Merge tag 'nfs-for-4.8-3' of git://git.linux-nfs.org/projects/trondmy/linux-nfs

Pull NFS client bugfixes from Trond Myklebust:
 "Highlights include:

  Stable patches:
   - Fix a refcount leak in nfs_callback_up_net
   - Fix an Oopsable condition when the flexfile pNFS driver connection
     to the DS fails
   - Fix an Oopsable condition in NFSv4.1 server callback races
   - Ensure pNFS clients stop doing I/O to the DS if their lease has
     expired, as required by the NFSv4.1 protocol

  Bugfixes:
   - Fix potential looping in the NFSv4.x migration code
   - Patch series to close callback races for OPEN, LAYOUTGET and
     LAYOUTRETURN
   - Silence WARN_ON when NFSv4.1 over RDMA is in use
   - Fix a LAYOUTCOMMIT race in the pNFS/blocks client
   - Fix pNFS timeout issues when the DS fails"

* tag 'nfs-for-4.8-3' of git://git.linux-nfs.org/projects/trondmy/linux-nfs:
  NFSv4.x: Fix a refcount leak in nfs_callback_up_net
  NFS4: Avoid migration loops
  pNFS/flexfiles: Fix an Oopsable condition when connection to the DS fails
  NFSv4.1: Remove obsolete and incorrrect assignment in nfs4_callback_sequence
  NFSv4.1: Close callback races for OPEN, LAYOUTGET and LAYOUTRETURN
  NFSv4.1: Defer bumping the slot sequence number until we free the slot
  NFSv4.1: Delay callback processing when there are referring triples
  NFSv4.1: Fix Oopsable condition in server callback races
  SUNRPC: Silence WARN_ON when NFSv4.1 over RDMA is in use
  pnfs/blocklayout: update last_write_offset atomically with extents
  pNFS: The client must not do I/O to the DS if it's lease has expired
  pNFS: Handle NFS4ERR_OLD_STATEID correctly in LAYOUTSTAT calls
  pNFS/flexfiles: Set reasonable default retrans values for the data channel
  NFS: Allow the mount option retrans=0
  pNFS/flexfiles: Fix layoutstat periodic reporting

Author: torvalds

fs/nfs/blocklayout/blocklayout.c

@@ -346,7 +346,7 @@ static void bl_write_cleanup(struct work_struct *work)
 			PAGE_SIZE - 1) & (loff_t)PAGE_MASK;
 
 		ext_tree_mark_written(bl, start >> SECTOR_SHIFT,
-					(end - start) >> SECTOR_SHIFT);
+					(end - start) >> SECTOR_SHIFT, end);
 	}
 
 	pnfs_ld_write_done(hdr);

fs/nfs/blocklayout/blocklayout.h

@@ -141,6 +141,7 @@ struct pnfs_block_layout {
 	struct rb_root		bl_ext_ro;
 	spinlock_t		bl_ext_lock;   /* Protects list manipulation */
 	bool			bl_scsi_layout;
+	u64			bl_lwb;
 };
 
 static inline struct pnfs_block_layout *
@@ -182,7 +183,7 @@ int ext_tree_insert(struct pnfs_block_layout *bl,
 int ext_tree_remove(struct pnfs_block_layout *bl, bool rw, sector_t start,
 		sector_t end);
 int ext_tree_mark_written(struct pnfs_block_layout *bl, sector_t start,
-		sector_t len);
+		sector_t len, u64 lwb);
 bool ext_tree_lookup(struct pnfs_block_layout *bl, sector_t isect,
 		struct pnfs_block_extent *ret, bool rw);
 int ext_tree_prepare_commit(struct nfs4_layoutcommit_args *arg);

fs/nfs/blocklayout/extent_tree.c

@@ -402,7 +402,7 @@ ext_tree_split(struct rb_root *root, struct pnfs_block_extent *be,
 
 int
 ext_tree_mark_written(struct pnfs_block_layout *bl, sector_t start,
-		sector_t len)
+		sector_t len, u64 lwb)
 {
 	struct rb_root *root = &bl->bl_ext_rw;
 	sector_t end = start + len;
@@ -471,6 +471,8 @@ ext_tree_mark_written(struct pnfs_block_layout *bl, sector_t start,
 		}
 	}
 out:
+	if (bl->bl_lwb < lwb)
+		bl->bl_lwb = lwb;
 	spin_unlock(&bl->bl_ext_lock);
 
