NFS client updates for Linux 3.19

Highlights include:
 
 Features:
 - NFSv4.2 client support for hole punching and preallocation.
 - Further RPC/RDMA client improvements.
 - Add more RPC transport debugging tracepoints.
 - Add RPC debugging tools in debugfs.
 
 Bugfixes:
 - Stable fix for layoutget error handling
 - Fix a change in COMMIT behaviour resulting from the recent io code updates
 -----BEGIN PGP SIGNATURE-----
 Version: GnuPG v1
 
 iQIcBAABAgAGBQJUhRVTAAoJEGcL54qWCgDyfeUP/RoFo3ImTMbGxfcPJqoELjcO
 lZbQ+27pOE/whFDkWgiOVTwlgGct5a0WRo7GCZmpYJA4q1kmSv4ngTb3nMTCUztt
 xMJ0mBr0BqttVs+ouKiVPm3cejQXedEhttwWcloIXS8lNenlpL29Zlrx2NHdU8UU
 13+souocj0dwIyTYYS/4Lm9KpuCYnpDBpP5ShvQjVaMe/GxJo6GyZu70c7FgwGNz
 Nh9onzZV3mz1elhfizlV38aVA7KWVXtLWIqOFIKlT2fa4nWB8Hc07miR5UeOK0/h
 r+icnF2qCQe83MbjOxYNxIKB6uiA/4xwVc90X4AQ7F0RX8XPWHIQWG5tlkC9jrCQ
 3RGzYshWDc9Ud2mXtLMyVQxHVVYlFAe1WtdP8ZWb1oxDInmhrarnWeNyECz9xGKu
 VzIDZzeq9G8slJXATWGRfPsYr+Ihpzcen4QQw58cakUBcqEJrYEhlEOfLovM71k3
 /S/jSHBAbQqiw4LPMw87bA5A6+ZKcVSsNE0XCtNnhmqFpLc1kKRrl5vaN+QMk5tJ
 v4/zR0fPqH7SGAJWYs4brdfahyejEo0TwgpDs7KHmu1W9zQ0LCVTaYnQuUmQjta6
 WyYwIy3TTibdfR191O0E3NOW82Q/k/NBD6ySvabN9HqQ9eSk6+rzrWAslXCbYohb
 BJfzcQfDdx+lsyhjeTx9
 =wOP3
 -----END PGP SIGNATURE-----

Merge tag 'nfs-for-3.19-1' of git://git.linux-nfs.org/projects/trondmy/linux-nfs

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

  Features:
   - NFSv4.2 client support for hole punching and preallocation.
   - Further RPC/RDMA client improvements.
   - Add more RPC transport debugging tracepoints.
   - Add RPC debugging tools in debugfs.

  Bugfixes:
   - Stable fix for layoutget error handling
   - Fix a change in COMMIT behaviour resulting from the recent io code
     updates"

* tag 'nfs-for-3.19-1' of git://git.linux-nfs.org/projects/trondmy/linux-nfs: (28 commits)
  sunrpc: add a debugfs rpc_xprt directory with an info file in it
  sunrpc: add debugfs file for displaying client rpc_task queue
  nfs: Add DEALLOCATE support
  nfs: Add ALLOCATE support
  NFS: Clean up nfs4_init_callback()
  NFS: SETCLIENTID XDR buffer sizes are incorrect
  SUNRPC: serialize iostats updates
  xprtrdma: Display async errors
  xprtrdma: Enable pad optimization
  xprtrdma: Re-write rpcrdma_flush_cqs()
  xprtrdma: Refactor tasklet scheduling
  xprtrdma: unmap all FMRs during transport disconnect
  xprtrdma: Cap req_cqinit
  xprtrdma: Return an errno from rpcrdma_register_external()
  nfs: define nfs_inc_fscache_stats and using it as possible
  nfs: replace nfs_add_stats with nfs_inc_stats when add one
  NFS: Deletion of unnecessary checks before the function call "nfs_put_client"
  sunrpc: eliminate RPC_TRACEPOINTS
  sunrpc: eliminate RPC_DEBUG
  lockd: eliminate LOCKD_DEBUG
  ...

fs/lockd/svclock.c

@@ -53,7 +53,7 @@ static const struct rpc_call_ops nlmsvc_grant_ops;
 static LIST_HEAD(nlm_blocked);
 static DEFINE_SPINLOCK(nlm_blocked_lock);
 
-#ifdef LOCKD_DEBUG
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 static const char *nlmdbg_cookie2a(const struct nlm_cookie *cookie)
 {
 	/*

fs/nfs/blocklayout/blocklayout.c

@@ -812,7 +812,7 @@ static u64 pnfs_num_cont_bytes(struct inode *inode, pgoff_t idx)
 
 	/* Optimize common case that writes from 0 to end of file */
 	end = DIV_ROUND_UP(i_size_read(inode), PAGE_CACHE_SIZE);
-	if (end != NFS_I(inode)->npages) {
+	if (end != inode->i_mapping->nrpages) {
 		rcu_read_lock();
 		end = page_cache_next_hole(mapping, idx + 1, ULONG_MAX);
 		rcu_read_unlock();

fs/nfs/callback_proc.c

@@ -49,7 +49,7 @@ __be32 nfs4_callback_getattr(struct cb_getattrargs *args,
 		goto out_iput;
 	res->size = i_size_read(inode);
 	res->change_attr = delegation->change_attr;
-	if (nfsi->npages != 0)
+	if (nfsi->nrequests != 0)
 		res->change_attr++;
 	res->ctime = inode->i_ctime;
 	res->mtime = inode->i_mtime;

fs/nfs/filelayout/filelayoutdev.c

@@ -204,8 +204,7 @@ destroy_ds(struct nfs4_pnfs_ds *ds)
 	ifdebug(FACILITY)
 		print_ds(ds);
 
-	if (ds->ds_clp)
+	nfs_put_client(ds->ds_clp);
-		nfs_put_client(ds->ds_clp);
 
 	while (!list_empty(&ds->ds_addrs)) {
 		da = list_first_entry(&ds->ds_addrs,

fs/nfs/fscache.c

@@ -269,8 +269,8 @@ int nfs_fscache_release_page(struct page *page, gfp_t gfp)
 		if (!fscache_maybe_release_page(cookie, page, gfp))
 			return 0;
 
-		nfs_add_fscache_stats(page->mapping->host,
+		nfs_inc_fscache_stats(page->mapping->host,
-				      NFSIOS_FSCACHE_PAGES_UNCACHED, 1);
+				      NFSIOS_FSCACHE_PAGES_UNCACHED);
 	}
 
 	return 1;
@@ -293,8 +293,8 @@ void __nfs_fscache_invalidate_page(struct page *page, struct inode *inode)
 
 	BUG_ON(!PageLocked(page));
 	fscache_uncache_page(cookie, page);
-	nfs_add_fscache_stats(page->mapping->host,
+	nfs_inc_fscache_stats(page->mapping->host,
-			      NFSIOS_FSCACHE_PAGES_UNCACHED, 1);
+			      NFSIOS_FSCACHE_PAGES_UNCACHED);
 }
 
 /*
@@ -343,19 +343,19 @@ int __nfs_readpage_from_fscache(struct nfs_open_context *ctx,
 	case 0: /* read BIO submitted (page in fscache) */
 		dfprintk(FSCACHE,
 			 "NFS:    readpage_from_fscache: BIO submitted\n");
-		nfs_add_fscache_stats(inode, NFSIOS_FSCACHE_PAGES_READ_OK, 1);
+		nfs_inc_fscache_stats(inode, NFSIOS_FSCACHE_PAGES_READ_OK);
 		return ret;
 
 	case -ENOBUFS: /* inode not in cache */
 	case -ENODATA: /* page not in cache */
-		nfs_add_fscache_stats(inode, NFSIOS_FSCACHE_PAGES_READ_FAIL, 1);
+		nfs_inc_fscache_stats(inode, NFSIOS_FSCACHE_PAGES_READ_FAIL);
 		dfprintk(FSCACHE,
 			 "NFS:    readpage_from_fscache %d\n", ret);
 		return 1;
 
 	default:
 		dfprintk(FSCACHE, "NFS:    readpage_from_fscache %d\n", ret);
-		nfs_add_fscache_stats(inode, NFSIOS_FSCACHE_PAGES_READ_FAIL, 1);
+		nfs_inc_fscache_stats(inode, NFSIOS_FSCACHE_PAGES_READ_FAIL);
 	}
 	return ret;
 }
@@ -429,11 +429,11 @@ void __nfs_readpage_to_fscache(struct inode *inode, struct page *page, int sync)
 
 	if (ret != 0) {
 		fscache_uncache_page(nfs_i_fscache(inode), page);
-		nfs_add_fscache_stats(inode,
+		nfs_inc_fscache_stats(inode,
-				      NFSIOS_FSCACHE_PAGES_WRITTEN_FAIL, 1);
+				      NFSIOS_FSCACHE_PAGES_WRITTEN_FAIL);
-		nfs_add_fscache_stats(inode, NFSIOS_FSCACHE_PAGES_UNCACHED, 1);
+		nfs_inc_fscache_stats(inode, NFSIOS_FSCACHE_PAGES_UNCACHED);
 	} else {
-		nfs_add_fscache_stats(inode,
+		nfs_inc_fscache_stats(inode,
-				      NFSIOS_FSCACHE_PAGES_WRITTEN_OK, 1);
+				      NFSIOS_FSCACHE_PAGES_WRITTEN_OK);
 	}
 }

fs/nfs/inode.c

@@ -192,6 +192,7 @@ void nfs_zap_caches(struct inode *inode)
 	nfs_zap_caches_locked(inode);
 	spin_unlock(&inode->i_lock);
 }
+EXPORT_SYMBOL_GPL(nfs_zap_caches);
 
 void nfs_zap_mapping(struct inode *inode, struct address_space *mapping)
 {
@@ -1149,7 +1150,7 @@ static unsigned long nfs_wcc_update_inode(struct inode *inode, struct nfs_fattr
 	if ((fattr->valid & NFS_ATTR_FATTR_PRESIZE)
 			&& (fattr->valid & NFS_ATTR_FATTR_SIZE)
 			&& i_size_read(inode) == nfs_size_to_loff_t(fattr->pre_size)
-			&& nfsi->npages == 0) {
+			&& nfsi->nrequests == 0) {
 		i_size_write(inode, nfs_size_to_loff_t(fattr->size));
 		ret |= NFS_INO_INVALID_ATTR;
 	}
@@ -1192,7 +1193,7 @@ static int nfs_check_inode_attributes(struct inode *inode, struct nfs_fattr *fat
 	if (fattr->valid & NFS_ATTR_FATTR_SIZE) {
 		cur_size = i_size_read(inode);
 		new_isize = nfs_size_to_loff_t(fattr->size);
-		if (cur_size != new_isize && nfsi->npages == 0)
+		if (cur_size != new_isize && nfsi->nrequests == 0)
 			invalid |= NFS_INO_INVALID_ATTR|NFS_INO_REVAL_PAGECACHE;
 	}
 
