Merge tag 'nfsd-5.5' of git://linux-nfs.org/~bfields/linux

Pull nfsd updates from Bruce Fields:
 "This is a relatively quiet cycle for nfsd, mainly various bugfixes.

  Possibly most interesting is Trond's fixes for some callback races
  that were due to my incomplete understanding of rpc client shutdown.
  Unfortunately at the last minute I've started noticing a new
  intermittent failure to send callbacks. As the logic seems basically
  correct, I'm leaving Trond's patches in for now, and hope to find a
  fix in the next week so I don't have to revert those patches"

* tag 'nfsd-5.5' of git://linux-nfs.org/~bfields/linux: (24 commits)
  nfsd: depend on CRYPTO_MD5 for legacy client tracking
  NFSD fixing possible null pointer derefering in copy offload
  nfsd: check for EBUSY from vfs_rmdir/vfs_unink.
  nfsd: Ensure CLONE persists data and metadata changes to the target file
  SUNRPC: Fix backchannel latency metrics
  nfsd: restore NFSv3 ACL support
  nfsd: v4 support requires CRYPTO_SHA256
  nfsd: Fix cld_net->cn_tfm initialization
  lockd: remove __KERNEL__ ifdefs
  sunrpc: remove __KERNEL__ ifdefs
  race in exportfs_decode_fh()
  nfsd: Drop LIST_HEAD where the variable it declares is never used.
  nfsd: document callback_wq serialization of callback code
  nfsd: mark cb path down on unknown errors
  nfsd: Fix races between nfsd4_cb_release() and nfsd4_shutdown_callback()
  nfsd: minor 4.1 callback cleanup
  SUNRPC: Fix svcauth_gss_proxy_init()
  SUNRPC: Trace gssproxy upcall results
  sunrpc: fix crash when cache_head become valid before update
  nfsd: remove private bin2hex implementation
  ...

12 files changed, 136 insertions, 52 deletions

diff --git a/fs/nfsd/Kconfig b/fs/nfsd/Kconfig
index 10cefb0c07c7..f2f81561ebb6 100644
--- a/fs/nfsd/Kconfig
+++ b/fs/nfsd/Kconfig
@@ -73,7 +73,8 @@ config NFSD_V4
 	select NFSD_V3
 	select FS_POSIX_ACL
 	select SUNRPC_GSS
-	select CRYPTO
+	select CRYPTO_MD5
+	select CRYPTO_SHA256
 	select GRACE_PERIOD
 	help
 	  This option enables support in your system's NFS server for
diff --git a/fs/nfsd/filecache.c b/fs/nfsd/filecache.c
index ef55e9b1cd4e..32a9bf22ac08 100644
--- a/fs/nfsd/filecache.c
+++ b/fs/nfsd/filecache.c
@@ -685,8 +685,6 @@ nfsd_file_cache_purge(struct net *net)
 void
 nfsd_file_cache_shutdown(void)
 {
-	LIST_HEAD(dispose);
-
 	set_bit(NFSD_FILE_SHUTDOWN, &nfsd_file_lru_flags);
 
 	lease_unregister_notifier(&nfsd_file_lease_notifier);
diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c
index 524111420b48..24534db87e86 100644
--- a/fs/nfsd/nfs4callback.c
+++ b/fs/nfsd/nfs4callback.c
@@ -826,6 +826,31 @@ static int max_cb_time(struct net *net)
 	return max(nn->nfsd4_lease/10, (time_t)1) * HZ;
 }
 