 	__ext_put_deviceids(&tmp);
@@ -518,7 +520,7 @@ static __be32 *encode_scsi_range(struct pnfs_block_extent *be, __be32 *p)
 }
 
 static int ext_tree_encode_commit(struct pnfs_block_layout *bl, __be32 *p,
-		size_t buffer_size, size_t *count)
+		size_t buffer_size, size_t *count, __u64 *lastbyte)
 {
 	struct pnfs_block_extent *be;
 	int ret = 0;
@@ -542,6 +544,8 @@ static int ext_tree_encode_commit(struct pnfs_block_layout *bl, __be32 *p,
 			p = encode_block_extent(be, p);
 		be->be_tag = EXTENT_COMMITTING;
 	}
+	*lastbyte = bl->bl_lwb - 1;
+	bl->bl_lwb = 0;
 	spin_unlock(&bl->bl_ext_lock);
 
 	return ret;
@@ -564,7 +568,7 @@ ext_tree_prepare_commit(struct nfs4_layoutcommit_args *arg)
 	arg->layoutupdate_pages = &arg->layoutupdate_page;
 
 retry:
-	ret = ext_tree_encode_commit(bl, start_p + 1, buffer_size, &count);
+	ret = ext_tree_encode_commit(bl, start_p + 1, buffer_size, &count, &arg->lastbytewritten);
 	if (unlikely(ret)) {
 		ext_tree_free_commitdata(arg, buffer_size);
 

fs/nfs/callback.c

@@ -275,6 +275,7 @@ static int nfs_callback_up_net(int minorversion, struct svc_serv *serv,
 err_socks:
 	svc_rpcb_cleanup(serv, net);
 err_bind:
+	nn->cb_users[minorversion]--;
 	dprintk("NFS: Couldn't create callback socket: err = %d; "
 			"net = %p\n", ret, net);
 	return ret;

fs/nfs/callback_proc.c

@@ -454,11 +454,8 @@ static bool referring_call_exists(struct nfs_client *clp,
 				((u32 *)&rclist->rcl_sessionid.data)[3],
 				ref->rc_sequenceid, ref->rc_slotid);
 
-			spin_lock(&tbl->slot_tbl_lock);
-			status = (test_bit(ref->rc_slotid, tbl->used_slots) &&
-				  tbl->slots[ref->rc_slotid].seq_nr ==
-					ref->rc_sequenceid);
-			spin_unlock(&tbl->slot_tbl_lock);
+			status = nfs4_slot_wait_on_seqid(tbl, ref->rc_slotid,
+					ref->rc_sequenceid, HZ >> 1) < 0;
 			if (status)
 				goto out;
 		}
@@ -487,7 +484,6 @@ __be32 nfs4_callback_sequence(struct cb_sequenceargs *args,
 		goto out;
 
 	tbl = &clp->cl_session->bc_slot_table;
-	slot = tbl->slots + args->csa_slotid;
 