@@ -1619,7 +1620,7 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr)
 		if (new_isize != cur_isize) {
 			/* Do we perhaps have any outstanding writes, or has
 			 * the file grown beyond our last write? */
-			if ((nfsi->npages == 0) || new_isize > cur_isize) {
+			if ((nfsi->nrequests == 0) || new_isize > cur_isize) {
 				i_size_write(inode, new_isize);
 				invalid |= NFS_INO_INVALID_ATTR|NFS_INO_INVALID_DATA;
 				invalid &= ~NFS_INO_REVAL_PAGECACHE;
@@ -1784,7 +1785,7 @@ static void init_once(void *foo)
 	INIT_LIST_HEAD(&nfsi->access_cache_entry_lru);
 	INIT_LIST_HEAD(&nfsi->access_cache_inode_lru);
 	INIT_LIST_HEAD(&nfsi->commit_info.list);
-	nfsi->npages = 0;
+	nfsi->nrequests = 0;
 	nfsi->commit_info.ncommit = 0;
 	atomic_set(&nfsi->commit_info.rpcs_out, 0);
 	atomic_set(&nfsi->silly_count, 1);

fs/nfs/iostat.h

@@ -55,6 +55,11 @@ static inline void nfs_add_fscache_stats(struct inode *inode,
 {
 	this_cpu_add(NFS_SERVER(inode)->io_stats->fscache[stat], addend);
 }
+static inline void nfs_inc_fscache_stats(struct inode *inode,
+					 enum nfs_stat_fscachecounters stat)
+{
+	this_cpu_inc(NFS_SERVER(inode)->io_stats->fscache[stat]);
+}
 #endif
 
 static inline struct nfs_iostats __percpu *nfs_alloc_iostats(void)

fs/nfs/nfs42.h

@@ -6,6 +6,8 @@
 #define __LINUX_FS_NFS_NFS4_2_H
 
 /* nfs4.2proc.c */
+int nfs42_proc_allocate(struct file *, loff_t, loff_t);
+int nfs42_proc_deallocate(struct file *, loff_t, loff_t);
 loff_t nfs42_proc_llseek(struct file *, loff_t, int);
 
 /* nfs4.2xdr.h */

fs/nfs/nfs42proc.c

@@ -32,6 +32,81 @@ static int nfs42_set_rw_stateid(nfs4_stateid *dst, struct file *file,
 	return ret;
 }
 
+static int _nfs42_proc_fallocate(struct rpc_message *msg, struct file *filep,
+				 loff_t offset, loff_t len)
+{
+	struct inode *inode = file_inode(filep);
+	struct nfs42_falloc_args args = {
+		.falloc_fh	= NFS_FH(inode),
+		.falloc_offset	= offset,
+		.falloc_length	= len,
+	};
+	struct nfs42_falloc_res res;
+	struct nfs_server *server = NFS_SERVER(inode);
+	int status;
+
+	msg->rpc_argp = &args;
+	msg->rpc_resp = &res;
+
+	status = nfs42_set_rw_stateid(&args.falloc_stateid, filep, FMODE_WRITE);
+	if (status)
+		return status;
+
+	return nfs4_call_sync(server->client, server, msg,
+			      &args.seq_args, &res.seq_res, 0);
+}
+
+static int nfs42_proc_fallocate(struct rpc_message *msg, struct file *filep,
+				loff_t offset, loff_t len)
+{
+	struct nfs_server *server = NFS_SERVER(file_inode(filep));
+	struct nfs4_exception exception = { };
+	int err;
+
+	do {
+		err = _nfs42_proc_fallocate(msg, filep, offset, len);
+		if (err == -ENOTSUPP)
+			return -EOPNOTSUPP;
+		err = nfs4_handle_exception(server, err, &exception);
+	} while (exception.retry);
+
+	return err;
+}
+
+int nfs42_proc_allocate(struct file *filep, loff_t offset, loff_t len)
+{
+	struct rpc_message msg = {
+		.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_ALLOCATE],
+	};
+	struct inode *inode = file_inode(filep);
+	int err;
+
+	if (!nfs_server_capable(inode, NFS_CAP_ALLOCATE))
+		return -EOPNOTSUPP;
+
+	err = nfs42_proc_fallocate(&msg, filep, offset, len);
+	if (err == -EOPNOTSUPP)
+		NFS_SERVER(inode)->caps &= ~NFS_CAP_ALLOCATE;
+	return err;
+}
+
+int nfs42_proc_deallocate(struct file *filep, loff_t offset, loff_t len)
+{
+	struct rpc_message msg = {
+		.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_DEALLOCATE],
+	};
+	struct inode *inode = file_inode(filep);
+	int err;
+
+	if (!nfs_server_capable(inode, NFS_CAP_DEALLOCATE))
+		return -EOPNOTSUPP;
+
+	err = nfs42_proc_fallocate(&msg, filep, offset, len);
+	if (err == -EOPNOTSUPP)
+		NFS_SERVER(inode)->caps &= ~NFS_CAP_DEALLOCATE;
+	return err;
+}
+
 loff_t nfs42_proc_llseek(struct file *filep, loff_t offset, int whence)
 {
 	struct inode *inode = file_inode(filep);
@@ -50,7 +125,7 @@ loff_t nfs42_proc_llseek(struct file *filep, loff_t offset, int whence)
 	struct nfs_server *server = NFS_SERVER(inode);
 	int status;
 
-	if (!(server->caps & NFS_CAP_SEEK))
+	if (!nfs_server_capable(inode, NFS_CAP_SEEK))
 		return -ENOTSUPP;
 
 	status = nfs42_set_rw_stateid(&args.sa_stateid, filep, FMODE_READ);

fs/nfs/nfs42xdr.c

@@ -4,6 +4,15 @@
 #ifndef __LINUX_FS_NFS_NFS4_2XDR_H
 #define __LINUX_FS_NFS_NFS4_2XDR_H
 
+#define encode_fallocate_maxsz		(encode_stateid_maxsz + \
+					 2 /* offset */ + \
+					 2 /* length */)
+#define encode_allocate_maxsz		(op_encode_hdr_maxsz + \
+					 encode_fallocate_maxsz)
+#define decode_allocate_maxsz		(op_decode_hdr_maxsz)
+#define encode_deallocate_maxsz		(op_encode_hdr_maxsz + \
+					 encode_fallocate_maxsz)
+#define decode_deallocate_maxsz		(op_decode_hdr_maxsz)
 #define encode_seek_maxsz		(op_encode_hdr_maxsz + \
 					 encode_stateid_maxsz + \
 					 2 /* offset */ + \
@@ -14,6 +23,18 @@
 					 2 /* offset */ + \
 					 2 /* length */)
 
+#define NFS4_enc_allocate_sz		(compound_encode_hdr_maxsz + \
+					 encode_putfh_maxsz + \
+					 encode_allocate_maxsz)
+#define NFS4_dec_allocate_sz		(compound_decode_hdr_maxsz + \
+					 decode_putfh_maxsz + \
+					 decode_allocate_maxsz)
+#define NFS4_enc_deallocate_sz		(compound_encode_hdr_maxsz + \
+					 encode_putfh_maxsz + \
+					 encode_deallocate_maxsz)
+#define NFS4_dec_deallocate_sz		(compound_decode_hdr_maxsz + \
+					 decode_putfh_maxsz + \
+					 decode_deallocate_maxsz)
 #define NFS4_enc_seek_sz		(compound_encode_hdr_maxsz + \
 					 encode_putfh_maxsz + \
 					 encode_seek_maxsz)
@@ -22,6 +43,30 @@
 					 decode_seek_maxsz)
 
 
+static void encode_fallocate(struct xdr_stream *xdr,
+			     struct nfs42_falloc_args *args)
+{
+	encode_nfs4_stateid(xdr, &args->falloc_stateid);
+	encode_uint64(xdr, args->falloc_offset);
+	encode_uint64(xdr, args->falloc_length);
+}
+
+static void encode_allocate(struct xdr_stream *xdr,
+			    struct nfs42_falloc_args *args,
+			    struct compound_hdr *hdr)
+{
+	encode_op_hdr(xdr, OP_ALLOCATE, decode_allocate_maxsz, hdr);
+	encode_fallocate(xdr, args);
+}
+
+static void encode_deallocate(struct xdr_stream *xdr,
+			      struct nfs42_falloc_args *args,
+			      struct compound_hdr *hdr)
+{
+	encode_op_hdr(xdr, OP_DEALLOCATE, decode_deallocate_maxsz, hdr);
+	encode_fallocate(xdr, args);
+}
+
 static void encode_seek(struct xdr_stream *xdr,
 			struct nfs42_seek_args *args,
 			struct compound_hdr *hdr)
@@ -32,6 +77,42 @@ static void encode_seek(struct xdr_stream *xdr,
 	encode_uint32(xdr, args->sa_what);
 }
 
