thr3ads.net - freebsd stable - rpc.lockd brokenness (2) [Mar 2006]

If this information is useful, please help other people find it:
Share via:

Jun Kuriyama

2006-Mar-06 14:43 UTC

rpc.lockd brokenness (2)

I'm not yet received enough information to track rpc.lockd problem.

As Kris posted before, here is a patch to backout my suspected
commit.  If someone can easily reproduce this problem, please try with
this patch on both of server/client side of rpc.lockd (I'm not sure
which of server/client side this affects).

http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/80389
http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/84953

Any reports about this patch (OK or still problem) are welcome!


Index: lock_proc.c
==================================================================RCS file:
/home/ncvs/src/usr.sbin/rpc.lockd/lock_proc.c,v
retrieving revision 1.18
retrieving revision 1.17
diff -u -r1.18 -r1.17
--- lock_proc.c	3 Feb 2005 22:21:19 -0000	1.18
+++ lock_proc.c	9 Oct 2004 15:36:13 -0000	1.17
@@ -62,8 +62,6 @@
 #define	CLIENT_CACHE_SIZE	64	/* No. of client sockets cached */
 #define	CLIENT_CACHE_LIFETIME	120	/* In seconds */
 
-#define	getrpcaddr(rqstp)	(struct sockaddr
*)(svc_getrpccaller((rqstp)->rq_xprt)->buf)
-
 static void	log_from_addr(const char *, struct svc_req *);
 static void	log_netobj(netobj *obj);
 static int	addrcmp(struct sockaddr *, struct sockaddr *);
@@ -196,7 +194,7 @@
 {
 	CLIENT *client;
 	struct timeval retry_time, time_now;
-	int error, i;
+	int i;
 	const char *netid;
 	struct netconfig *nconf;
 	char host[NI_MAXHOST];
@@ -243,11 +241,9 @@
 	 * Need a host string for clnt_tp_create. Use NI_NUMERICHOST
 	 * to avoid DNS lookups.
 	 */
-	error = getnameinfo(host_addr, host_addr->sa_len, host, sizeof host,
-			    NULL, 0, NI_NUMERICHOST);
-	if (error != 0) {
-		syslog(LOG_ERR, "unable to get name string for caller: %s",
-		       gai_strerror(error));
+	if (getnameinfo(host_addr, host_addr->sa_len, host, sizeof host,
+	    NULL, 0, NI_NUMERICHOST) != 0) {
+		syslog(LOG_ERR, "unable to get name string for caller");
 		return NULL;
 	}
 
@@ -566,7 +562,8 @@
 
 	res.cookie = arg->cookie;
 	res.stat.stat = getlock(&arg4, rqstp, LOCK_ASYNC | LOCK_MON);
-	transmit_result(NLM_LOCK_RES, &res, getrpcaddr(rqstp));
+	transmit_result(NLM_LOCK_RES, &res,
+	    (struct sockaddr *)svc_getcaller(rqstp->rq_xprt));
 
 	return (NULL);
 }
@@ -620,7 +617,8 @@
 	 * a lock to cancel, so this call always fails.
 	 */
 	res.stat.stat = unlock(&arg4, LOCK_CANCEL);
-	transmit_result(NLM_CANCEL_RES, &res, getrpcaddr(rqstp));
+	transmit_result(NLM_CANCEL_RES, &res,
+	    (struct sockaddr *)svc_getcaller(rqstp->rq_xprt));
 	return (NULL);
 }
 
@@ -667,7 +665,8 @@
 	res.stat.stat = unlock(&arg4, 0);
 	res.cookie = arg->cookie;
 
-	transmit_result(NLM_UNLOCK_RES, &res, getrpcaddr(rqstp));
+	transmit_result(NLM_UNLOCK_RES, &res,
+	    (struct sockaddr *)svc_getcaller(rqstp->rq_xprt));
 	return (NULL);
 }
 
@@ -724,7 +723,8 @@
 		nlm_granted, NULL, NLM_VERS) == 0 ?
 		nlm_granted : nlm_denied;
 
-	transmit_result(NLM_GRANTED_RES, &res, getrpcaddr(rqstp));
+	transmit_result(NLM_GRANTED_RES, &res,
+	    (struct sockaddr *)svc_getcaller(rqstp->rq_xprt));
 	return (NULL);
 }
 
@@ -1067,7 +1067,8 @@
 
 	res.cookie = arg->cookie;
 	res.stat.stat = getlock(arg, rqstp, LOCK_MON | LOCK_ASYNC | LOCK_V4);
-	transmit4_result(NLM4_LOCK_RES, &res, getrpcaddr(rqstp));
+	transmit4_result(NLM4_LOCK_RES, &res,
+	    (struct sockaddr *)svc_getcaller(rqstp->rq_xprt));
 
 	return (NULL);
 }
@@ -1115,7 +1116,8 @@
 	 * a lock to cancel, so this call always fails.
 	 */
 	res.stat.stat = unlock(&arg->alock, LOCK_CANCEL | LOCK_V4);
-	transmit4_result(NLM4_CANCEL_RES, &res, getrpcaddr(rqstp));
+	transmit4_result(NLM4_CANCEL_RES, &res,
+	    (struct sockaddr *)svc_getcaller(rqstp->rq_xprt));
 	return (NULL);
 }
 
