All of lore.kernel.org
 help / color / mirror / Atom feed
From: Benny Halevy <bhalevy@panasas.com>
To: andros@netapp.com
Cc: pnfs@linux-nfs.org, linux-nfs@vger.kernel.org
Subject: Re: [pnfs] [PATCH 2/2] nfs41: fix race on session reset
Date: Thu, 22 Oct 2009 15:20:18 +0200	[thread overview]
Message-ID: <4AE05C12.7070403@panasas.com> (raw)
In-Reply-To: <1256149492-25481-3-git-send-email-andros@netapp.com>

On Oct. 21, 2009, 20:24 +0200, andros@netapp.com wrote:
> From: Andy Adamson <andros@netapp.com>
> 
> Do not clear the NFS4CLNT_SESSION_SETUP bit until after the session has been
> reset (a possible EXCHANGE_ID, a DESTROY_SESSION, and a CREATE_SESSION) to
> prevent a race with nfs41_setup_sequence assigning a slot on a
> partially reset session.
> 
> Signed-off-by: Andy Adamson <andros@netapp.com>
> ---
>  fs/nfs/nfs4proc.c  |    3 +++
>  fs/nfs/nfs4state.c |   15 +++++++++------
>  2 files changed, 12 insertions(+), 6 deletions(-)
> 
> diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
> index eb245a1..80764e2 100644
> --- a/fs/nfs/nfs4proc.c
> +++ b/fs/nfs/nfs4proc.c
> @@ -4788,6 +4788,9 @@ int nfs4_proc_create_session(struct nfs_client *clp, int reset)
>  	if (status)
>  		goto out;
>  
> +	/* Signal nfs41_setup_sequence that the session is ready for use */
> +	clear_bit(NFS4CLNT_SESSION_SETUP, &clp->cl_state);
> +
>  	ptr = (unsigned *)&session->sess_id.data[0];
>  	dprintk("%s client>seqid %d sessionid %u:%u:%u:%u\n", __func__,
>  		clp->cl_seqid, ptr[0], ptr[1], ptr[2], ptr[3]);
> diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c
> index 1394dfb..09ca30b 100644
> --- a/fs/nfs/nfs4state.c
> +++ b/fs/nfs/nfs4state.c
> @@ -1252,17 +1252,20 @@ static void nfs4_state_manager(struct nfs_client *clp)
>  				continue;
>  		}
>  		/* Initialize or reset the session */
> -		if (test_and_clear_bit(NFS4CLNT_SESSION_SETUP, &clp->cl_state)
> -		   && nfs4_has_session(clp)) {
> +		if (test_bit(NFS4CLNT_SESSION_SETUP, &clp->cl_state)
> +		    && nfs4_has_session(clp)) {
>  			if (clp->cl_cons_state == NFS_CS_SESSION_INITING)
>  				status = nfs4_initialize_session(clp);
>  			else
>  				status = nfs4_reset_session(clp);
> -			if (status) {
> -				if (status == -NFS4ERR_STALE_CLIENTID)
> -					continue;
> +			if (status == -NFS4ERR_STALE_CLIENTID)
> +				continue;
> +			/* For error case. On success the bit is cleared in
> +			 * nfs4_proc_create_session */

Why the separation?
Why not handle both success and error cases here?

Benny

> +			clear_bit(NFS4CLNT_SESSION_SETUP, &clp->cl_state);
> +
> +			if (status)
>  				goto out_error;
> -			}
>  		}
>  		/* First recover reboot state... */
>  		if (test_and_clear_bit(NFS4CLNT_RECLAIM_REBOOT, &clp->cl_state)) {

  reply	other threads:[~2009-10-22 13:20 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2009-10-21 18:24 [PATCH 0/2] nfs41 client session bug fixes andros
2009-10-21 18:24 ` [PATCH 1/2] nfs41: fix multiple free slot calls andros
2009-10-21 18:24   ` [PATCH 2/2] nfs41: fix race on session reset andros
2009-10-22 13:20     ` Benny Halevy [this message]
2009-10-22 13:32       ` [pnfs] " William A. (Andy) Adamson
     [not found]         ` <89c397150910220632o5708b8efhb7223541315b4e33-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
2009-10-22 13:47           ` Benny Halevy
2009-10-22 14:14             ` William A. (Andy) Adamson
     [not found]               ` <89c397150910220714jae5e335qa8313a994e440660-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
2009-11-03  6:22                 ` Benny Halevy

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=4AE05C12.7070403@panasas.com \
    --to=bhalevy@panasas.com \
    --cc=andros@netapp.com \
    --cc=linux-nfs@vger.kernel.org \
    --cc=pnfs@linux-nfs.org \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.