+/*
+ * Encode ALLOCATE request
+ */
+static void nfs4_xdr_enc_allocate(struct rpc_rqst *req,
+				  struct xdr_stream *xdr,
+				  struct nfs42_falloc_args *args)
+{
+	struct compound_hdr hdr = {
+		.minorversion = nfs4_xdr_minorversion(&args->seq_args),
+	};
+
+	encode_compound_hdr(xdr, req, &hdr);
+	encode_sequence(xdr, &args->seq_args, &hdr);
+	encode_putfh(xdr, args->falloc_fh, &hdr);
+	encode_allocate(xdr, args, &hdr);
+	encode_nops(&hdr);
+}
+
+/*
+ * Encode DEALLOCATE request
+ */
+static void nfs4_xdr_enc_deallocate(struct rpc_rqst *req,
+				    struct xdr_stream *xdr,
+				    struct nfs42_falloc_args *args)
+{
+	struct compound_hdr hdr = {
+		.minorversion = nfs4_xdr_minorversion(&args->seq_args),
+	};
+
+	encode_compound_hdr(xdr, req, &hdr);
+	encode_sequence(xdr, &args->seq_args, &hdr);
+	encode_putfh(xdr, args->falloc_fh, &hdr);
+	encode_deallocate(xdr, args, &hdr);
+	encode_nops(&hdr);
+}
+
 /*
  * Encode SEEK request
  */
@@ -50,6 +131,16 @@ static void nfs4_xdr_enc_seek(struct rpc_rqst *req,
 	encode_nops(&hdr);
 }
 
+static int decode_allocate(struct xdr_stream *xdr, struct nfs42_falloc_res *res)
+{
+	return decode_op_hdr(xdr, OP_ALLOCATE);
+}
+
+static int decode_deallocate(struct xdr_stream *xdr, struct nfs42_falloc_res *res)
+{
+	return decode_op_hdr(xdr, OP_DEALLOCATE);
+}
+
 static int decode_seek(struct xdr_stream *xdr, struct nfs42_seek_res *res)
 {
 	int status;
@@ -72,6 +163,54 @@ out_overflow:
 	return -EIO;
 }
 
+/*
+ * Decode ALLOCATE request
+ */
+static int nfs4_xdr_dec_allocate(struct rpc_rqst *rqstp,
+				 struct xdr_stream *xdr,
+				 struct nfs42_falloc_res *res)
+{
+	struct compound_hdr hdr;
+	int status;
+
+	status = decode_compound_hdr(xdr, &hdr);
+	if (status)
+		goto out;
+	status = decode_sequence(xdr, &res->seq_res, rqstp);
+	if (status)
+		goto out;
+	status = decode_putfh(xdr);
+	if (status)
+		goto out;
+	status = decode_allocate(xdr, res);
+out:
+	return status;
+}
+
+/*
+ * Decode DEALLOCATE request
+ */
+static int nfs4_xdr_dec_deallocate(struct rpc_rqst *rqstp,
+				   struct xdr_stream *xdr,
+				   struct nfs42_falloc_res *res)
+{
+	struct compound_hdr hdr;
+	int status;
+
+	status = decode_compound_hdr(xdr, &hdr);
+	if (status)
+		goto out;
+	status = decode_sequence(xdr, &res->seq_res, rqstp);
+	if (status)
+		goto out;
+	status = decode_putfh(xdr);
+	if (status)
+		goto out;
+	status = decode_deallocate(xdr, res);
+out:
+	return status;
+}
+
 /*
  * Decode SEEK request
  */

fs/nfs/nfs4_fs.h

@@ -226,6 +226,7 @@ int nfs4_replace_transport(struct nfs_server *server,
 				const struct nfs4_fs_locations *locations);
 
 /* nfs4proc.c */
+extern int nfs4_handle_exception(struct nfs_server *, int, struct nfs4_exception *);
 extern int nfs4_call_sync(struct rpc_clnt *, struct nfs_server *,
 			  struct rpc_message *, struct nfs4_sequence_args *,
 			  struct nfs4_sequence_res *, int);

fs/nfs/nfs4client.c

@@ -241,28 +241,25 @@ void nfs4_free_client(struct nfs_client *clp)
  */
 static int nfs4_init_callback(struct nfs_client *clp)
 {
+	struct rpc_xprt *xprt;
 	int error;
 
-	if (clp->rpc_ops->version == 4) {
+	xprt = rcu_dereference_raw(clp->cl_rpcclient->cl_xprt);
-		struct rpc_xprt *xprt;
 
-		xprt = rcu_dereference_raw(clp->cl_rpcclient->cl_xprt);
+	if (nfs4_has_session(clp)) {
-
+		error = xprt_setup_backchannel(xprt, NFS41_BC_MIN_CALLBACKS);
-		if (nfs4_has_session(clp)) {
+		if (error < 0)
-			error = xprt_setup_backchannel(xprt,
-						NFS41_BC_MIN_CALLBACKS);
-			if (error < 0)
-				return error;
-		}
-
-		error = nfs_callback_up(clp->cl_mvops->minor_version, xprt);
-		if (error < 0) {
-			dprintk("%s: failed to start callback. Error = %d\n",
-				__func__, error);
 			return error;
-		}
-		__set_bit(NFS_CS_CALLBACK, &clp->cl_res_state);
 	}
+
+	error = nfs_callback_up(clp->cl_mvops->minor_version, xprt);
+	if (error < 0) {
+		dprintk("%s: failed to start callback. Error = %d\n",
+			__func__, error);
+		return error;
+	}
+	__set_bit(NFS_CS_CALLBACK, &clp->cl_res_state);
+
 	return 0;
 }
 
@@ -498,8 +495,7 @@ int nfs40_walk_client_list(struct nfs_client *new,
 			atomic_inc(&pos->cl_count);
 			spin_unlock(&nn->nfs_client_lock);
 
-			if (prev)
+			nfs_put_client(prev);
-				nfs_put_client(prev);
 			prev = pos;
 
 			status = nfs_wait_client_init_complete(pos);
@@ -517,8 +513,7 @@ int nfs40_walk_client_list(struct nfs_client *new,
 		atomic_inc(&pos->cl_count);
 		spin_unlock(&nn->nfs_client_lock);
 
-		if (prev)
+		nfs_put_client(prev);
-			nfs_put_client(prev);
 		prev = pos;
 
 		status = nfs4_proc_setclientid_confirm(pos, &clid, cred);
@@ -549,8 +544,7 @@ int nfs40_walk_client_list(struct nfs_client *new,
 
 	/* No match found. The server lost our clientid */
 out:
-	if (prev)
+	nfs_put_client(prev);
-		nfs_put_client(prev);
 	dprintk("NFS: <-- %s status = %d\n", __func__, status);
 	return status;
 }
@@ -641,8 +635,7 @@ int nfs41_walk_client_list(struct nfs_client *new,
 			atomic_inc(&pos->cl_count);
 			spin_unlock(&nn->nfs_client_lock);
 
-			if (prev)
+			nfs_put_client(prev);
-				nfs_put_client(prev);
 			prev = pos;
 
 			status = nfs_wait_client_init_complete(pos);
@@ -675,8 +668,7 @@ int nfs41_walk_client_list(struct nfs_client *new,
 	/* No matching nfs_client found. */
 	spin_unlock(&nn->nfs_client_lock);
 	dprintk("NFS: <-- %s status = %d\n", __func__, status);
-	if (prev)
+	nfs_put_client(prev);
-		nfs_put_client(prev);
 	return status;
 }
 #endif	/* CONFIG_NFS_V4_1 */

fs/nfs/nfs4file.c

@@ -3,6 +3,8 @@
  *
  *  Copyright (C) 1992  Rick Sladkey
  */
+#include <linux/fs.h>
+#include <linux/falloc.h>
 #include <linux/nfs_fs.h>
 #include "internal.h"
 #include "fscache.h"
@@ -134,6 +136,32 @@ static loff_t nfs4_file_llseek(struct file *filep, loff_t offset, int whence)
 		return nfs_file_llseek(filep, offset, whence);
 	}
 }
+
+static long nfs42_fallocate(struct file *filep, int mode, loff_t offset, loff_t len)
+{
+	struct inode *inode = file_inode(filep);
+	long ret;
+
+	if (!S_ISREG(inode->i_mode))
+		return -EOPNOTSUPP;
+
+	if ((mode != 0) && (mode != (FALLOC_FL_PUNCH_HOLE | FALLOC_FL_KEEP_SIZE)))
+		return -EOPNOTSUPP;
+
+	ret = inode_newsize_ok(inode, offset + len);
+	if (ret < 0)
+		return ret;
+
+	mutex_lock(&inode->i_mutex);
+	if (mode & FALLOC_FL_PUNCH_HOLE)
+		ret = nfs42_proc_deallocate(filep, offset, len);
+	else
+		ret = nfs42_proc_allocate(filep, offset, len);
+	mutex_unlock(&inode->i_mutex);
+
+	nfs_zap_caches(inode);
+	return ret;
+}
 #endif /* CONFIG_NFS_V4_2 */
 
 const struct file_operations nfs4_file_operations = {
@@ -155,6 +183,9 @@ const struct file_operations nfs4_file_operations = {
 	.flock		= nfs_flock,
 	.splice_read	= nfs_file_splice_read,
 	.splice_write	= iter_file_splice_write,
+#ifdef CONFIG_NFS_V4_2
+	.fallocate	= nfs42_fallocate,
+#endif /* CONFIG_NFS_V4_2 */
 	.check_flags	= nfs_check_flags,
 	.setlease	= simple_nosetlease,
 };

fs/nfs/nfs4proc.c

@@ -158,8 +158,6 @@ static int nfs4_map_errors(int err)
 		return -EACCES;
 	case -NFS4ERR_MINOR_VERS_MISMATCH:
 		return -EPROTONOSUPPORT;
-	case -NFS4ERR_ACCESS:
-		return -EACCES;
 	case -NFS4ERR_FILE_OPEN:
 		return -EBUSY;
 	default:
@@ -344,7 +342,7 @@ static int nfs4_delay(struct rpc_clnt *clnt, long *timeout)
 /* This is the error handling routine for processes that are allowed
  * to sleep.
  */
-static int nfs4_handle_exception(struct nfs_server *server, int errorcode, struct nfs4_exception *exception)
+int nfs4_handle_exception(struct nfs_server *server, int errorcode, struct nfs4_exception *exception)
 {
 	struct nfs_client *clp = server->nfs_client;
 	struct nfs4_state *state = exception->state;
@@ -7704,6 +7702,9 @@ nfs4_proc_layoutget(struct nfs4_layoutget *lgp, gfp_t gfp_flags)
 
 	dprintk("--> %s\n", __func__);
 
