan