[PATCH] nfsd: update callback done processing

[PATCH] nfsd: update callback done processing

[Scott Mayhew]

Instead of having the convention where individual nfsd4_callback_ops->done
operations return -1 to indicate the callback path is down, move the check
to nfsd4_cb_done.  Only mark the callback path down on transport-level
errors, not NFS-level errors.

Additionally, handle NFS4ERR_DELAY in nfsd4_cb_recall_done.  The client
returns NFS4ERR_DELAY if it is already in the process of returning the
delegation.

Signed-off-by: Scott Mayhew <smayhew@redhat.com>
---
 fs/nfsd/nfs4callback.c | 9 +++++----
 fs/nfsd/nfs4layouts.c  | 2 +-
 fs/nfsd/nfs4state.c    | 5 ++++-
 3 files changed, 10 insertions(+), 6 deletions(-)

diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c
index 7caa3801ce72..b9cbd7189d74 100644
--- a/fs/nfsd/nfs4callback.c
+++ b/fs/nfsd/nfs4callback.c
@@ -1122,10 +1122,11 @@ static void nfsd4_cb_done(struct rpc_task *task, void *calldata)
 		rpc_restart_call_prepare(task);
 		return;
 	case 1:
-		break;
-	case -1:
-		/* Network partition? */
-		nfsd4_mark_cb_down(clp, task->tk_status);
+		switch (task->tk_status) {
+		case -EIO:
+		case -ETIMEDOUT:
+			nfsd4_mark_cb_down(clp, task->tk_status);
+		}
 		break;
 	default:
 		BUG();
diff --git a/fs/nfsd/nfs4layouts.c b/fs/nfsd/nfs4layouts.c
index 44517fb5c0de..a79e24b79095 100644
--- a/fs/nfsd/nfs4layouts.c
+++ b/fs/nfsd/nfs4layouts.c
@@ -693,7 +693,7 @@ nfsd4_cb_layout_done(struct nfsd4_callback *cb, struct rpc_task *task)
 			ops->fence_client(ls);
 		else
 			nfsd4_cb_layout_fail(ls);
-		return -1;
+		return 1;
 	case -NFS4ERR_NOMATCHING_LAYOUT:
 		trace_nfsd_layout_recall_done(&ls->ls_stid.sc_stateid);
 		task->tk_status = 0;
diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c
index f056b1d3fecd..4805b83ac5e5 100644
--- a/fs/nfsd/nfs4state.c
+++ b/fs/nfsd/nfs4state.c
@@ -3957,6 +3957,9 @@ static int nfsd4_cb_recall_done(struct nfsd4_callback *cb,
 	switch (task->tk_status) {
 	case 0:
 		return 1;
+	case -NFS4ERR_DELAY:
+		rpc_delay(task, 2 * HZ);
+		return 0;
 	case -EBADHANDLE:
 	case -NFS4ERR_BAD_STATEID:
 		/*
@@ -3969,7 +3972,7 @@ static int nfsd4_cb_recall_done(struct nfsd4_callback *cb,
 		}
 		/*FALLTHRU*/
 	default:
-		return -1;
+		return 1;
 	}
 }
 
-- 
2.17.2

Re: [PATCH] nfsd: update callback done processing

[J. Bruce Fields]

Thanks, looks good!  Added one more paragraph with what you'd originally
seen (see below), and applied for 5.2.

I haven't added a stable cc, since I didn't think you'd established that
this in itself was causing a major problem--tell me if I'm wrong.

--b.

commit 1c73b9d24f80
Author: Scott Mayhew <smayhew@redhat.com>
Date:   Thu May 2 13:32:12 2019 -0400

    nfsd: update callback done processing
    
    Instead of having the convention where individual nfsd4_callback_ops->done
    operations return -1 to indicate the callback path is down, move the check
    to nfsd4_cb_done.  Only mark the callback path down on transport-level
    errors, not NFS-level errors.
    
    The existing logic causes the server to set SEQ4_STATUS_CB_PATH_DOWN
    just because the client returned an error to a CB_RECALL for a
    delegation that the client had already done a FREE_STATEID for.  But
    clearly that error doesn't mean that there's anything wrong with the
    backchannel.
    
    Additionally, handle NFS4ERR_DELAY in nfsd4_cb_recall_done.  The client
    returns NFS4ERR_DELAY if it is already in the process of returning the
    delegation.
    
    Signed-off-by: Scott Mayhew <smayhew@redhat.com>
    Signed-off-by: J. Bruce Fields <bfields@redhat.com>

diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c
index 7caa3801ce72..b9cbd7189d74 100644
--- a/fs/nfsd/nfs4callback.c
+++ b/fs/nfsd/nfs4callback.c
@@ -1122,10 +1122,11 @@ static void nfsd4_cb_done(struct rpc_task *task, void *calldata)
 		rpc_restart_call_prepare(task);
 		return;
 	case 1:
-		break;
-	case -1:
-		/* Network partition? */
-		nfsd4_mark_cb_down(clp, task->tk_status);
+		switch (task->tk_status) {
+		case -EIO:
+		case -ETIMEDOUT:
+			nfsd4_mark_cb_down(clp, task->tk_status);
+		}
 		break;
 	default:
 		BUG();
diff --git a/fs/nfsd/nfs4layouts.c b/fs/nfsd/nfs4layouts.c
index 44517fb5c0de..a79e24b79095 100644
--- a/fs/nfsd/nfs4layouts.c
+++ b/fs/nfsd/nfs4layouts.c
@@ -693,7 +693,7 @@ nfsd4_cb_layout_done(struct nfsd4_callback *cb, struct rpc_task *task)
 			ops->fence_client(ls);
 		else
 			nfsd4_cb_layout_fail(ls);
-		return -1;
+		return 1;
 	case -NFS4ERR_NOMATCHING_LAYOUT:
 		trace_nfsd_layout_recall_done(&ls->ls_stid.sc_stateid);
 		task->tk_status = 0;
diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c
index 8078314981f5..dfe4b596c2e5 100644
--- a/fs/nfsd/nfs4state.c
+++ b/fs/nfsd/nfs4state.c
@@ -3970,6 +3970,9 @@ static int nfsd4_cb_recall_done(struct nfsd4_callback *cb,
 	switch (task->tk_status) {
 	case 0:
 		return 1;
+	case -NFS4ERR_DELAY:
+		rpc_delay(task, 2 * HZ);
+		return 0;
 	case -EBADHANDLE:
 	case -NFS4ERR_BAD_STATEID:
 		/*
@@ -3982,7 +3985,7 @@ static int nfsd4_cb_recall_done(struct nfsd4_callback *cb,
 		}
 		/*FALLTHRU*/
 	default:
-		return -1;
+		return 1;
 	}
 }