+	/* nfs4_layoutget_release calls pnfs_put_layout_hdr */
+	pnfs_get_layout_hdr(NFS_I(inode)->layout);
+
 	lgp->args.layout.pages = nfs4_alloc_pages(max_pages, gfp_flags);
 	if (!lgp->args.layout.pages) {
 		nfs4_layoutget_release(lgp);
@@ -7716,9 +7717,6 @@ nfs4_proc_layoutget(struct nfs4_layoutget *lgp, gfp_t gfp_flags)
 	lgp->res.seq_res.sr_slot = NULL;
 	nfs4_init_sequence(&lgp->args.seq_args, &lgp->res.seq_res, 0);
 
-	/* nfs4_layoutget_release calls pnfs_put_layout_hdr */
-	pnfs_get_layout_hdr(NFS_I(inode)->layout);
-
 	task = rpc_run_task(&task_setup_data);
 	if (IS_ERR(task))
 		return ERR_CAST(task);
@@ -8426,6 +8424,8 @@ static const struct nfs4_minor_version_ops nfs_v4_2_minor_ops = {
 		| NFS_CAP_POSIX_LOCK
 		| NFS_CAP_STATEID_NFSV41
 		| NFS_CAP_ATOMIC_OPEN_V1
+		| NFS_CAP_ALLOCATE
+		| NFS_CAP_DEALLOCATE
 		| NFS_CAP_SEEK,
 	.init_client = nfs41_init_client,
 	.shutdown_client = nfs41_shutdown_client,

fs/nfs/nfs4xdr.c

@@ -141,13 +141,15 @@ static int nfs4_stat_to_errno(int);
 				XDR_QUADLEN(NFS4_VERIFIER_SIZE) + \
 				XDR_QUADLEN(NFS4_SETCLIENTID_NAMELEN) + \
 				1 /* sc_prog */ + \
-				XDR_QUADLEN(RPCBIND_MAXNETIDLEN) + \
+				1 + XDR_QUADLEN(RPCBIND_MAXNETIDLEN) + \
-				XDR_QUADLEN(RPCBIND_MAXUADDRLEN) + \
+				1 + XDR_QUADLEN(RPCBIND_MAXUADDRLEN) + \
 				1) /* sc_cb_ident */
 #define decode_setclientid_maxsz \
 				(op_decode_hdr_maxsz + \
-				2 + \
+				2 /* clientid */ + \
-				1024) /* large value for CLID_INUSE */
+				XDR_QUADLEN(NFS4_VERIFIER_SIZE) + \
+				1 + XDR_QUADLEN(RPCBIND_MAXNETIDLEN) + \
+				1 + XDR_QUADLEN(RPCBIND_MAXUADDRLEN))
 #define encode_setclientid_confirm_maxsz \
 				(op_encode_hdr_maxsz + \
 				3 + (NFS4_VERIFIER_SIZE >> 2))
@@ -7394,6 +7396,8 @@ struct rpc_procinfo	nfs4_procedures[] = {
 #endif /* CONFIG_NFS_V4_1 */
 #ifdef CONFIG_NFS_V4_2
 	PROC(SEEK,		enc_seek,		dec_seek),
+	PROC(ALLOCATE,		enc_allocate,		dec_allocate),
+	PROC(DEALLOCATE,	enc_deallocate,		dec_deallocate),
 #endif /* CONFIG_NFS_V4_2 */
 };
 

fs/nfs/pagelist.c

@@ -258,6 +258,7 @@ bool nfs_page_group_sync_on_bit(struct nfs_page *req, unsigned int bit)
 static inline void
 nfs_page_group_init(struct nfs_page *req, struct nfs_page *prev)
 {
+	struct inode *inode;
 	WARN_ON_ONCE(prev == req);
 
 	if (!prev) {
@@ -276,12 +277,16 @@ nfs_page_group_init(struct nfs_page *req, struct nfs_page *prev)
 		 * nfs_page_group_destroy is called */
 		kref_get(&req->wb_head->wb_kref);
 
-		/* grab extra ref if head request has extra ref from
+		/* grab extra ref and bump the request count if head request
-		 * the write/commit path to handle handoff between write
+		 * has extra ref from the write/commit path to handle handoff
-		 * and commit lists */
+		 * between write and commit lists. */
 		if (test_bit(PG_INODE_REF, &prev->wb_head->wb_flags)) {
+			inode = page_file_mapping(req->wb_page)->host;
 			set_bit(PG_INODE_REF, &req->wb_flags);
 			kref_get(&req->wb_kref);
+			spin_lock(&inode->i_lock);
+			NFS_I(inode)->nrequests++;
+			spin_unlock(&inode->i_lock);
 		}
 	}
 }

fs/nfs/read.c

@@ -269,7 +269,7 @@ int nfs_readpage(struct file *file, struct page *page)
 	dprintk("NFS: nfs_readpage (%p %ld@%lu)\n",
 		page, PAGE_CACHE_SIZE, page_file_index(page));
 	nfs_inc_stats(inode, NFSIOS_VFSREADPAGE);
-	nfs_add_stats(inode, NFSIOS_READPAGES, 1);
+	nfs_inc_stats(inode, NFSIOS_READPAGES);
 
 	/*
 	 * Try to flush any pending writes to the file..

fs/nfs/write.c

@@ -575,7 +575,7 @@ static int nfs_do_writepage(struct page *page, struct writeback_control *wbc, st
 	int ret;
 
 	nfs_inc_stats(inode, NFSIOS_VFSWRITEPAGE);
-	nfs_add_stats(inode, NFSIOS_WRITEPAGES, 1);
+	nfs_inc_stats(inode, NFSIOS_WRITEPAGES);
 
 	nfs_pageio_cond_complete(pgio, page_file_index(page));
 	ret = nfs_page_async_flush(pgio, page, wbc->sync_mode == WB_SYNC_NONE);
@@ -670,7 +670,8 @@ static void nfs_inode_add_request(struct inode *inode, struct nfs_page *req)
 	nfs_lock_request(req);
 
 	spin_lock(&inode->i_lock);
-	if (!nfsi->npages && NFS_PROTO(inode)->have_delegation(inode, FMODE_WRITE))
+	if (!nfsi->nrequests &&
+	    NFS_PROTO(inode)->have_delegation(inode, FMODE_WRITE))
 		inode->i_version++;
 	/*
 	 * Swap-space should not get truncated. Hence no need to plug the race
@@ -681,9 +682,11 @@ static void nfs_inode_add_request(struct inode *inode, struct nfs_page *req)
 		SetPagePrivate(req->wb_page);
 		set_page_private(req->wb_page, (unsigned long)req);
 	}
-	nfsi->npages++;
+	nfsi->nrequests++;
 	/* this a head request for a page group - mark it as having an
-	 * extra reference so sub groups can follow suit */
+	 * extra reference so sub groups can follow suit.
+	 * This flag also informs pgio layer when to bump nrequests when
+	 * adding subrequests. */
 	WARN_ON(test_and_set_bit(PG_INODE_REF, &req->wb_flags));
 	kref_get(&req->wb_kref);
 	spin_unlock(&inode->i_lock);
@@ -709,7 +712,11 @@ static void nfs_inode_remove_request(struct nfs_page *req)
 			wake_up_page(head->wb_page, PG_private);
 			clear_bit(PG_MAPPED, &head->wb_flags);
 		}
-		nfsi->npages--;
+		nfsi->nrequests--;
+		spin_unlock(&inode->i_lock);
+	} else {
+		spin_lock(&inode->i_lock);
+		nfsi->nrequests--;
 		spin_unlock(&inode->i_lock);
 	}
 
@@ -1735,7 +1742,7 @@ static int nfs_commit_unstable_pages(struct inode *inode, struct writeback_contr
 		/* Don't commit yet if this is a non-blocking flush and there
 		 * are a lot of outstanding writes for this mapping.
 		 */
-		if (nfsi->commit_info.ncommit <= (nfsi->npages >> 1))
+		if (nfsi->commit_info.ncommit <= (nfsi->nrequests >> 1))
 			goto out_mark_dirty;
 
 		/* don't wait for the COMMIT response */

include/linux/lockd/debug.h

@@ -17,12 +17,8 @@
  * Enable lockd debugging.
  * Requires RPC_DEBUG.
  */
-#ifdef RPC_DEBUG
-# define LOCKD_DEBUG		1
-#endif
-
 #undef ifdebug
-#if defined(RPC_DEBUG) && defined(LOCKD_DEBUG)
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 # define ifdebug(flag)		if (unlikely(nlm_debug & NLMDBG_##flag))
 #else
 # define ifdebug(flag)		if (0)

include/linux/nfs4.h

@@ -490,6 +490,8 @@ enum {
 
 	/* nfs42 */
 	NFSPROC4_CLNT_SEEK,
+	NFSPROC4_CLNT_ALLOCATE,
+	NFSPROC4_CLNT_DEALLOCATE,
 };
 
 /* nfs41 types */

include/linux/nfs_fs.h

@@ -163,7 +163,7 @@ struct nfs_inode {
 	 */
 	__be32			cookieverf[2];
 
-	unsigned long		npages;
+	unsigned long		nrequests;
 	struct nfs_mds_commit_info commit_info;
 
 	/* Open contexts for shared mmap writes */
@@ -520,7 +520,7 @@ extern void nfs_commit_free(struct nfs_commit_data *data);
 static inline int
 nfs_have_writebacks(struct inode *inode)
 {
-	return NFS_I(inode)->npages != 0;
+	return NFS_I(inode)->nrequests != 0;
 }
 
