Commit Graph

6 Commits

Author SHA1 Message Date
David Teigland
36b71a8bfb dlm: fix deadlock between dlm_send and dlm_controld
A deadlock sometimes occurs between dlm_controld closing
a lowcomms connection through configfs and dlm_send looking
up the address for a new connection in configfs.

dlm_controld does a configfs rmdir which calls
dlm_lowcomms_close which waits for dlm_send to
cancel work on the workqueues.

The dlm_send workqueue thread has called
tcp_connect_to_sock which calls dlm_nodeid_to_addr
which does a configfs lookup and blocks on a lock
held by dlm_controld in the rmdir path.

The solution here is to save the node addresses within
the lowcomms code so that the lowcomms workqueue does
not need to step through configfs to get a node address.

dlm_controld:
wait_for_completion+0x1d/0x20
__cancel_work_timer+0x1b3/0x1e0
cancel_work_sync+0x10/0x20
dlm_lowcomms_close+0x4c/0xb0 [dlm]
drop_comm+0x22/0x60 [dlm]
client_drop_item+0x26/0x50 [configfs]
configfs_rmdir+0x180/0x230 [configfs]
vfs_rmdir+0xbd/0xf0
do_rmdir+0x103/0x120
sys_rmdir+0x16/0x20

dlm_send:
mutex_lock+0x2b/0x50
get_comm+0x34/0x140 [dlm]
dlm_nodeid_to_addr+0x18/0xd0 [dlm]
tcp_connect_to_sock+0xf4/0x2d0 [dlm]
process_send_sockets+0x1d2/0x260 [dlm]
worker_thread+0x170/0x2a0

Signed-off-by: David Teigland <teigland@redhat.com>
2012-08-08 11:33:35 -05:00
Christine Caulfield
391fbdc5d5 dlm: connect to nodes earlier
Make network connections to other nodes earlier, in the context of
dlm_recoverd.  This avoids connecting to nodes from dlm_send where we
try to avoid allocations which could possibly deadlock if memory reclaim
goes into the cluster fs which may try to do a dlm operation.

Signed-off-by: Christine Caulfield <ccaulfie@redhat.com>
Signed-off-by: David Teigland <teigland@redhat.com>
2009-05-15 09:34:12 -05:00
Patrick Caulfield
ac33d07105 [DLM] Clean up lowcomms
This fixes up most of the things pointed out by akpm and Pavel Machek
with comments below indicating why some things have been left:

Andrew Morton wrote:
>
>> +static struct nodeinfo *nodeid2nodeinfo(int nodeid, gfp_t alloc)
>> +{
>> +	struct nodeinfo *ni;
>> +	int r;
>> +	int n;
>> +
>> +	down_read(&nodeinfo_lock);
>
> Given that this function can sleep, I wonder if `alloc' is useful.
>
> I see lots of callers passing in a literal "0" for `alloc'.  That's in fact
> a secret (GFP_ATOMIC & ~__GFP_HIGH).  I doubt if that's what you really
> meant.  Particularly as the code could at least have used __GFP_WAIT (aka
> GFP_NOIO) which is much, much more reliable than "0".  In fact "0" is the
> least reliable mode possible.
>
> IOW, this is all bollixed up.

When 0 is passed into nodeid2nodeinfo the function does not try to allocate a
new structure at all. it's an indication that the caller only wants the nodeinfo
struct for that nodeid if there actually is one in existance.
I've tidied the function itself so it's more obvious, (and tidier!)

>> +/* Data received from remote end */
>> +static int receive_from_sock(void)
>> +{
>> +	int ret = 0;
>> +	struct msghdr msg;
>> +	struct kvec iov[2];
>> +	unsigned len;
>> +	int r;
>> +	struct sctp_sndrcvinfo *sinfo;
>> +	struct cmsghdr *cmsg;
>> +	struct nodeinfo *ni;
>> +
>> +	/* These two are marginally too big for stack allocation, but this
>> +	 * function is (currently) only called by dlm_recvd so static should be
>> +	 * OK.
>> +	 */
>> +	static struct sockaddr_storage msgname;
>> +	static char incmsg[CMSG_SPACE(sizeof(struct sctp_sndrcvinfo))];
>
> whoa.  This is globally singly-threaded code??

Yes. it is only ever run in the context of dlm_recvd.
>>
>> +static void initiate_association(int nodeid)
>> +{
>> +	struct sockaddr_storage rem_addr;
>> +	static char outcmsg[CMSG_SPACE(sizeof(struct sctp_sndrcvinfo))];
>
> Another static buffer to worry about.  Globally singly-threaded code?

Yes. Only ever called by dlm_sendd.

>> +
>> +/* Send a message */
>> +static int send_to_sock(struct nodeinfo *ni)
>> +{
>> +	int ret = 0;
>> +	struct writequeue_entry *e;
>> +	int len, offset;
>> +	struct msghdr outmsg;
>> +	static char outcmsg[CMSG_SPACE(sizeof(struct sctp_sndrcvinfo))];
>
> Singly-threaded?

Yep.

>>
>> +static void dealloc_nodeinfo(void)
>> +{
>> +	int i;
>> +
>> +	for (i=1; i<=max_nodeid; i++) {
>> +		struct nodeinfo *ni = nodeid2nodeinfo(i, 0);
>> +		if (ni) {
>> +			idr_remove(&nodeinfo_idr, i);
>
> Didn't that need locking?

Not. it's only ever called at DLM shutdown after all the other threads
have been stopped.

>>
>> +static int write_list_empty(void)
>> +{
>> +	int status;
>> +
>> +	spin_lock_bh(&write_nodes_lock);
>> +	status = list_empty(&write_nodes);
>> +	spin_unlock_bh(&write_nodes_lock);
>> +
>> +	return status;
>> +}
>
> This function's return value is meaningless.  As soon as the lock gets
> dropped, the return value can get out of sync with reality.
>
> Looking at the caller, this _might_ happen to be OK, but it's a nasty and
> dangerous thing.  Really the locking should be moved into the caller.

It's just an optimisation to allow the caller to schedule if there is no work
to do. if something arrives immediately afterwards then it will get picked up
when the process re-awakes (and it will be woken by that arrival).

The 'accepting' atomic has gone completely. as Andrew pointed out it didn't
really achieve much anyway. I suspect it was a plaster over some other
startup or shutdown bug to be honest.


Signed-off-by: Patrick Caulfield <pcaulfie@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
Cc: Andrew Morton <akpm@osdl.org>
Cc: Pavel Machek <pavel@ucw.cz>
2006-12-07 09:25:13 -05:00
Al Viro
38d6fd26ea [PATCH] dlm gfp_t annotations
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-10-09 14:19:08 -07:00
David Teigland
1c032c0311 [DLM] PATCH 2/3 dlm: lowcomms close
When a node is removed from a lockspace configuration, close our
connection to it, clearing any remaining messages for it.

Signed-off-by: David Teigland <teigland@redhat.com>
Signed-off-by: Patrick Caulfield <pcaulfie@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
2006-04-28 10:50:41 -04:00
David Teigland
e7fd41792f [DLM] The core of the DLM for GFS2/CLVM
This is the core of the distributed lock manager which is required
to use GFS2 as a cluster filesystem. It is also used by CLVM and
can be used as a standalone lock manager independantly of either
of these two projects.

It implements VAX-style locking modes.

Signed-off-by: David Teigland <teigland@redhat.com>
Signed-off-by: Steve Whitehouse <swhiteho@redhat.com>
2006-01-18 09:30:29 +00:00