cluster-devel.redhat.com archive mirror
 help / color / mirror / Atom feed
From: dann frazier <dann.frazier@canonical.com>
To: cluster-devel.redhat.com
Subject: [Cluster-devel] GPF in dlm_lowcomms_stop
Date: Fri, 30 Mar 2012 11:17:56 -0600	[thread overview]
Message-ID: <20120330171756.GC4584@dannf.org> (raw)
In-Reply-To: <20120330164240.GF20636@redhat.com>

On Fri, Mar 30, 2012 at 12:42:40PM -0400, David Teigland wrote:
> On Fri, Mar 30, 2012 at 11:42:56AM -0400, David Teigland wrote:
> > Hi Dan, I'm not very familiar with this code either, but I've talked with
> > Chrissie and she suggested we try something like this:

Yeah, that's the mechanism I was thinking of as well.

> A second version that addresses a potentially similar problem in
> start.

Oh, good catch! I hadn't considered that path.

Reviewed-by: dann frazier <dann.frazier@canonical.com>

> 
> diff --git a/fs/dlm/lowcomms.c b/fs/dlm/lowcomms.c
> index 133ef6d..5c1b0e3 100644
> --- a/fs/dlm/lowcomms.c
> +++ b/fs/dlm/lowcomms.c
> @@ -142,6 +142,7 @@ struct writequeue_entry {
>  
>  static struct sockaddr_storage *dlm_local_addr[DLM_MAX_ADDR_COUNT];
>  static int dlm_local_count;
> +static int dlm_allow_conn;
>  
>  /* Work queues */
>  static struct workqueue_struct *recv_workqueue;
> @@ -710,6 +711,13 @@ static int tcp_accept_from_sock(struct connection *con)
>  	struct connection *newcon;
>  	struct connection *addcon;
>  
> +	mutex_lock(&connections_lock);
> +	if (!dlm_allow_conn) {
> +		mutex_unlock(&connections_lock);
> +		return -1;
> +	}
> +	mutex_unlock(&connections_lock);
> +
>  	memset(&peeraddr, 0, sizeof(peeraddr));
>  	result = sock_create_kern(dlm_local_addr[0]->ss_family, SOCK_STREAM,
>  				  IPPROTO_TCP, &newsock);
> @@ -1503,6 +1511,7 @@ void dlm_lowcomms_stop(void)
>  	   socket activity.
>  	*/
>  	mutex_lock(&connections_lock);
> +	dlm_allow_conn = 0;
>  	foreach_conn(stop_conn);
>  	mutex_unlock(&connections_lock);
>  
> @@ -1530,7 +1539,7 @@ int dlm_lowcomms_start(void)
>  	if (!dlm_local_count) {
>  		error = -ENOTCONN;
>  		log_print("no local IP address has been set");
> -		goto out;
> +		goto fail;
>  	}
>  
>  	error = -ENOMEM;
> @@ -1538,7 +1547,13 @@ int dlm_lowcomms_start(void)
>  				      __alignof__(struct connection), 0,
>  				      NULL);
>  	if (!con_cache)
> -		goto out;
> +		goto fail;
> +
> +	error = work_start();
> +	if (error)
> +		goto fail_destroy;
> +
> +	dlm_allow_conn = 1;
>  
>  	/* Start listening */
>  	if (dlm_config.ci_protocol == 0)
> @@ -1548,20 +1563,17 @@ int dlm_lowcomms_start(void)
>  	if (error)
>  		goto fail_unlisten;
>  
> -	error = work_start();
> -	if (error)
> -		goto fail_unlisten;
> -
>  	return 0;
>  
>  fail_unlisten:
> +	dlm_allow_conn = 0;
>  	con = nodeid2con(0,0);
>  	if (con) {
>  		close_connection(con, false);
>  		kmem_cache_free(con_cache, con);
>  	}
> +fail_destroy:
>  	kmem_cache_destroy(con_cache);
> -
> -out:
> +fail:
>  	return error;
>  }



  reply	other threads:[~2012-03-30 17:17 UTC|newest]

Thread overview: 6+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-03-22  1:59 [Cluster-devel] GPF in dlm_lowcomms_stop dann frazier
2012-03-30 15:42 ` David Teigland
2012-03-30 16:42   ` David Teigland
2012-03-30 17:17     ` dann frazier [this message]
2012-05-04 17:33       ` dann frazier
2012-05-04 17:51         ` David Teigland

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20120330171756.GC4584@dannf.org \
    --to=dann.frazier@canonical.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).