 /*

include/linux/nfs_fs_sb.h

@@ -231,5 +231,7 @@ struct nfs_server {
 #define NFS_CAP_ATOMIC_OPEN_V1	(1U << 17)
 #define NFS_CAP_SECURITY_LABEL	(1U << 18)
 #define NFS_CAP_SEEK		(1U << 19)
+#define NFS_CAP_ALLOCATE	(1U << 20)
+#define NFS_CAP_DEALLOCATE	(1U << 21)
 
 #endif

include/linux/nfs_xdr.h

@@ -1243,6 +1243,20 @@ nfs_free_pnfs_ds_cinfo(struct pnfs_ds_commit_info *cinfo)
 #endif /* CONFIG_NFS_V4_1 */
 
 #ifdef CONFIG_NFS_V4_2
+struct nfs42_falloc_args {
+	struct nfs4_sequence_args	seq_args;
+
+	struct nfs_fh			*falloc_fh;
+	nfs4_stateid			 falloc_stateid;
+	u64				 falloc_offset;
+	u64				 falloc_length;
+};
+
+struct nfs42_falloc_res {
+	struct nfs4_sequence_res	seq_res;
+	unsigned int			status;
+};
+
 struct nfs42_seek_args {
 	struct nfs4_sequence_args	seq_args;
 

include/linux/sunrpc/auth.h

@@ -53,7 +53,7 @@ struct rpc_cred {
 	struct rcu_head		cr_rcu;
 	struct rpc_auth *	cr_auth;
 	const struct rpc_credops *cr_ops;
-#ifdef RPC_DEBUG
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 	unsigned long		cr_magic;	/* 0x0f4aa4f0 */
 #endif
 	unsigned long		cr_expire;	/* when to gc */

include/linux/sunrpc/clnt.h

@@ -63,6 +63,9 @@ struct rpc_clnt {
 	struct rpc_rtt		cl_rtt_default;
 	struct rpc_timeout	cl_timeout_default;
 	const struct rpc_program *cl_program;
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
+	struct dentry		*cl_debugfs;	/* debugfs directory */
+#endif
 };
 
 /*
@@ -176,5 +179,6 @@ size_t		rpc_peeraddr(struct rpc_clnt *, struct sockaddr *, size_t);
 const char	*rpc_peeraddr2str(struct rpc_clnt *, enum rpc_display_format_t);
 int		rpc_localaddr(struct rpc_clnt *, struct sockaddr *, size_t);
 
+const char *rpc_proc_name(const struct rpc_task *task);
 #endif /* __KERNEL__ */
 #endif /* _LINUX_SUNRPC_CLNT_H */

include/linux/sunrpc/debug.h

@@ -10,22 +10,10 @@
 
 #include <uapi/linux/sunrpc/debug.h>
 
-
-/*
- * Enable RPC debugging/profiling.
- */
-#ifdef CONFIG_SUNRPC_DEBUG
-#define  RPC_DEBUG
-#endif
-#ifdef CONFIG_TRACEPOINTS
-#define RPC_TRACEPOINTS
-#endif
-/* #define  RPC_PROFILE */
-
 /*
  * Debugging macros etc
  */
-#ifdef RPC_DEBUG
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 extern unsigned int		rpc_debug;
 extern unsigned int		nfs_debug;
 extern unsigned int		nfsd_debug;
@@ -36,7 +24,7 @@ extern unsigned int		nlm_debug;
 #define dprintk_rcu(args...)	dfprintk_rcu(FACILITY, ## args)
 
 #undef ifdebug
-#ifdef RPC_DEBUG			
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 # define ifdebug(fac)		if (unlikely(rpc_debug & RPCDBG_##fac))
 
 # define dfprintk(fac, args...)	\
@@ -65,9 +53,55 @@ extern unsigned int		nlm_debug;
 /*
  * Sysctl interface for RPC debugging
  */
-#ifdef RPC_DEBUG
+
+struct rpc_clnt;
+struct rpc_xprt;
+
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 void		rpc_register_sysctl(void);
 void		rpc_unregister_sysctl(void);
+int		sunrpc_debugfs_init(void);
+void		sunrpc_debugfs_exit(void);
+int		rpc_clnt_debugfs_register(struct rpc_clnt *);
+void		rpc_clnt_debugfs_unregister(struct rpc_clnt *);
+int		rpc_xprt_debugfs_register(struct rpc_xprt *);
+void		rpc_xprt_debugfs_unregister(struct rpc_xprt *);
+#else
+static inline int
+sunrpc_debugfs_init(void)
+{
+	return 0;
+}
+
+static inline void
+sunrpc_debugfs_exit(void)
+{
+	return;
+}
+
+static inline int
+rpc_clnt_debugfs_register(struct rpc_clnt *clnt)
+{
+	return 0;
+}
+
+static inline void
+rpc_clnt_debugfs_unregister(struct rpc_clnt *clnt)
+{
+	return;
+}
+
+static inline int
+rpc_xprt_debugfs_register(struct rpc_xprt *xprt)
+{
+	return 0;
+}
+
+static inline void
+rpc_xprt_debugfs_unregister(struct rpc_xprt *xprt)
+{
+	return;
+}
 #endif
 
 #endif /* _LINUX_SUNRPC_DEBUG_H_ */

include/linux/sunrpc/metrics.h

@@ -27,10 +27,13 @@
 
 #include <linux/seq_file.h>
 #include <linux/ktime.h>
+#include <linux/spinlock.h>
 
 #define RPC_IOSTATS_VERS	"1.0"
 
 struct rpc_iostats {
+	spinlock_t		om_lock;
+
 	/*
 	 * These counters give an idea about how many request
 	 * transmissions are required, on average, to complete that

include/linux/sunrpc/sched.h

@@ -79,7 +79,7 @@ struct rpc_task {
 	unsigned short		tk_flags;	/* misc flags */
 	unsigned short		tk_timeouts;	/* maj timeouts */
 
-#if defined(RPC_DEBUG) || defined(RPC_TRACEPOINTS)
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG) || IS_ENABLED(CONFIG_TRACEPOINTS)
 	unsigned short		tk_pid;		/* debugging aid */
 #endif
 	unsigned char		tk_priority : 2,/* Task priority */
@@ -187,7 +187,7 @@ struct rpc_wait_queue {
 	unsigned char		nr;			/* # tasks remaining for cookie */
 	unsigned short		qlen;			/* total # tasks waiting in queue */
 	struct rpc_timer	timer_list;
-#if defined(RPC_DEBUG) || defined(RPC_TRACEPOINTS)
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG) || IS_ENABLED(CONFIG_TRACEPOINTS)
 	const char *		name;
 #endif
 };
@@ -237,7 +237,7 @@ void		rpc_free(void *);
 int		rpciod_up(void);
 void		rpciod_down(void);
 int		__rpc_wait_for_completion_task(struct rpc_task *task, wait_bit_action_f *);
-#ifdef RPC_DEBUG
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 struct net;
 void		rpc_show_tasks(struct net *);
 #endif
@@ -251,7 +251,7 @@ static inline int rpc_wait_for_completion_task(struct rpc_task *task)
 	return __rpc_wait_for_completion_task(task, NULL);
 }
 
-#if defined(RPC_DEBUG) || defined (RPC_TRACEPOINTS)
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG) || IS_ENABLED(CONFIG_TRACEPOINTS)
 static inline const char * rpc_qname(const struct rpc_wait_queue *q)
 {
 	return ((q && q->name) ? q->name : "unknown");

include/linux/sunrpc/xprt.h

@@ -239,6 +239,9 @@ struct rpc_xprt {
 	struct net		*xprt_net;
 	const char		*servername;
 	const char		*address_strings[RPC_DISPLAY_MAX];
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
+	struct dentry		*debugfs;		/* debugfs directory */
+#endif
 };
 
 #if defined(CONFIG_SUNRPC_BACKCHANNEL)

include/linux/sunrpc/xprtsock.h

@@ -17,6 +17,65 @@ void		cleanup_socket_xprt(void);
 #define RPC_DEF_MIN_RESVPORT	(665U)
 #define RPC_DEF_MAX_RESVPORT	(1023U)
 
+struct sock_xprt {
+	struct rpc_xprt		xprt;
+
+	/*
+	 * Network layer
+	 */
+	struct socket *		sock;
+	struct sock *		inet;
+
+	/*
+	 * State of TCP reply receive
+	 */
+	__be32			tcp_fraghdr,
+				tcp_xid,
+				tcp_calldir;
+
+	u32			tcp_offset,
+				tcp_reclen;
+
+	unsigned long		tcp_copied,
+				tcp_flags;
+
+	/*
+	 * Connection of transports
+	 */
+	struct delayed_work	connect_worker;
+	struct sockaddr_storage	srcaddr;
+	unsigned short		srcport;
+
+	/*
+	 * UDP socket buffer size parameters
+	 */
+	size_t			rcvsize,
+				sndsize;
+
+	/*
+	 * Saved socket callback addresses
+	 */
+	void			(*old_data_ready)(struct sock *);
+	void			(*old_state_change)(struct sock *);
+	void			(*old_write_space)(struct sock *);
+	void			(*old_error_report)(struct sock *);
+};
+
+/*
+ * TCP receive state flags
+ */
+#define TCP_RCV_LAST_FRAG	(1UL << 0)
+#define TCP_RCV_COPY_FRAGHDR	(1UL << 1)
+#define TCP_RCV_COPY_XID	(1UL << 2)
+#define TCP_RCV_COPY_DATA	(1UL << 3)
+#define TCP_RCV_READ_CALLDIR	(1UL << 4)
+#define TCP_RCV_COPY_CALLDIR	(1UL << 5)
+
+/*
+ * TCP RPC flags
+ */
+#define TCP_RPC_REPLY		(1UL << 6)
+
 #endif /* __KERNEL__ */
 
