On 2012/12/17 14:19, mar...@katamail.com wrote:
> Ettercap crashes on OpenBSD 5.2 i386.
>
> Steps to reproduce:
>
> # ettercap -C
>
> Select "Sniffing"
>
> Select "Unified sniffing..."
>
> Confirm the network interface
>
> After these steps ettercap quits without printing any error message.
>
> Tested on two machines and I get the same result. Ettercap worked fine on 5.1
> i386.
I don't know how to fix it but the way ettercap is doing mutex locking
in ec_thread_new() triggers strict mutex checking code in rthreads.
#0 0x00001b5ca5729d4a in kill () at <stdin>:2
#1 0x00001b5ca57904aa in abort () at /usr/src/lib/libc/stdlib/abort.c:70
#2 0x00001b5ca6118505 in _rthread_mutex_lock (mutexp=0x1b5a9ad99320,
trywait=0, abstime=0x0)
at /usr/src/lib/librthread/rthread_sync.c:127
#3 0x00001b5a9ab3a4ce in ec_thread_new (name=0x1b5a9ac8502e "top_half",
desc=0x1b5a9ac8501b "dispatching module", function=0x1b5a9ab1d73c
<top_half>, args=0x0) at ec_threads.c:210
#4 0x00001b5a9ab27455 in main (argc=2, argv=0x7f7ffffd5fe0) at ec_main.c:169
ettercap ec_threads.c:
185 * creates a new thread on the given function
186 */
187
188 pthread_t ec_thread_new(char *name, char *desc, void *(*function)(void *),
v oid *args)
189 {
190 pthread_t id;
191
192 DEBUG_MSG("ec_thread_new -- %s", name);
193
194 /*
195 * lock the mutex to syncronize with the new thread.
196 * the newly created thread will perform INIT_UNLOCK
197 * so at the end of this function we are sure that the
198 * thread had be initialized
199 */
200 INIT_LOCK;
201
202 if (pthread_create(&id, NULL, function, args) != 0)
203 ERROR_MSG("not enough resources to create a new thread in this
process ");
204
205 ec_thread_register(id, name, desc);
206
207 DEBUG_MSG("ec_thread_new -- %lu created ", PTHREAD_ID(id));
208
209 /* the new thread will unlock this */
210 INIT_LOCK;
211 INIT_UNLOCK;
212
213 return id;
214 }
and this is the check which is triggered.
124 /* self-deadlock is disallowed by strict */
125 if (mutex->type == PTHREAD_MUTEX_STRICT_NP &&
126 abstime == NULL)
127 abort();
