PERFORCE change 141676 for review

Vincenzo Iozzo snagg at FreeBSD.org
Fri May 16 16:46:21 UTC 2008


You're right Christian, thanks for the review

Il giorno 16/mag/08, alle ore 16:18, Christian S.J. Peron ha scritto:

> On Thu, May 15, 2008 at 09:45:32PM +0000, Vincenzo Iozzo wrote:
> [..]
>> ==== //depot/projects/soc2008/snagg-audit/sys/security/audit/ 
>> audit_pipe.c#9 (text) ====
>>
>> @@ -435,10 +435,6 @@
>> 	if (app != NULL) {
>> 		TAILQ_REMOVE(&ap->ap_preselect_list, app, app_list);
>> 		mtx_unlock(&audit_pipe_mtx);
>> -	}
>> -	
>> -	mtx_unlock(&audit_pipe_mtx);
>> -	if (app != NULL) {
>> 		for(i = 0; i < app->app_event_len; i++)
>
> Now we have eliminated the unlock which means we are leaking a  
> mutex. Why not
> try something like:
>
> [..]
> 	if (app != NULL)
> 		TAILQ_REMOVE(&ap->ap_preselect_list, app, app_list);
> 	mtx_unlock(&audit_pipe_mtx);
>
> This way we dont need to worry about conditionally dropping the mutex.



More information about the p4-projects mailing list