On Sat, Aug 21, 2010 at 8:01 PM, Tetsuo Handa
<penguin-kernel@i-love.sakura.ne.jp> wrote:
quoted text
> I was browsing unix_autobind() and wondered what happens if all names in
> Unix domain socket's abstract namespace were in use.
>
> Below part is unix_autobind() from linux-2.6.36-rc1/net/unix/af_unix.c
>
> 710 retry:
> 711         addr->len = sprintf(addr->name->sun_path+1, "%05x", ordernum) + 1 + sizeof(short);
> 712         addr->hash = unix_hash_fold(csum_partial(addr->name, addr->len, 0));
> 713
> 714         spin_lock(&unix_table_lock);
> 715         ordernum = (ordernum+1)&0xFFFFF;
> 716
> 717         if (__unix_find_socket_byname(net, addr->name, addr->len, sock->type,
> 718                                       addr->hash)) {
> 719                 spin_unlock(&unix_table_lock);
> 720                 /* Sanity yield. It is unusual case, but yet... */
> 721                 if (!(ordernum&0xFF))
> 722                         yield();
> 723                 goto retry;
> 724         }
>
> We can see that unix_autobind() allows 1048576 names.
>
> A machine with 256MB RAM:
>
>  # cat /proc/sys/fs/file-max
>  24109
>  # cat /proc/sys/fs/file-nr
>  608 0 24109
>
> A machine with 1736MB RAM:
>
>  # cat /proc/sys/fs/file-max
>  174347
>  # cat /proc/sys/fs/file-nr
>  96 0 174347
>
> /proc/sys/fs/file-max seems to be proportional to the amount of RAM.
> I don't have access to a machine with 10GB RAM (where /proc/sys/fs/file-max
> becomes larger than 1048576 by default). So, I manually set
>
>  # echo 1050000 > /proc/sys/fs/file-max
>
> and executed below program as non-root user.
>
> ---------- Test program start ----------
> #include <sys/socket.h>
> #include <sys/un.h>
> #include <unistd.h>
>
> int main(int argc, char *argv[])
> {
>        int i;
>        for (i = 0; i < 1030; i++) {
>                switch (fork()) {
>                case 0:
>                        sleep(5);
>                        close(0);
>                        close(1);
>                        close(2);
>                        for (i = 0; i < 1024; i++) {
>                                struct sockaddr_un addr;
>                                int fd = socket(PF_UNIX, SOCK_DGRAM, 0);
>                                addr.sun_family = AF_UNIX;
>                                bind(fd, (struct sockaddr *) &addr, sizeof(addr.sun_family));
>                        }
>                        while (1)
>                                sleep(1000);
>                case -1:
>                        write(1, "fork() failed\n", 14);
>                        return 1;
>                }
>        }
>        return 0;
> }
> ---------- Test program end ----------
>
> If there is not enough memory, OOM killer was invoked. OOM killer killed
> /usr/sbin/httpd process rather than above test program. (Oops. Non-root user
> was able to terminate other user's processes via OOM killer.)
>
> If there is enough memory (I can't test it), OOM killer will not be invoked.
> But if all names were occupied, I guess subsequent unix_autobind() by other
> users will loop forever because it loops until a name becomes available.
> (I had to type "killall a.out" before above program occupies all names, for
> the machine became very dull due to unix_autobind().)
>
> Maybe some safeguard is wanted.

You can add a counter of the names, and check it before allocating a new name.

-- 
Regards,
Changli Gao(xiaosuo@gmail.com)
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html