 #endif /* _LINUX_SUNRPC_XPRTSOCK_H */

include/trace/events/sunrpc.h

@@ -6,6 +6,8 @@
 
 #include <linux/sunrpc/sched.h>
 #include <linux/sunrpc/clnt.h>
+#include <linux/sunrpc/svc.h>
+#include <linux/sunrpc/xprtsock.h>
 #include <net/tcp_states.h>
 #include <linux/net.h>
 #include <linux/tracepoint.h>
@@ -306,6 +308,164 @@ DEFINE_RPC_SOCKET_EVENT_DONE(rpc_socket_reset_connection);
 DEFINE_RPC_SOCKET_EVENT(rpc_socket_close);
 DEFINE_RPC_SOCKET_EVENT(rpc_socket_shutdown);
 
+DECLARE_EVENT_CLASS(rpc_xprt_event,
+	TP_PROTO(struct rpc_xprt *xprt, __be32 xid, int status),
+
+	TP_ARGS(xprt, xid, status),
+
+	TP_STRUCT__entry(
+		__field(__be32, xid)
+		__field(int, status)
+		__string(addr, xprt->address_strings[RPC_DISPLAY_ADDR])
+		__string(port, xprt->address_strings[RPC_DISPLAY_PORT])
+	),
+
+	TP_fast_assign(
+		__entry->xid = xid;
+		__entry->status = status;
+		__assign_str(addr, xprt->address_strings[RPC_DISPLAY_ADDR]);
+		__assign_str(port, xprt->address_strings[RPC_DISPLAY_PORT]);
+	),
+
+	TP_printk("peer=[%s]:%s xid=0x%x status=%d", __get_str(addr),
+			__get_str(port), be32_to_cpu(__entry->xid),
+			__entry->status)
+);
+
+DEFINE_EVENT(rpc_xprt_event, xprt_lookup_rqst,
+	TP_PROTO(struct rpc_xprt *xprt, __be32 xid, int status),
+	TP_ARGS(xprt, xid, status));
+
+DEFINE_EVENT(rpc_xprt_event, xprt_transmit,
+	TP_PROTO(struct rpc_xprt *xprt, __be32 xid, int status),
+	TP_ARGS(xprt, xid, status));
+
+DEFINE_EVENT(rpc_xprt_event, xprt_complete_rqst,
+	TP_PROTO(struct rpc_xprt *xprt, __be32 xid, int status),
+	TP_ARGS(xprt, xid, status));
+
+TRACE_EVENT(xs_tcp_data_ready,
+	TP_PROTO(struct rpc_xprt *xprt, int err, unsigned int total),
+
+	TP_ARGS(xprt, err, total),
+
+	TP_STRUCT__entry(
+		__field(int, err)
+		__field(unsigned int, total)
+		__string(addr, xprt ? xprt->address_strings[RPC_DISPLAY_ADDR] :
+				"(null)")
+		__string(port, xprt ? xprt->address_strings[RPC_DISPLAY_PORT] :
+				"(null)")
+	),
+
+	TP_fast_assign(
+		__entry->err = err;
+		__entry->total = total;
+		__assign_str(addr, xprt ?
+			xprt->address_strings[RPC_DISPLAY_ADDR] : "(null)");
+		__assign_str(port, xprt ?
+			xprt->address_strings[RPC_DISPLAY_PORT] : "(null)");
+	),
+
+	TP_printk("peer=[%s]:%s err=%d total=%u", __get_str(addr),
+			__get_str(port), __entry->err, __entry->total)
+);
+
+#define rpc_show_sock_xprt_flags(flags) \
+	__print_flags(flags, "|", \
+		{ TCP_RCV_LAST_FRAG, "TCP_RCV_LAST_FRAG" }, \
+		{ TCP_RCV_COPY_FRAGHDR, "TCP_RCV_COPY_FRAGHDR" }, \
+		{ TCP_RCV_COPY_XID, "TCP_RCV_COPY_XID" }, \
+		{ TCP_RCV_COPY_DATA, "TCP_RCV_COPY_DATA" }, \
+		{ TCP_RCV_READ_CALLDIR, "TCP_RCV_READ_CALLDIR" }, \
+		{ TCP_RCV_COPY_CALLDIR, "TCP_RCV_COPY_CALLDIR" }, \
+		{ TCP_RPC_REPLY, "TCP_RPC_REPLY" })
+
+TRACE_EVENT(xs_tcp_data_recv,
+	TP_PROTO(struct sock_xprt *xs),
+
+	TP_ARGS(xs),
+
+	TP_STRUCT__entry(
+		__string(addr, xs->xprt.address_strings[RPC_DISPLAY_ADDR])
+		__string(port, xs->xprt.address_strings[RPC_DISPLAY_PORT])
+		__field(__be32, xid)
+		__field(unsigned long, flags)
+		__field(unsigned long, copied)
+		__field(unsigned int, reclen)
+		__field(unsigned long, offset)
+	),
+
+	TP_fast_assign(
+		__assign_str(addr, xs->xprt.address_strings[RPC_DISPLAY_ADDR]);
+		__assign_str(port, xs->xprt.address_strings[RPC_DISPLAY_PORT]);
+		__entry->xid = xs->tcp_xid;
+		__entry->flags = xs->tcp_flags;
+		__entry->copied = xs->tcp_copied;
+		__entry->reclen = xs->tcp_reclen;
+		__entry->offset = xs->tcp_offset;
+	),
+
+	TP_printk("peer=[%s]:%s xid=0x%x flags=%s copied=%lu reclen=%u offset=%lu",
+			__get_str(addr), __get_str(port), be32_to_cpu(__entry->xid),
+			rpc_show_sock_xprt_flags(__entry->flags),
+			__entry->copied, __entry->reclen, __entry->offset)
+);
+
+TRACE_EVENT(svc_recv,
+	TP_PROTO(struct svc_rqst *rqst, int status),
+
+	TP_ARGS(rqst, status),
+
+	TP_STRUCT__entry(
+		__field(struct sockaddr *, addr)
+		__field(__be32, xid)
+		__field(int, status)
+	),
+
+	TP_fast_assign(
+		__entry->addr = (struct sockaddr *)&rqst->rq_addr;
+		__entry->xid = status > 0 ? rqst->rq_xid : 0;
+		__entry->status = status;
+	),
+
+	TP_printk("addr=%pIScp xid=0x%x status=%d", __entry->addr,
+			be32_to_cpu(__entry->xid), __entry->status)
+);
+
+DECLARE_EVENT_CLASS(svc_rqst_status,
+
+	TP_PROTO(struct svc_rqst *rqst, int status),
+
+	TP_ARGS(rqst, status),
+
+	TP_STRUCT__entry(
+		__field(struct sockaddr *, addr)
+		__field(__be32, xid)
+		__field(int, dropme)
+		__field(int, status)
+	),
+
+	TP_fast_assign(
+		__entry->addr = (struct sockaddr *)&rqst->rq_addr;
+		__entry->xid = rqst->rq_xid;
+		__entry->dropme = (int)rqst->rq_dropme;
+		__entry->status = status;
+	),
+
+	TP_printk("addr=%pIScp rq_xid=0x%x dropme=%d status=%d",
+		__entry->addr, be32_to_cpu(__entry->xid), __entry->dropme,
+		__entry->status)
+);
+
+DEFINE_EVENT(svc_rqst_status, svc_process,
+	TP_PROTO(struct svc_rqst *rqst, int status),
+	TP_ARGS(rqst, status));
+
+DEFINE_EVENT(svc_rqst_status, svc_send,
+	TP_PROTO(struct svc_rqst *rqst, int status),
+	TP_ARGS(rqst, status));
+
 #endif /* _TRACE_SUNRPC_H */
 
 #include <trace/define_trace.h>

include/uapi/linux/nfsd/debug.h

@@ -15,7 +15,7 @@
  * Enable debugging for nfsd.
  * Requires RPC_DEBUG.
  */
-#ifdef RPC_DEBUG
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 # define NFSD_DEBUG		1
 #endif
 

net/sunrpc/Kconfig

@@ -35,6 +35,7 @@ config RPCSEC_GSS_KRB5
 config SUNRPC_DEBUG
 	bool "RPC: Enable dprintk debugging"
 	depends on SUNRPC && SYSCTL
+	select DEBUG_FS
 	help
 	  This option enables a sysctl-based debugging interface
 	  that is be used by the 'rpcdebug' utility to turn on or off

net/sunrpc/Makefile

@@ -14,6 +14,7 @@ sunrpc-y := clnt.o xprt.o socklib.o xprtsock.o sched.o \
 	    addr.o rpcb_clnt.o timer.o xdr.o \
 	    sunrpc_syms.o cache.o rpc_pipe.o \
 	    svc_xprt.o
+sunrpc-$(CONFIG_SUNRPC_DEBUG) += debugfs.o
 sunrpc-$(CONFIG_SUNRPC_BACKCHANNEL) += backchannel_rqst.o bc_svc.o
 sunrpc-$(CONFIG_PROC_FS) += stats.o
 sunrpc-$(CONFIG_SYSCTL) += sysctl.o

net/sunrpc/auth.c

@@ -16,7 +16,7 @@
 #include <linux/sunrpc/gss_api.h>
 #include <linux/spinlock.h>
 
-#ifdef RPC_DEBUG
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 # define RPCDBG_FACILITY	RPCDBG_AUTH
 #endif
 
@@ -646,7 +646,7 @@ rpcauth_init_cred(struct rpc_cred *cred, const struct auth_cred *acred,
 	cred->cr_auth = auth;
 	cred->cr_ops = ops;
 	cred->cr_expire = jiffies;
-#ifdef RPC_DEBUG
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 	cred->cr_magic = RPCAUTH_CRED_MAGIC;
 #endif
 	cred->cr_uid = acred->uid;

net/sunrpc/auth_generic.c

@@ -14,7 +14,7 @@
 #include <linux/sunrpc/debug.h>
 #include <linux/sunrpc/sched.h>
 