 	/* Set up res before grabbing the spinlock */
 	memcpy(&res->csr_sessionid, &args->csa_sessionid,

fs/nfs/client.c

@@ -426,17 +426,17 @@ EXPORT_SYMBOL_GPL(nfs_mark_client_ready);
  * Initialise the timeout values for a connection
  */
 void nfs_init_timeout_values(struct rpc_timeout *to, int proto,
-				    unsigned int timeo, unsigned int retrans)
+				    int timeo, int retrans)
 {
 	to->to_initval = timeo * HZ / 10;
 	to->to_retries = retrans;
 
 	switch (proto) {
 	case XPRT_TRANSPORT_TCP:
 	case XPRT_TRANSPORT_RDMA:
-		if (to->to_retries == 0)
+		if (retrans == NFS_UNSPEC_RETRANS)
 			to->to_retries = NFS_DEF_TCP_RETRANS;
-		if (to->to_initval == 0)
+		if (timeo == NFS_UNSPEC_TIMEO || to->to_retries == 0)
 			to->to_initval = NFS_DEF_TCP_TIMEO * HZ / 10;
 		if (to->to_initval > NFS_MAX_TCP_TIMEOUT)
 			to->to_initval = NFS_MAX_TCP_TIMEOUT;
@@ -449,9 +449,9 @@ void nfs_init_timeout_values(struct rpc_timeout *to, int proto,
 		to->to_exponential = 0;
 		break;
 	case XPRT_TRANSPORT_UDP:
-		if (to->to_retries == 0)
+		if (retrans == NFS_UNSPEC_RETRANS)
 			to->to_retries = NFS_DEF_UDP_RETRANS;
-		if (!to->to_initval)
+		if (timeo == NFS_UNSPEC_TIMEO || to->to_initval == 0)
 			to->to_initval = NFS_DEF_UDP_TIMEO * HZ / 10;
 		if (to->to_initval > NFS_MAX_UDP_TIMEOUT)
 			to->to_initval = NFS_MAX_UDP_TIMEOUT;

fs/nfs/flexfilelayout/flexfilelayout.c

@@ -37,6 +37,7 @@ ff_layout_alloc_layout_hdr(struct inode *inode, gfp_t gfp_flags)
 	if (ffl) {
 		INIT_LIST_HEAD(&ffl->error_list);
 		INIT_LIST_HEAD(&ffl->mirrors);
+		ffl->last_report_time = ktime_get();
 		return &ffl->generic_hdr;
 	} else
 		return NULL;
@@ -640,19 +641,18 @@ nfs4_ff_layoutstat_start_io(struct nfs4_ff_layout_mirror *mirror,
 {
 	static const ktime_t notime = {0};
 	s64 report_interval = FF_LAYOUTSTATS_REPORT_INTERVAL;
+	struct nfs4_flexfile_layout *ffl = FF_LAYOUT_FROM_HDR(mirror->layout);
 
 	nfs4_ff_start_busy_timer(&layoutstat->busy_timer, now);
 	if (ktime_equal(mirror->start_time, notime))
 		mirror->start_time = now;
-	if (ktime_equal(mirror->last_report_time, notime))
-		mirror->last_report_time = now;
 	if (mirror->report_interval != 0)
 		report_interval = (s64)mirror->report_interval * 1000LL;
 	else if (layoutstats_timer != 0)
 		report_interval = (s64)layoutstats_timer * 1000LL;
-	if (ktime_to_ms(ktime_sub(now, mirror->last_report_time)) >=
+	if (ktime_to_ms(ktime_sub(now, ffl->last_report_time)) >=
 			report_interval) {
-		mirror->last_report_time = now;
+		ffl->last_report_time = now;
 		return true;
 	}
 
@@ -806,11 +806,14 @@ ff_layout_choose_best_ds_for_read(struct pnfs_layout_segment *lseg,
 {
 	struct nfs4_ff_layout_segment *fls = FF_LAYOUT_LSEG(lseg);
 	struct nfs4_pnfs_ds *ds;
+	bool fail_return = false;
 	int idx;
 
 	/* mirrors are sorted by efficiency */
 	for (idx = start_idx; idx < fls->mirror_array_cnt; idx++) {
-		ds = nfs4_ff_layout_prepare_ds(lseg, idx, false);
+		if (idx+1 == fls->mirror_array_cnt)
+			fail_return = true;
+		ds = nfs4_ff_layout_prepare_ds(lseg, idx, fail_return);
 		if (ds) {
 			*best_idx = idx;
 			return ds;
@@ -859,6 +862,7 @@ ff_layout_pg_init_read(struct nfs_pageio_descriptor *pgio,
 	struct nfs4_pnfs_ds *ds;
 	int ds_idx;
 
