[Libreoffice] [Patch] [Pushed] Question about sal/qa/osl/process/osl_Thread.cxx
Julien Nabet
serval2412 at yahoo.fr
Sat Jun 18 05:09:04 PDT 2011
Le 18/06/2011 13:12, Norbert Thiebaud a écrit :
> On Sat, Jun 18, 2011 at 5:39 AM, Julien Nabet<serval2412 at yahoo.fr> wrote:
>> Hello,
>>
>> Last "git version" of cppcheck says this :
>> [sal/qa/osl/process/osl_Thread.cxx:1099]: (warning) Redundant assignment of
>> "sPrioStr" in switch
>> Here is the switch :
>> switch (_aPriority)
>> {
>> case osl_Thread_PriorityHighest:
>> sPrioStr = "Highest";
>> break;
>>
>> case osl_Thread_PriorityAboveNormal:
>> sPrioStr = "AboveNormal";
>>
>> case osl_Thread_PriorityNormal:
>> sPrioStr = "Normal";
>>
>> case osl_Thread_PriorityBelowNormal:
>> sPrioStr = "BelowNormal";
>> break;
>>
>> case osl_Thread_PriorityLowest:
>> sPrioStr = "Lowest";
>> break;
>> default:
>> sPrioStr = "unknown";
>> }
>>
>> Do the break instructions lack for "AboveNormal" and "Normal" cases ?
>> If no, it could be better to have this :
>> // we want to have the same treatment for these 3 cases
>> case osl_Thread_PriorityAboveNormal:
>> case osl_Thread_PriorityNormal:
>> case osl_Thread_PriorityBelowNormal:
>> sPrioStr = "BelowNormal";
>> break;
>>
>> Somebody to answer this ?
> * It looks like this has been like that since day-1 (i.e for as long
> as we have history for it.
> * That function seems to be only used for display purpose at only place only.
> * It looks like there should be break after each case and no
> fall-through. I don't see the reason why one would want to obfuscate
> the true value of priority in a trace message.
>
Ok, so I added the 2 break instructions, commited and pushed on master.
Thank you !
> Norbert
>
More information about the LibreOffice
mailing list