-#ifdef RPC_DEBUG
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 # define RPCDBG_FACILITY	RPCDBG_AUTH
 #endif
 

net/sunrpc/auth_gss/auth_gss.c

@@ -66,7 +66,7 @@ static unsigned int gss_expired_cred_retry_delay = GSS_RETRY_EXPIRED;
 #define GSS_KEY_EXPIRE_TIMEO 240
 static unsigned int gss_key_expire_timeo = GSS_KEY_EXPIRE_TIMEO;
 
-#ifdef RPC_DEBUG
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 # define RPCDBG_FACILITY	RPCDBG_AUTH
 #endif
 

net/sunrpc/auth_gss/gss_generic_token.c

@@ -38,7 +38,7 @@
 #include <linux/sunrpc/gss_asn1.h>
 
 
-#ifdef RPC_DEBUG
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 # define RPCDBG_FACILITY        RPCDBG_AUTH
 #endif
 

net/sunrpc/auth_gss/gss_krb5_crypto.c

@@ -45,7 +45,7 @@
 #include <linux/sunrpc/gss_krb5.h>
 #include <linux/sunrpc/xdr.h>
 
-#ifdef RPC_DEBUG
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 # define RPCDBG_FACILITY        RPCDBG_AUTH
 #endif
 

net/sunrpc/auth_gss/gss_krb5_keys.c

@@ -61,7 +61,7 @@
 #include <linux/sunrpc/xdr.h>
 #include <linux/lcm.h>
 
-#ifdef RPC_DEBUG
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 # define RPCDBG_FACILITY        RPCDBG_AUTH
 #endif
 

net/sunrpc/auth_gss/gss_krb5_mech.c

@@ -45,7 +45,7 @@
 #include <linux/crypto.h>
 #include <linux/sunrpc/gss_krb5_enctypes.h>
 
-#ifdef RPC_DEBUG
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 # define RPCDBG_FACILITY	RPCDBG_AUTH
 #endif
 

net/sunrpc/auth_gss/gss_krb5_seal.c

@@ -64,7 +64,7 @@
 #include <linux/random.h>
 #include <linux/crypto.h>
 
-#ifdef RPC_DEBUG
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 # define RPCDBG_FACILITY        RPCDBG_AUTH
 #endif
 

net/sunrpc/auth_gss/gss_krb5_seqnum.c

@@ -35,7 +35,7 @@
 #include <linux/sunrpc/gss_krb5.h>
 #include <linux/crypto.h>
 
-#ifdef RPC_DEBUG
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 # define RPCDBG_FACILITY        RPCDBG_AUTH
 #endif
 

net/sunrpc/auth_gss/gss_krb5_unseal.c

@@ -62,7 +62,7 @@
 #include <linux/sunrpc/gss_krb5.h>
 #include <linux/crypto.h>
 
-#ifdef RPC_DEBUG
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 # define RPCDBG_FACILITY        RPCDBG_AUTH
 #endif
 

net/sunrpc/auth_gss/gss_krb5_wrap.c

@@ -35,7 +35,7 @@
 #include <linux/pagemap.h>
 #include <linux/crypto.h>
 
-#ifdef RPC_DEBUG
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 # define RPCDBG_FACILITY	RPCDBG_AUTH
 #endif
 

net/sunrpc/auth_gss/gss_mech_switch.c

@@ -46,7 +46,7 @@
 #include <linux/sunrpc/gss_api.h>
 #include <linux/sunrpc/clnt.h>
 
-#ifdef RPC_DEBUG
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 # define RPCDBG_FACILITY        RPCDBG_AUTH
 #endif
 

net/sunrpc/auth_gss/gss_rpc_xdr.h

@@ -25,7 +25,7 @@
 #include <linux/sunrpc/clnt.h>
 #include <linux/sunrpc/xprtsock.h>
 
-#ifdef RPC_DEBUG
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 # define RPCDBG_FACILITY	RPCDBG_AUTH
 #endif
 

net/sunrpc/auth_gss/svcauth_gss.c

@@ -51,7 +51,7 @@
 #include "gss_rpc_upcall.h"
 
 
-#ifdef RPC_DEBUG
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 # define RPCDBG_FACILITY	RPCDBG_AUTH
 #endif
 

net/sunrpc/auth_null.c

@@ -10,7 +10,7 @@
 #include <linux/module.h>
 #include <linux/sunrpc/clnt.h>
 
-#ifdef RPC_DEBUG
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 # define RPCDBG_FACILITY	RPCDBG_AUTH
 #endif
 
@@ -138,7 +138,7 @@ struct rpc_cred null_cred = {
 	.cr_ops		= &null_credops,
 	.cr_count	= ATOMIC_INIT(1),
 	.cr_flags	= 1UL << RPCAUTH_CRED_UPTODATE,
-#ifdef RPC_DEBUG
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 	.cr_magic	= RPCAUTH_CRED_MAGIC,
 #endif
 };

net/sunrpc/auth_unix.c

@@ -25,7 +25,7 @@ struct unx_cred {
 
 #define UNX_WRITESLACK		(21 + (UNX_MAXNODENAME >> 2))
 
-#ifdef RPC_DEBUG
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 # define RPCDBG_FACILITY	RPCDBG_AUTH
 #endif
 

net/sunrpc/backchannel_rqst.c

@@ -27,7 +27,7 @@ SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 #include <linux/export.h>
 #include <linux/sunrpc/bc_xprt.h>
 
-#ifdef RPC_DEBUG
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 #define RPCDBG_FACILITY	RPCDBG_TRANS
 #endif
 

net/sunrpc/clnt.c

@@ -42,7 +42,7 @@
 #include "sunrpc.h"
 #include "netns.h"
 
-#ifdef RPC_DEBUG
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 # define RPCDBG_FACILITY	RPCDBG_CALL
 #endif
 
@@ -305,6 +305,10 @@ static int rpc_client_register(struct rpc_clnt *clnt,
 	struct super_block *pipefs_sb;
 	int err;
 
+	err = rpc_clnt_debugfs_register(clnt);
+	if (err)
+		return err;
+
 	pipefs_sb = rpc_get_sb_net(net);
 	if (pipefs_sb) {
 		err = rpc_setup_pipedir(pipefs_sb, clnt);
@@ -331,6 +335,7 @@ err_auth:
 out:
 	if (pipefs_sb)
 		rpc_put_sb_net(net);
+	rpc_clnt_debugfs_unregister(clnt);
 	return err;
 }
 
@@ -670,6 +675,7 @@ int rpc_switch_client_transport(struct rpc_clnt *clnt,
 
 	rpc_unregister_client(clnt);
 	__rpc_clnt_remove_pipedir(clnt);
+	rpc_clnt_debugfs_unregister(clnt);
 
 	/*
 	 * A new transport was created.  "clnt" therefore
@@ -771,6 +777,7 @@ rpc_free_client(struct rpc_clnt *clnt)
 			rcu_dereference(clnt->cl_xprt)->servername);
 	if (clnt->cl_parent != clnt)
 		parent = clnt->cl_parent;
+	rpc_clnt_debugfs_unregister(clnt);
 	rpc_clnt_remove_pipedir(clnt);
 	rpc_unregister_client(clnt);
 	rpc_free_iostats(clnt->cl_metrics);
@@ -1396,8 +1403,9 @@ rpc_restart_call(struct rpc_task *task)
 }
 EXPORT_SYMBOL_GPL(rpc_restart_call);
 
-#ifdef RPC_DEBUG
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
-static const char *rpc_proc_name(const struct rpc_task *task)
+const char
+*rpc_proc_name(const struct rpc_task *task)
 {
 	const struct rpc_procinfo *proc = task->tk_msg.rpc_proc;
 
@@ -2421,7 +2429,7 @@ struct rpc_task *rpc_call_null(struct rpc_clnt *clnt, struct rpc_cred *cred, int
 }
 EXPORT_SYMBOL_GPL(rpc_call_null);
 
-#ifdef RPC_DEBUG
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 static void rpc_show_header(void)
 {
 	printk(KERN_INFO "-pid- flgs status -client- --rqstp- "