+retry:
 	/* Use full layout for now */
 	if (!pgio->pg_lseg)
 		ff_layout_pg_get_read(pgio, req, false);
@@ -871,10 +875,13 @@ ff_layout_pg_init_read(struct nfs_pageio_descriptor *pgio,
 
 	ds = ff_layout_choose_best_ds_for_read(pgio->pg_lseg, 0, &ds_idx);
 	if (!ds) {
-		if (ff_layout_no_fallback_to_mds(pgio->pg_lseg))
-			goto out_pnfs;
-		else
+		if (!ff_layout_no_fallback_to_mds(pgio->pg_lseg))
 			goto out_mds;
+		pnfs_put_lseg(pgio->pg_lseg);
+		pgio->pg_lseg = NULL;
+		/* Sleep for 1 second before retrying */
+		ssleep(1);
+		goto retry;
 	}
 
 	mirror = FF_LAYOUT_COMP(pgio->pg_lseg, ds_idx);
@@ -890,12 +897,6 @@ ff_layout_pg_init_read(struct nfs_pageio_descriptor *pgio,
 	pnfs_put_lseg(pgio->pg_lseg);
 	pgio->pg_lseg = NULL;
 	nfs_pageio_reset_read_mds(pgio);
-	return;
-
-out_pnfs:
-	pnfs_set_lo_fail(pgio->pg_lseg);
-	pnfs_put_lseg(pgio->pg_lseg);
-	pgio->pg_lseg = NULL;
 }
 
 static void
@@ -909,6 +910,7 @@ ff_layout_pg_init_write(struct nfs_pageio_descriptor *pgio,
 	int i;
 	int status;
 
+retry:
 	if (!pgio->pg_lseg) {
 		pgio->pg_lseg = pnfs_update_layout(pgio->pg_inode,
 						   req->wb_context,
@@ -940,10 +942,13 @@ ff_layout_pg_init_write(struct nfs_pageio_descriptor *pgio,
 	for (i = 0; i < pgio->pg_mirror_count; i++) {
 		ds = nfs4_ff_layout_prepare_ds(pgio->pg_lseg, i, true);
 		if (!ds) {
-			if (ff_layout_no_fallback_to_mds(pgio->pg_lseg))
-				goto out_pnfs;
-			else
+			if (!ff_layout_no_fallback_to_mds(pgio->pg_lseg))
 				goto out_mds;
+			pnfs_put_lseg(pgio->pg_lseg);
+			pgio->pg_lseg = NULL;
+			/* Sleep for 1 second before retrying */
+			ssleep(1);
+			goto retry;
 		}
 		pgm = &pgio->pg_mirrors[i];
 		mirror = FF_LAYOUT_COMP(pgio->pg_lseg, i);
@@ -956,12 +961,6 @@ ff_layout_pg_init_write(struct nfs_pageio_descriptor *pgio,
 	pnfs_put_lseg(pgio->pg_lseg);
 	pgio->pg_lseg = NULL;
 	nfs_pageio_reset_write_mds(pgio);
-	return;
-
-out_pnfs:
-	pnfs_set_lo_fail(pgio->pg_lseg);
-	pnfs_put_lseg(pgio->pg_lseg);
-	pgio->pg_lseg = NULL;
 }
 
 static unsigned int

fs/nfs/flexfilelayout/flexfilelayout.h

@@ -84,7 +84,6 @@ struct nfs4_ff_layout_mirror {
 	struct nfs4_ff_layoutstat	read_stat;
 	struct nfs4_ff_layoutstat	write_stat;
 	ktime_t				start_time;
-	ktime_t				last_report_time;
 	u32				report_interval;
 };
 