@@ -1156,7 +1158,8 @@
 	res.stat.stat = unlock(&arg->alock, LOCK_V4);
 	res.cookie = arg->cookie;
 
-	transmit4_result(NLM4_UNLOCK_RES, &res, getrpcaddr(rqstp));
+	transmit4_result(NLM4_UNLOCK_RES, &res,
+	    (struct sockaddr *)svc_getcaller(rqstp->rq_xprt));
 	return (NULL);
 }
 
@@ -1212,7 +1215,8 @@
 	res.stat.stat = lock_answer(arg->alock.svid, &arg->cookie,
 		nlm4_granted, NULL, NLM_VERS4) == 0 ?
 		nlm4_granted : nlm4_denied;
-	transmit4_result(NLM4_GRANTED_RES, &res, getrpcaddr(rqstp));
+	transmit4_result(NLM4_GRANTED_RES, &res,
+	    (struct sockaddr *)svc_getrpccaller(rqstp->rq_xprt)->buf);
 	return (NULL);
 }
 

-- 
Jun Kuriyama <kuriyama@imgsrc.co.jp> // IMG SRC, Inc.
             <kuriyama@FreeBSD.org> // FreeBSD Project

Miguel Ramos

2006-Mar-07 09:40 UTC

head link

rpc.lockd brokenness (2)

The problem I'm having is not exactly what is described in pr bin/80389,
although it is very much related. So, I'm unable to verify if the patch
you gave corrects the problem, since the problem is not visible on my
systems.

I do have a lockd hang, however... Here is as much data as I can gather:

1- Only one client machine of all I have, the only one which is remote
booted, hangs on startup with rpc.lockd/rpc.statd enabled.
The hang occurs when daemons try to open their pid file using
pidfile_open functions after statd and lockd have been started.

2- The problem does not appear when I do "touch test; lockf test echo
1".

3- The only machine that shows this problem is the only machine
where /var is an NFS mount.

4- Applying the patch to revert to the Oct/2004 version, either to the
client or the server, didn't change a thing.

5- I'm led to believe that the hang on my machine is related to a
problem on the pidfile_open functions, not on rpc.lockd.

Can anybody else verify if the system hangs on pidfile_open when:
- /var is nfs mounted
- lockd and statd are running
- a daemon is started (perhaps /etc/rc.d/cron start)

Ter, 2006-03-07 ?s 08:10 +0900, Jun Kuriyama escreveu:> At Mon, 06 Mar 2006 22:56:46 +0000,
> Miguel Ramos wrote:
> > I'm getting rpc.lockd related hangs on a single machine which is
remote
> > booted. I'm goind to test you patch as soon as possible.
> 
> Thanks!
> 
Thanks.

-- 
Miguel Ramos <miguel@anjos.strangled.net>

Miguel Lopes Santos Ramos

2006-Mar-07 14:05 UTC

head link

rpc.lockd brokenness (2)

> From: Kris Kennaway <kris@obsecurity.org>
> Subject: Re: rpc.lockd brokenness (2)
>
> > Ok. There are two versions:
> >     http://mega.ist.utl.pt/~mlsr/nfs.dump
> >         is the output of tcpdump -vvv host targa and udp port nfs
> >     http://mega.ist.utl.pt/~mlsr/nfsx.dump
> >         is the output of tcpdump -X -vvv host targa and udp port nfs
>
> Hmm, looks like you need -s 0 in addition to -X -vvv.
There. http://mega.ist.utl.pt/~mlsr/nfsxs.dump

I did just cron, instead of /etc/rc.d/cron start. It has much less garbage now.

Miguel

Rong-En Fan

2006-Apr-08 05:29 UTC

head link

rpc.lockd brokenness (2)

On 3/6/06, Jun Kuriyama <kuriyama@imgsrc.co.jp>
wrote:>
> I'm not yet received enough information to track rpc.lockd problem.
>
> As Kris posted before, here is a patch to backout my suspected
> commit.  If someone can easily reproduce this problem, please try with
> this patch on both of server/client side of rpc.lockd (I'm not sure
> which of server/client side this affects).
>
> http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/80389
> http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/84953
>
> Any reports about this patch (OK or still problem) are welcome!
Hi,

Somehow I have problems with lockd after 3 boxes upgraded from
Feb's RELENG_6 to Apr 6's. One of them has problems with lockd.
For example, mutt and irssi will stuck in lockd (shown by
top). I tried to back out changes in revision 1.18 for lock_proc.c,
and do /etc/rc.d/nfslocking stop then a start. After backout it,
mutt and irssi work well. If I put 1.18 back, mutt and irssi will stuck
in lockd again.

Last month, I played with the test program/script in those two PRs,
found that revision 1.18 does not make any difference. I'm not 100%
sure the problem I encountered now is related to rev 1.18. But
it is a report that  backout 1.18 really helps.

For record, all my clients involved in this mail are running RELENG_6.
Server is RELENG_5 as of March 9. Only IPv4 here, no IPv6.

Regards,
Rong-En Fan

freebsd stable - Mar 2006 - rpc.lockd brokenness (2)

rpc.lockd brokenness (2)

rpc.lockd brokenness (2)

rpc.lockd brokenness (2)

rpc.lockd brokenness (2)