+static struct workqueue_struct *callback_wq;
+
+static bool nfsd4_queue_cb(struct nfsd4_callback *cb)
+{
+	return queue_work(callback_wq, &cb->cb_work);
+}
+
+static void nfsd41_cb_inflight_begin(struct nfs4_client *clp)
+{
+	atomic_inc(&clp->cl_cb_inflight);
+}
+
+static void nfsd41_cb_inflight_end(struct nfs4_client *clp)
+{
+
+	if (atomic_dec_and_test(&clp->cl_cb_inflight))
+		wake_up_var(&clp->cl_cb_inflight);
+}
+
+static void nfsd41_cb_inflight_wait_complete(struct nfs4_client *clp)
+{
+	wait_var_event(&clp->cl_cb_inflight,
+			!atomic_read(&clp->cl_cb_inflight));
+}
+
 static const struct cred *get_backchannel_cred(struct nfs4_client *clp, struct rpc_clnt *client, struct nfsd4_session *ses)
 {
 	if (clp->cl_minorversion == 0) {
@@ -937,14 +962,21 @@ static void nfsd4_cb_probe_done(struct rpc_task *task, void *calldata)
 		clp->cl_cb_state = NFSD4_CB_UP;
 }
 
+static void nfsd4_cb_probe_release(void *calldata)
+{
+	struct nfs4_client *clp = container_of(calldata, struct nfs4_client, cl_cb_null);
+
+	nfsd41_cb_inflight_end(clp);
+
+}
+
 static const struct rpc_call_ops nfsd4_cb_probe_ops = {
 	/* XXX: release method to ensure we set the cb channel down if
 	 * necessary on early failure? */
 	.rpc_call_done = nfsd4_cb_probe_done,
+	.rpc_release = nfsd4_cb_probe_release,
 };
 
-static struct workqueue_struct *callback_wq;
-
 /*
  * Poke the callback thread to process any updates to the callback
  * parameters, and send a null probe.
@@ -975,9 +1007,12 @@ void nfsd4_change_callback(struct nfs4_client *clp, struct nfs4_cb_conn *conn)
  * If the slot is available, then mark it busy.  Otherwise, set the
  * thread for sleeping on the callback RPC wait queue.
  */
-static bool nfsd41_cb_get_slot(struct nfs4_client *clp, struct rpc_task *task)
+static bool nfsd41_cb_get_slot(struct nfsd4_callback *cb, struct rpc_task *task)
 {
-	if (test_and_set_bit(0, &clp->cl_cb_slot_busy) != 0) {
+	struct nfs4_client *clp = cb->cb_clp;
+
+	if (!cb->cb_holds_slot &&
+	    test_and_set_bit(0, &clp->cl_cb_slot_busy) != 0) {
 		rpc_sleep_on(&clp->cl_cb_waitq, task, NULL);
 		/* Race breaker */
 		if (test_and_set_bit(0, &clp->cl_cb_slot_busy) != 0) {
@@ -986,9 +1021,31 @@ static bool nfsd41_cb_get_slot(struct nfs4_client *clp, struct rpc_task *task)
 		}
 		rpc_wake_up_queued_task(&clp->cl_cb_waitq, task);
 	}
+	cb->cb_holds_slot = true;
 	return true;
 }
 
+static void nfsd41_cb_release_slot(struct nfsd4_callback *cb)
+{
+	struct nfs4_client *clp = cb->cb_clp;
+
+	if (cb->cb_holds_slot) {
+		cb->cb_holds_slot = false;
+		clear_bit(0, &clp->cl_cb_slot_busy);
+		rpc_wake_up_next(&clp->cl_cb_waitq);
+	}
+}
+
+static void nfsd41_destroy_cb(struct nfsd4_callback *cb)
+{
+	struct nfs4_client *clp = cb->cb_clp;
+
+	nfsd41_cb_release_slot(cb);
+	if (cb->cb_ops && cb->cb_ops->release)
+		cb->cb_ops->release(cb);
+	nfsd41_cb_inflight_end(clp);
+}
+
 /*
  * TODO: cb_sequence should support referring call lists, cachethis, multiple
  * slots, and mark callback channel down on communication errors.
@@ -1005,11 +1062,8 @@ static void nfsd4_cb_prepare(struct rpc_task *task, void *calldata)
 	 */
 	cb->cb_seq_status = 1;
 	cb->cb_status = 0;
-	if (minorversion) {
-		if (!cb->cb_holds_slot && !nfsd41_cb_get_slot(clp, task))
-			return;
-		cb->cb_holds_slot = true;
-	}
+	if (minorversion && !nfsd41_cb_get_slot(cb, task))
+		return;
 	rpc_call_start(task);
 }
 
