Subject: Re: [PATCH] NFS: add more context to state manager error mesgs
Mime-Version: 1.0 (Apple Message framework v1278)
Content-Type: text/plain; charset=us-ascii
From: Chuck Lever <chuck.lever@oracle.com>
In-Reply-To: <1338321479-3830-1-git-send-email-dros@netapp.com>
Date: Tue, 29 May 2012 15:59:44 -0400
Cc: Trond.Myklebust@netapp.com, linux-nfs@vger.kernel.org
Message-Id: <79585983-5563-43B1-95A5-978F0F89E8C5@oracle.com>
References: <1338321479-3830-1-git-send-email-dros@netapp.com>
To: Weston Andros Adamson <dros@netapp.com>
Sender: linux-nfs-owner@vger.kernel.org


On May 29, 2012, at 3:57 PM, Weston Andros Adamson wrote:

> 
> Signed-off-by: Weston Andros Adamson <dros@netapp.com>
> ---
> This has been useful for me in the past - time to share it!

Agree, I also have a patch that does something klutzy but very similar.

> Very simple, but should help narrow down the scope problems in the state manager
> thread on systems without rpcdebug enabled.
> 
> fs/nfs/nfs4state.c |   16 ++++++++++++++--
> 1 files changed, 14 insertions(+), 2 deletions(-)
> 
> diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c
> index c679b9e..b9963d8 100644
> --- a/fs/nfs/nfs4state.c
> +++ b/fs/nfs/nfs4state.c
> @@ -1855,10 +1855,12 @@ static int nfs4_bind_conn_to_session(struct nfs_client *clp)
> static void nfs4_state_manager(struct nfs_client *clp)
> {
> 	int status = 0;
> +	const char *section = "", *section_sep = "";
> 
> 	/* Ensure exclusive access to NFSv4 state */
> 	do {
> 		if (test_bit(NFS4CLNT_PURGE_STATE, &clp->cl_state)) {
> +			section = "purge state";
> 			status = nfs4_reclaim_lease(clp);
> 			if (status < 0)
> 				goto out_error;
> @@ -1867,6 +1869,7 @@ static void nfs4_state_manager(struct nfs_client *clp)
> 		}
> 
> 		if (test_and_clear_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state)) {
> +			section = "lease expired";
> 			/* We're going to have to re-establish a clientid */
> 			status = nfs4_reclaim_lease(clp);
> 			if (status < 0)
> @@ -1886,6 +1889,7 @@ static void nfs4_state_manager(struct nfs_client *clp)
> 		}
> 
> 		if (test_and_clear_bit(NFS4CLNT_CHECK_LEASE, &clp->cl_state)) {
> +			section = "check lease";
> 			status = nfs4_check_lease(clp);
> 			if (status < 0)
> 				goto out_error;
> @@ -1896,6 +1900,7 @@ static void nfs4_state_manager(struct nfs_client *clp)
> 		/* Initialize or reset the session */
> 		if (test_and_clear_bit(NFS4CLNT_SESSION_RESET, &clp->cl_state)
> 		   && nfs4_has_session(clp)) {
> +			section = "reset session";
> 			status = nfs4_reset_session(clp);
> 			if (test_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state))
> 				continue;
> @@ -1906,6 +1911,7 @@ static void nfs4_state_manager(struct nfs_client *clp)
> 		/* Send BIND_CONN_TO_SESSION */
> 		if (test_and_clear_bit(NFS4CLNT_BIND_CONN_TO_SESSION,
> 				&clp->cl_state) && nfs4_has_session(clp)) {
> +			section = "bind conn to session";
> 			status = nfs4_bind_conn_to_session(clp);
> 			if (status < 0)
> 				goto out_error;
> @@ -1914,6 +1920,7 @@ static void nfs4_state_manager(struct nfs_client *clp)
> 
> 		/* First recover reboot state... */
> 		if (test_bit(NFS4CLNT_RECLAIM_REBOOT, &clp->cl_state)) {
> +			section = "reclaim reboot";
> 			status = nfs4_do_reclaim(clp,
> 				clp->cl_mvops->reboot_recovery_ops);
> 			if (test_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state) ||
> @@ -1928,6 +1935,7 @@ static void nfs4_state_manager(struct nfs_client *clp)
> 
> 		/* Now recover expired state... */
> 		if (test_and_clear_bit(NFS4CLNT_RECLAIM_NOGRACE, &clp->cl_state)) {
> +			section = "reclaim nograce";
> 			status = nfs4_do_reclaim(clp,
> 				clp->cl_mvops->nograce_recovery_ops);
> 			if (test_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state) ||
> @@ -1946,6 +1954,7 @@ static void nfs4_state_manager(struct nfs_client *clp)
> 		/* Recall session slots */
> 		if (test_and_clear_bit(NFS4CLNT_RECALL_SLOT, &clp->cl_state)
> 		   && nfs4_has_session(clp)) {
> +			section = "recall slot";
> 			status = nfs4_recall_slot(clp);
> 			if (status < 0)
> 				goto out_error;
> @@ -1962,8 +1971,11 @@ static void nfs4_state_manager(struct nfs_client *clp)
> 	} while (atomic_read(&clp->cl_count) > 1);
> 	return;
> out_error:
> -	pr_warn_ratelimited("NFS: state manager failed on NFSv4 server %s"
> -			" with error %d\n", clp->cl_hostname, -status);
> +	if (strlen(section))
> +		section_sep = ": ";
> +	pr_warn_ratelimited("NFS: state manager%s%s failed on NFSv4 server %s"
> +			" with error %d\n", section_sep, section,
> +			clp->cl_hostname, -status);
> 	nfs4_end_drain_session(clp);
> 	nfs4_clear_state_manager_bit(clp);
> }
> -- 
> 1.7.4.4
> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

-- 
Chuck Lever
chuck[dot]lever[at]oracle[dot]com