iv/linux
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Audit: stop deadlock from signals under load

Browse Source 
A deadlock is possible between kauditd and auditd under load if auditd
receives a signal.  When auditd receives a signal it sends a netlink
message to the kernel asking for information about the sender of the
signal.  In that same context the audit system will attempt to send a
netlink message back to the userspace auditd.  If kauditd has already
filled the socket buffer (see netlink_attachskb()) auditd will now put
itself to sleep waiting for room to send the message.  Since auditd is
responsible for draining that socket we have a deadlock.  The fix, since
the response from the kernel does not need to be synchronous is to send
the signal information back to auditd in a separate thread.  And thus
auditd can continue to drain the audit queue normally.

Signed-off-by: Eric Paris <eparis@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
This commit is contained in:
Eric Paris 2008-04-18 10:11:04 -04:00 committed by Al Viro
parent f3d357b092
commit f09ac9db2a
1 changed files with 35 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

38
kernel/audit.c
View File

@ -156,6 +156,11 @@ struct audit_buffer {
gfp_t gfp_mask;
};
struct audit_reply {
int pid;
struct sk_buff *skb;
};
static void audit_set_pid(struct audit_buffer *ab, pid_t pid)
{
if (ab) {
@ -528,6 +533,19 @@ nlmsg_failure: /* Used by NLMSG_PUT */
return NULL;
}
static int audit_send_reply_thread(void *arg)
{
struct audit_reply *reply = (struct audit_reply *)arg;
mutex_lock(&audit_cmd_mutex);
mutex_unlock(&audit_cmd_mutex);
/* Ignore failure. It'll only happen if the sender goes away,
because our timeout is set to infinite. */
netlink_unicast(audit_sock, reply->skb, reply->pid, 0);
kfree(reply);
return 0;
}
/**
* audit_send_reply - send an audit reply message via netlink
* @pid: process id to send reply to
@ -545,13 +563,25 @@ void audit_send_reply(int pid, int seq, int type, int done, int multi,
void *payload, int size)
{
struct sk_buff *skb;
struct task_struct *tsk;
struct audit_reply *reply = kmalloc(sizeof(struct audit_reply),
GFP_KERNEL);
if (!reply)
return;
skb = audit_make_reply(pid, seq, type, done, multi, payload, size);
if (!skb)
return;
/* Ignore failure. It'll only happen if the sender goes away,
because our timeout is set to infinite. */
netlink_unicast(audit_sock, skb, pid, 0);
return;
reply->pid = pid;
reply->skb = skb;
tsk = kthread_run(audit_send_reply_thread, reply, "audit_send_reply");
if (IS_ERR(tsk)) {
kfree(reply);
kfree_skb(skb);
}
}
/*