@@ -1072,13 +1126,12 @@ static bool nfsd4_cb_sequence_done(struct rpc_task *task, struct nfsd4_callback
 		}
 		break;
 	default:
+		nfsd4_mark_cb_fault(cb->cb_clp, cb->cb_seq_status);
 		dprintk("%s: unprocessed error %d\n", __func__,
 			cb->cb_seq_status);
 	}
 
-	cb->cb_holds_slot = false;
-	clear_bit(0, &clp->cl_cb_slot_busy);
-	rpc_wake_up_next(&clp->cl_cb_waitq);
+	nfsd41_cb_release_slot(cb);
 	dprintk("%s: freed slot, new seqid=%d\n", __func__,
 		clp->cl_cb_session->se_cb_seq_nr);
 
@@ -1091,8 +1144,10 @@ retry_nowait:
 		ret = false;
 	goto out;
 need_restart:
-	task->tk_status = 0;
-	cb->cb_need_restart = true;
+	if (!test_bit(NFSD4_CLIENT_CB_KILL, &clp->cl_flags)) {
+		task->tk_status = 0;
+		cb->cb_need_restart = true;
+	}
 	return false;
 }
 
@@ -1134,9 +1189,9 @@ static void nfsd4_cb_release(void *calldata)
 	struct nfsd4_callback *cb = calldata;
 
 	if (cb->cb_need_restart)
-		nfsd4_run_cb(cb);
+		nfsd4_queue_cb(cb);
 	else
-		cb->cb_ops->release(cb);
+		nfsd41_destroy_cb(cb);
 
 }
 
@@ -1170,6 +1225,7 @@ void nfsd4_shutdown_callback(struct nfs4_client *clp)
 	 */
 	nfsd4_run_cb(&clp->cl_cb_null);
 	flush_workqueue(callback_wq);
+	nfsd41_cb_inflight_wait_complete(clp);
 }
 
 /* requires cl_lock: */
@@ -1187,6 +1243,12 @@ static struct nfsd4_conn * __nfsd4_find_backchannel(struct nfs4_client *clp)
 	return NULL;
 }
 
+/*
+ * Note there isn't a lot of locking in this code; instead we depend on
+ * the fact that it is run from the callback_wq, which won't run two
+ * work items at once.  So, for example, callback_wq handles all access
+ * of cl_cb_client and all calls to rpc_create or rpc_shutdown_client.
+ */
 static void nfsd4_process_cb_update(struct nfsd4_callback *cb)
 {
 	struct nfs4_cb_conn conn;
@@ -1255,8 +1317,7 @@ nfsd4_run_cb_work(struct work_struct *work)
 	clnt = clp->cl_cb_client;
 	if (!clnt) {
 		/* Callback channel broken, or client killed; give up: */
-		if (cb->cb_ops && cb->cb_ops->release)
-			cb->cb_ops->release(cb);
+		nfsd41_destroy_cb(cb);
 		return;
 	}
 
@@ -1265,6 +1326,7 @@ nfsd4_run_cb_work(struct work_struct *work)
 	 */
 	if (!cb->cb_ops && clp->cl_minorversion) {
 		clp->cl_cb_state = NFSD4_CB_UP;
+		nfsd41_destroy_cb(cb);
 		return;
 	}
 
