Hi, all.
My program process multiple NFQUEUEs by creating a separate thread
for every NFQUEUE. An each thread do recv() and nfq_set_verdict2():
But i catch a strange behaviour when more than 1 thread(queue) are
used. In those cases one queue "freeze" while others works right.
For example, for 5 queues(here a queue with id 1 "freeze"):
# cat /proc/net/netfilter/nfnetlink_queue
0 -4216 4 2 65531 0 0 533513 1
1 19395 1024 2 65531 7319 0 1024 1
2 -4217 0 2 65531 0 0 62736 1
3 -4218 0 2 65531 0 0 150717 1
4 -4215 19 2 65531 0 0 1990176 1
# grep 'start thread' /var/log/messages
Aug 26 17:52:52 kvm trfl[19395]: start thread 2[19398] for nfqueue 2
Aug 26 17:52:52 kvm trfl[19395]: start thread 0[19396] for nfqueue 0
Aug 26 17:52:52 kvm trfl[19395]: start thread 1[19397] for nfqueue 1
Aug 26 17:52:52 kvm trfl[19395]: start thread 4[19400] for nfqueue 4
Aug 26 17:52:52 kvm trfl[19395]: start thread 3[19399] for nfqueue 3
# strace -p 19397
strace: Process 19397 attached
recvfrom(6, ^Cstrace: Process 19397 detached
<detached ...>
This behaviour is not constantly reproducible.
If i use SO_RCVTIMEO:
tv.tv_sec = 2;
tv.tv_usec = 0;
ret = setsockopt(nfq_fd(h), SOL_SOCKET, SO_RCVTIMEO, &tv, sizeof(tv));
....
fd = nfq_fd(h);
do {
while ((n = recv(fd, pkt_buf, 80000, 0)) > 0) {
nfq_handle_packet(h, pkt_buf, n);
}
fprintf(stderr, "%u: RECV ERR: %s", thread_idx, strerror(errno));
} while (errno == EWOULDBLOCK);
then i get every 2 seconds the next message:
0: RECV ERR: Resource temporarily unavailable
for case when thread with id = 0 freeze on recvfrom().
How can i resolve this problem?
Thanks!
--
Олег Неманов (Oleg Nemanov)
--
To unsubscribe from this list: send the line "unsubscribe netfilter-devel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