@@ -101,6 +100,7 @@ struct nfs4_flexfile_layout {
 	struct pnfs_ds_commit_info commit_info;
 	struct list_head	mirrors;
 	struct list_head	error_list; /* nfs4_ff_layout_ds_err */
+	ktime_t			last_report_time; /* Layoutstat report times */
 };
 
 static inline struct nfs4_flexfile_layout *

fs/nfs/flexfilelayout/flexfilelayoutdev.c

@@ -17,8 +17,8 @@
 
 #define NFSDBG_FACILITY		NFSDBG_PNFS_LD
 
-static unsigned int dataserver_timeo = NFS4_DEF_DS_TIMEO;
-static unsigned int dataserver_retrans = NFS4_DEF_DS_RETRANS;
+static unsigned int dataserver_timeo = NFS_DEF_TCP_RETRANS;
+static unsigned int dataserver_retrans;
 
 void nfs4_ff_layout_put_deviceid(struct nfs4_ff_layout_ds *mirror_ds)
 {
@@ -379,7 +379,7 @@ nfs4_ff_layout_prepare_ds(struct pnfs_layout_segment *lseg, u32 ds_idx,
 
 	devid = &mirror->mirror_ds->id_node;
 	if (ff_layout_test_devid_unavailable(devid))
-		goto out;
+		goto out_fail;
 
 	ds = mirror->mirror_ds->ds;
 	/* matching smp_wmb() in _nfs4_pnfs_v3/4_ds_connect */
@@ -405,15 +405,16 @@ nfs4_ff_layout_prepare_ds(struct pnfs_layout_segment *lseg, u32 ds_idx,
 			mirror->mirror_ds->ds_versions[0].rsize = max_payload;
 		if (mirror->mirror_ds->ds_versions[0].wsize > max_payload)
 			mirror->mirror_ds->ds_versions[0].wsize = max_payload;
-	} else {
-		ff_layout_track_ds_error(FF_LAYOUT_FROM_HDR(lseg->pls_layout),
-					 mirror, lseg->pls_range.offset,
-					 lseg->pls_range.length, NFS4ERR_NXIO,
-					 OP_ILLEGAL, GFP_NOIO);
-		if (fail_return || !ff_layout_has_available_ds(lseg))
-			pnfs_error_mark_layout_for_return(ino, lseg);
-		ds = NULL;
+		goto out;
 	}
+	ff_layout_track_ds_error(FF_LAYOUT_FROM_HDR(lseg->pls_layout),
+				 mirror, lseg->pls_range.offset,
+				 lseg->pls_range.length, NFS4ERR_NXIO,
+				 OP_ILLEGAL, GFP_NOIO);
+out_fail:
+	if (fail_return || !ff_layout_has_available_ds(lseg))
+		pnfs_error_mark_layout_for_return(ino, lseg);
+	ds = NULL;
 out:
 	return ds;
 }

fs/nfs/internal.h

@@ -58,6 +58,9 @@ struct nfs_clone_mount {
  */
 #define NFS_UNSPEC_PORT		(-1)
 
+#define NFS_UNSPEC_RETRANS	(UINT_MAX)
+#define NFS_UNSPEC_TIMEO	(UINT_MAX)
+
 /*
  * Maximum number of pages that readdir can use for creating
  * a vmapped array of pages.
@@ -156,7 +159,7 @@ struct nfs_client *nfs_get_client(const struct nfs_client_initdata *,
 int nfs_probe_fsinfo(struct nfs_server *server, struct nfs_fh *, struct nfs_fattr *);
 void nfs_server_insert_lists(struct nfs_server *);
 void nfs_server_remove_lists(struct nfs_server *);
-void nfs_init_timeout_values(struct rpc_timeout *, int, unsigned int, unsigned int);
+void nfs_init_timeout_values(struct rpc_timeout *to, int proto, int timeo, int retrans);
 int nfs_init_server_rpcclient(struct nfs_server *, const struct rpc_timeout *t,
 		rpc_authflavor_t);
 struct nfs_server *nfs_alloc_server(void);