@@ -1290,5 +1352,9 @@ void nfsd4_init_cb(struct nfsd4_callback *cb, struct nfs4_client *clp,
 
 void nfsd4_run_cb(struct nfsd4_callback *cb)
 {
-	queue_work(callback_wq, &cb->cb_work);
+	struct nfs4_client *clp = cb->cb_clp;
+
+	nfsd41_cb_inflight_begin(clp);
+	if (!nfsd4_queue_cb(cb))
+		nfsd41_cb_inflight_end(clp);
 }
diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c
index 4e3e77b76411..4798667af647 100644
--- a/fs/nfsd/nfs4proc.c
+++ b/fs/nfsd/nfs4proc.c
@@ -1077,7 +1077,8 @@ nfsd4_clone(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 		goto out;
 
 	status = nfsd4_clone_file_range(src->nf_file, clone->cl_src_pos,
-			dst->nf_file, clone->cl_dst_pos, clone->cl_count);
+			dst->nf_file, clone->cl_dst_pos, clone->cl_count,
+			EX_ISSYNC(cstate->current_fh.fh_export));
 
 	nfsd_file_put(dst);
 	nfsd_file_put(src);
@@ -1297,7 +1298,8 @@ nfsd4_copy(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 out:
 	return status;
 out_err:
-	cleanup_async_copy(async_copy);
+	if (async_copy)
+		cleanup_async_copy(async_copy);
 	goto out;
 }
 
diff --git a/fs/nfsd/nfs4recover.c b/fs/nfsd/nfs4recover.c
index cdc75ad4438b..2481e7662128 100644
--- a/fs/nfsd/nfs4recover.c
+++ b/fs/nfsd/nfs4recover.c
@@ -1578,6 +1578,7 @@ nfsd4_cld_tracking_init(struct net *net)
 	struct nfsd_net *nn = net_generic(net, nfsd_net_id);
 	bool running;
 	int retries = 10;
+	struct crypto_shash *tfm;
 
 	status = nfs4_cld_state_init(net);
 	if (status)
@@ -1586,11 +1587,6 @@ nfsd4_cld_tracking_init(struct net *net)
 	status = __nfsd4_init_cld_pipe(net);
 	if (status)
 		goto err_shutdown;
-	nn->cld_net->cn_tfm = crypto_alloc_shash("sha256", 0, 0);
-	if (IS_ERR(nn->cld_net->cn_tfm)) {
-		status = PTR_ERR(nn->cld_net->cn_tfm);
-		goto err_remove;
-	}
 
 	/*
 	 * rpc pipe upcalls take 30 seconds to time out, so we don't want to
@@ -1607,6 +1603,12 @@ nfsd4_cld_tracking_init(struct net *net)
 		status = -ETIMEDOUT;
 		goto err_remove;
 	}
+	tfm = crypto_alloc_shash("sha256", 0, 0);
+	if (IS_ERR(tfm)) {
+		status = PTR_ERR(tfm);
+		goto err_remove;
+	}
+	nn->cld_net->cn_tfm = tfm;
 
 	status = nfsd4_cld_get_version(nn);
 	if (status == -EOPNOTSUPP)
@@ -1850,19 +1852,14 @@ nfsd4_umh_cltrack_upcall(char *cmd, char *arg, char *env0, char *env1)
 static char *
 bin_to_hex_dup(const unsigned char *src, int srclen)
 {
-	int i;
-	char *buf, *hex;
+	char *buf;
 
 	/* +1 for terminating NULL */
-	buf = kmalloc((srclen * 2) + 1, GFP_KERNEL);
+	buf = kzalloc((srclen * 2) + 1, GFP_KERNEL);
 	if (!buf)
 		return buf;
 
-	hex = buf;
-	for (i = 0; i < srclen; i++) {
-		sprintf(hex, "%2.2x", *src++);
-		hex += 2;
-	}
+	bin2hex(buf, src, srclen);
 	return buf;
 }
 
diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c
index c65aeaa812d4..369e574c5092 100644
--- a/fs/nfsd/nfs4state.c
+++ b/fs/nfsd/nfs4state.c
@@ -2382,10 +2382,10 @@ static int nfs4_show_open(struct seq_file *s, struct nfs4_stid *st)
 	access = bmap_to_share_mode(ols->st_access_bmap);
 	deny   = bmap_to_share_mode(ols->st_deny_bmap);
 
