more uninitialized mutexes in rc.39

Carter Bullard carter at qosient.com
Tue Mar 6 23:43:28 EST 2007


Hey Christoph,
Fixed in rc.41, which I should have up tomorrow.
Carter

Christoph Badura wrote:

>Hey Carter,
>
>here are the remaining places rc.40 where hash tables are allocated without
>ArgusNewHashTable().
>
>In the clients distribution:
>
>clients/racount.c:98:struct ArgusHashTable ArgusSrcAddrTable, ArgusDstAddrTable;
>clients/racount.c:135:      if ((ArgusSrcAddrTable.array = (struct ArgusHashTabl
>clients/racount.c:139:      if ((ArgusDstAddrTable.array = (struct ArgusHashTabl
>
>clients/rapath.c:72:struct ArgusHashTable ArgusHashTable;
>clients/rapath.c:243:      if ((ArgusHashTable.array = (struct ArgusHashTableHdr **) ArgusCalloc (RABINS_HASHTABLESIZE,
>clients/rapath.c:244:                                    sizeof (struct ArgusHashTableHdr *))) == NULL) {
>
>In the server distribution:
>
>argus/ArgusModeler.c:78:   if ((retn->ArgusHashTable = (struct ArgusHashTableStruct *) ArgusCalloc (1, sizeof (struct ArgusHashTableStruct))) == NULL)
>argus/ArgusModeler.c:87:   pthread_mutex_init(&retn->ArgusHashTable->lock, NULL);
>argus/ArgusModeler.c:172:         model->ArgusHashTable = NULL;
>
>But, of course, the lock is not normally taken.
>
>Thanks for looking into this!
>--chris
>
>  
>




More information about the argus mailing list