-	seq_printf(s, "access: \%s\%s, ",
+	seq_printf(s, "access: %s%s, ",
 		access & NFS4_SHARE_ACCESS_READ ? "r" : "-",
 		access & NFS4_SHARE_ACCESS_WRITE ? "w" : "-");
-	seq_printf(s, "deny: \%s\%s, ",
+	seq_printf(s, "deny: %s%s, ",
 		deny & NFS4_SHARE_ACCESS_READ ? "r" : "-",
 		deny & NFS4_SHARE_ACCESS_WRITE ? "w" : "-");
 
@@ -3548,12 +3548,17 @@ static bool replay_matches_cache(struct svc_rqst *rqstp,
 	    (bool)seq->cachethis)
 		return false;
 	/*
-	 * If there's an error than the reply can have fewer ops than
-	 * the call.  But if we cached a reply with *more* ops than the
-	 * call you're sending us now, then this new call is clearly not
-	 * really a replay of the old one:
+	 * If there's an error then the reply can have fewer ops than
+	 * the call.
 	 */
-	if (slot->sl_opcnt < argp->opcnt)
+	if (slot->sl_opcnt < argp->opcnt && !slot->sl_status)
+		return false;
+	/*
+	 * But if we cached a reply with *more* ops than the call you're
+	 * sending us now, then this new call is clearly not really a
+	 * replay of the old one:
+	 */
+	if (slot->sl_opcnt > argp->opcnt)
 		return false;
 	/* This is the only check explicitly called by spec: */
 	if (!same_creds(&rqstp->rq_cred, &slot->sl_cred))
diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c
index b09237431ae2..d2dc4c0e22e8 100644
--- a/fs/nfsd/nfs4xdr.c
+++ b/fs/nfsd/nfs4xdr.c
@@ -3452,7 +3452,6 @@ static __be32 nfsd4_encode_splice_read(
 	struct xdr_stream *xdr = &resp->xdr;
 	struct xdr_buf *buf = xdr->buf;
 	u32 eof;
-	long len;
 	int space_left;
 	__be32 nfserr;
 	__be32 *p = xdr->p - 2;
@@ -3461,7 +3460,6 @@ static __be32 nfsd4_encode_splice_read(
 	if (xdr->end - xdr->p < 1)
 		return nfserr_resource;
 
-	len = maxcount;
 	nfserr = nfsd_splice_read(read->rd_rqstp, read->rd_fhp,
 				  file, read->rd_offset, &maxcount, &eof);
 	read->rd_length = maxcount;
diff --git a/fs/nfsd/nfsd.h b/fs/nfsd/nfsd.h
index af2947551e9c..57b93d95fa5c 100644
--- a/fs/nfsd/nfsd.h
+++ b/fs/nfsd/nfsd.h
@@ -280,7 +280,8 @@ void		nfsd_lockd_shutdown(void);
 #define nfserr_union_notsupp		cpu_to_be32(NFS4ERR_UNION_NOTSUPP)
 #define nfserr_offload_denied		cpu_to_be32(NFS4ERR_OFFLOAD_DENIED)
 #define nfserr_wrong_lfs		cpu_to_be32(NFS4ERR_WRONG_LFS)
-#define nfserr_badlabel		cpu_to_be32(NFS4ERR_BADLABEL)
+#define nfserr_badlabel			cpu_to_be32(NFS4ERR_BADLABEL)
+#define nfserr_file_open		cpu_to_be32(NFS4ERR_FILE_OPEN)
 
 /* error codes for internal use */
 /* if a request fails due to kmalloc failure, it gets dropped.
diff --git a/fs/nfsd/nfssvc.c b/fs/nfsd/nfssvc.c
index fdf7ed4bd5dd..e8bee8ff30c5 100644
--- a/fs/nfsd/nfssvc.c
+++ b/fs/nfsd/nfssvc.c
@@ -95,12 +95,11 @@ static const struct svc_version *nfsd_acl_version[] = {
 
 #define NFSD_ACL_MINVERS            2
 #define NFSD_ACL_NRVERS		ARRAY_SIZE(nfsd_acl_version)
-static const struct svc_version *nfsd_acl_versions[NFSD_ACL_NRVERS];
 
 static struct svc_program	nfsd_acl_program = {
 	.pg_prog		= NFS_ACL_PROGRAM,
 	.pg_nvers		= NFSD_ACL_NRVERS,
-	.pg_vers		= nfsd_acl_versions,
+	.pg_vers		= nfsd_acl_version,
 	.pg_name		= "nfsacl",
 	.pg_class		= "nfsd",
 	.pg_stats		= &nfsd_acl_svcstats,
diff --git a/fs/nfsd/state.h b/fs/nfsd/state.h
index 46f56afb6cb8..d61b83b9654c 100644
--- a/fs/nfsd/state.h
+++ b/fs/nfsd/state.h
@@ -367,6 +367,7 @@ struct nfs4_client {
 	struct net		*net;
 	struct list_head	async_copies;	/* list of async copies */
 	spinlock_t		async_lock;	/* lock for async copies */
+	atomic_t		cl_cb_inflight;	/* Outstanding callbacks */
 };
 
 /* struct nfs4_client_reset
diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c
index bd0a385df3fc..c0dc491537a6 100644
--- a/fs/nfsd/vfs.c
+++ b/fs/nfsd/vfs.c
@@ -525,7 +525,7 @@ __be32 nfsd4_set_nfs4_label(struct svc_rqst *rqstp, struct svc_fh *fhp,
 #endif
 
 __be32 nfsd4_clone_file_range(struct file *src, u64 src_pos, struct file *dst,
-		u64 dst_pos, u64 count)
+		u64 dst_pos, u64 count, bool sync)
 {
 	loff_t cloned;
 
@@ -534,6 +534,12 @@ __be32 nfsd4_clone_file_range(struct file *src, u64 src_pos, struct file *dst,
 		return nfserrno(cloned);
 	if (count && cloned != count)
 		return nfserrno(-EINVAL);
+	if (sync) {
+		loff_t dst_end = count ? dst_pos + count - 1 : LLONG_MAX;
+		int status = vfs_fsync_range(dst, dst_pos, dst_end, 0);
+		if (status < 0)
+			return nfserrno(status);
+	}
 	return 0;
 }
 
@@ -1809,7 +1815,17 @@ nfsd_unlink(struct svc_rqst *rqstp, struct svc_fh *fhp, int type,
 out_drop_write:
 	fh_drop_write(fhp);
 out_nfserr:
-	err = nfserrno(host_err);
+	if (host_err == -EBUSY) {
+		/* name is mounted-on. There is no perfect
+		 * error status.
+		 */
+		if (nfsd_v4client(rqstp))
+			err = nfserr_file_open;
+		else
+			err = nfserr_acces;
+	} else {
+		err = nfserrno(host_err);
+	}
 out:
 	return err;
 }
diff --git a/fs/nfsd/vfs.h b/fs/nfsd/vfs.h
index a13fd9d7e1f5..cc110a10bfe8 100644
--- a/fs/nfsd/vfs.h
+++ b/fs/nfsd/vfs.h
@@ -56,7 +56,7 @@ __be32          nfsd4_set_nfs4_label(struct svc_rqst *, struct svc_fh *,
 __be32		nfsd4_vfs_fallocate(struct svc_rqst *, struct svc_fh *,
 				    struct file *, loff_t, loff_t, int);
 __be32		nfsd4_clone_file_range(struct file *, u64, struct file *,
-			u64, u64);
+				       u64, u64, bool);
 #endif /* CONFIG_NFSD_V4 */
 __be32		nfsd_create_locked(struct svc_rqst *, struct svc_fh *,
 				char *name, int len, struct iattr *attrs,