net/sunrpc/debugfs.c

@@ -0,0 +1,292 @@
+/**
+ * debugfs interface for sunrpc
+ *
+ * (c) 2014 Jeff Layton <jlayton@primarydata.com>
+ */
+
+#include <linux/debugfs.h>
+#include <linux/sunrpc/sched.h>
+#include <linux/sunrpc/clnt.h>
+#include "netns.h"
+
+static struct dentry *topdir;
+static struct dentry *rpc_clnt_dir;
+static struct dentry *rpc_xprt_dir;
+
+struct rpc_clnt_iter {
+	struct rpc_clnt	*clnt;
+	loff_t		pos;
+};
+
+static int
+tasks_show(struct seq_file *f, void *v)
+{
+	u32 xid = 0;
+	struct rpc_task *task = v;
+	struct rpc_clnt *clnt = task->tk_client;
+	const char *rpc_waitq = "none";
+
+	if (RPC_IS_QUEUED(task))
+		rpc_waitq = rpc_qname(task->tk_waitqueue);
+
+	if (task->tk_rqstp)
+		xid = be32_to_cpu(task->tk_rqstp->rq_xid);
+
+	seq_printf(f, "%5u %04x %6d 0x%x 0x%x %8ld %ps %sv%u %s a:%ps q:%s\n",
+		task->tk_pid, task->tk_flags, task->tk_status,
+		clnt->cl_clid, xid, task->tk_timeout, task->tk_ops,
+		clnt->cl_program->name, clnt->cl_vers, rpc_proc_name(task),
+		task->tk_action, rpc_waitq);
+	return 0;
+}
+
+static void *
+tasks_start(struct seq_file *f, loff_t *ppos)
+	__acquires(&clnt->cl_lock)
+{
+	struct rpc_clnt_iter *iter = f->private;
+	loff_t pos = *ppos;
+	struct rpc_clnt *clnt = iter->clnt;
+	struct rpc_task *task;
+
+	iter->pos = pos + 1;
+	spin_lock(&clnt->cl_lock);
+	list_for_each_entry(task, &clnt->cl_tasks, tk_task)
+		if (pos-- == 0)
+			return task;
+	return NULL;
+}
+
+static void *
+tasks_next(struct seq_file *f, void *v, loff_t *pos)
+{
+	struct rpc_clnt_iter *iter = f->private;
+	struct rpc_clnt *clnt = iter->clnt;
+	struct rpc_task *task = v;
+	struct list_head *next = task->tk_task.next;
+
+	++iter->pos;
+	++*pos;
+
+	/* If there's another task on list, return it */
+	if (next == &clnt->cl_tasks)
+		return NULL;
+	return list_entry(next, struct rpc_task, tk_task);
+}
+
+static void
+tasks_stop(struct seq_file *f, void *v)
+	__releases(&clnt->cl_lock)
+{
+	struct rpc_clnt_iter *iter = f->private;
+	struct rpc_clnt *clnt = iter->clnt;
+
+	spin_unlock(&clnt->cl_lock);
+}
+
+static const struct seq_operations tasks_seq_operations = {
+	.start	= tasks_start,
+	.next	= tasks_next,
+	.stop	= tasks_stop,
+	.show	= tasks_show,
+};
+
+static int tasks_open(struct inode *inode, struct file *filp)
+{
+	int ret = seq_open_private(filp, &tasks_seq_operations,
+					sizeof(struct rpc_clnt_iter));
+
+	if (!ret) {
+		struct seq_file *seq = filp->private_data;
+		struct rpc_clnt_iter *iter = seq->private;
+
+		iter->clnt = inode->i_private;
+
+		if (!atomic_inc_not_zero(&iter->clnt->cl_count)) {
+			seq_release_private(inode, filp);
+			ret = -EINVAL;
+		}
+	}
+
+	return ret;
+}
+
+static int
+tasks_release(struct inode *inode, struct file *filp)
+{
+	struct seq_file *seq = filp->private_data;
+	struct rpc_clnt_iter *iter = seq->private;
+
+	rpc_release_client(iter->clnt);
+	return seq_release_private(inode, filp);
+}
+
+static const struct file_operations tasks_fops = {
+	.owner		= THIS_MODULE,
+	.open		= tasks_open,
+	.read		= seq_read,
+	.llseek		= seq_lseek,
+	.release	= tasks_release,
+};
+
+int
+rpc_clnt_debugfs_register(struct rpc_clnt *clnt)
+{
+	int len, err;
+	char name[24]; /* enough for "../../rpc_xprt/ + 8 hex digits + NULL */
+
+	/* Already registered? */
+	if (clnt->cl_debugfs)
+		return 0;
+
+	len = snprintf(name, sizeof(name), "%x", clnt->cl_clid);
+	if (len >= sizeof(name))
+		return -EINVAL;
+
+	/* make the per-client dir */
+	clnt->cl_debugfs = debugfs_create_dir(name, rpc_clnt_dir);
+	if (!clnt->cl_debugfs)
+		return -ENOMEM;
+
+	/* make tasks file */
+	err = -ENOMEM;
+	if (!debugfs_create_file("tasks", S_IFREG | S_IRUSR, clnt->cl_debugfs,
+				 clnt, &tasks_fops))
+		goto out_err;
+
+	err = -EINVAL;
+	rcu_read_lock();
+	len = snprintf(name, sizeof(name), "../../rpc_xprt/%s",
+			rcu_dereference(clnt->cl_xprt)->debugfs->d_name.name);
+	rcu_read_unlock();
+	if (len >= sizeof(name))
+		goto out_err;
+
+	err = -ENOMEM;
+	if (!debugfs_create_symlink("xprt", clnt->cl_debugfs, name))
+		goto out_err;
+
+	return 0;
+out_err:
+	debugfs_remove_recursive(clnt->cl_debugfs);
+	clnt->cl_debugfs = NULL;
+	return err;
+}
+
+void
+rpc_clnt_debugfs_unregister(struct rpc_clnt *clnt)
+{
+	debugfs_remove_recursive(clnt->cl_debugfs);
+	clnt->cl_debugfs = NULL;
+}
+
+static int
+xprt_info_show(struct seq_file *f, void *v)
+{
+	struct rpc_xprt *xprt = f->private;
+
+	seq_printf(f, "netid: %s\n", xprt->address_strings[RPC_DISPLAY_NETID]);
+	seq_printf(f, "addr:  %s\n", xprt->address_strings[RPC_DISPLAY_ADDR]);
+	seq_printf(f, "port:  %s\n", xprt->address_strings[RPC_DISPLAY_PORT]);
+	seq_printf(f, "state: 0x%lx\n", xprt->state);
+	return 0;
+}
+
+static int
+xprt_info_open(struct inode *inode, struct file *filp)
+{
+	int ret;
+	struct rpc_xprt *xprt = inode->i_private;
+
+	ret = single_open(filp, xprt_info_show, xprt);
+
+	if (!ret) {
+		if (!xprt_get(xprt)) {
+			single_release(inode, filp);
+			ret = -EINVAL;
+		}
+	}
+	return ret;
+}
+
+static int
+xprt_info_release(struct inode *inode, struct file *filp)
+{
+	struct rpc_xprt *xprt = inode->i_private;
+
+	xprt_put(xprt);
+	return single_release(inode, filp);
+}
+
+static const struct file_operations xprt_info_fops = {
+	.owner		= THIS_MODULE,
+	.open		= xprt_info_open,
+	.read		= seq_read,
+	.llseek		= seq_lseek,
+	.release	= xprt_info_release,
+};
+
+int
+rpc_xprt_debugfs_register(struct rpc_xprt *xprt)
+{
+	int len, id;
+	static atomic_t	cur_id;
+	char		name[9]; /* 8 hex digits + NULL term */
+
+	id = (unsigned int)atomic_inc_return(&cur_id);
+
+	len = snprintf(name, sizeof(name), "%x", id);
+	if (len >= sizeof(name))
+		return -EINVAL;
+
+	/* make the per-client dir */
+	xprt->debugfs = debugfs_create_dir(name, rpc_xprt_dir);
+	if (!xprt->debugfs)
+		return -ENOMEM;
+
+	/* make tasks file */
+	if (!debugfs_create_file("info", S_IFREG | S_IRUSR, xprt->debugfs,
+				 xprt, &xprt_info_fops)) {
+		debugfs_remove_recursive(xprt->debugfs);
+		xprt->debugfs = NULL;
+		return -ENOMEM;
+	}
+
+	return 0;
+}
+
+void
+rpc_xprt_debugfs_unregister(struct rpc_xprt *xprt)
+{
+	debugfs_remove_recursive(xprt->debugfs);
+	xprt->debugfs = NULL;
+}
+
+void __exit
+sunrpc_debugfs_exit(void)
+{
+	debugfs_remove_recursive(topdir);
+}
+
+int __init
+sunrpc_debugfs_init(void)
+{
+	topdir = debugfs_create_dir("sunrpc", NULL);
+	if (!topdir)
+		goto out;
+
+	rpc_clnt_dir = debugfs_create_dir("rpc_clnt", topdir);
+	if (!rpc_clnt_dir)
+		goto out_remove;
+
+	rpc_xprt_dir = debugfs_create_dir("rpc_xprt", topdir);
+	if (!rpc_xprt_dir)
+		goto out_remove;
+
+	return 0;
+out_remove:
+	debugfs_remove_recursive(topdir);
+	topdir = NULL;
+out:
+	return -ENOMEM;
+}

net/sunrpc/rpcb_clnt.c

@@ -32,7 +32,7 @@
 
 #include "netns.h"
 
-#ifdef RPC_DEBUG
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 # define RPCDBG_FACILITY	RPCDBG_BIND
 #endif
 

net/sunrpc/sched.c

@@ -24,7 +24,7 @@
 
 #include "sunrpc.h"
 
-#ifdef RPC_DEBUG
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 #define RPCDBG_FACILITY		RPCDBG_SCHED
 #endif
 
@@ -258,7 +258,7 @@ static int rpc_wait_bit_killable(struct wait_bit_key *key)
 	return 0;
 }
 
-#if defined(RPC_DEBUG) || defined(RPC_TRACEPOINTS)
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG) || IS_ENABLED(CONFIG_TRACEPOINTS)
 static void rpc_task_set_debuginfo(struct rpc_task *task)
 {
 	static atomic_t rpc_pid;

net/sunrpc/stats.c

@@ -116,7 +116,15 @@ EXPORT_SYMBOL_GPL(svc_seq_show);
  */
 struct rpc_iostats *rpc_alloc_iostats(struct rpc_clnt *clnt)
 {
-	return kcalloc(clnt->cl_maxproc, sizeof(struct rpc_iostats), GFP_KERNEL);
+	struct rpc_iostats *stats;
+	int i;
+
+	stats = kcalloc(clnt->cl_maxproc, sizeof(*stats), GFP_KERNEL);
+	if (stats) {
+		for (i = 0; i < clnt->cl_maxproc; i++)
+			spin_lock_init(&stats[i].om_lock);
+	}
+	return stats;
 }
 EXPORT_SYMBOL_GPL(rpc_alloc_iostats);
 
@@ -135,20 +143,21 @@ EXPORT_SYMBOL_GPL(rpc_free_iostats);
  * rpc_count_iostats - tally up per-task stats
  * @task: completed rpc_task
  * @stats: array of stat structures
- *
- * Relies on the caller for serialization.
  */
 void rpc_count_iostats(const struct rpc_task *task, struct rpc_iostats *stats)
 {
 	struct rpc_rqst *req = task->tk_rqstp;
 	struct rpc_iostats *op_metrics;
-	ktime_t delta;
+	ktime_t delta, now;
 
 	if (!stats || !req)
 		return;
 
+	now = ktime_get();
 	op_metrics = &stats[task->tk_msg.rpc_proc->p_statidx];
 
+	spin_lock(&op_metrics->om_lock);
+
 	op_metrics->om_ops++;
 	op_metrics->om_ntrans += req->rq_ntrans;
 	op_metrics->om_timeouts += task->tk_timeouts;
@@ -161,8 +170,10 @@ void rpc_count_iostats(const struct rpc_task *task, struct rpc_iostats *stats)
 
 	op_metrics->om_rtt = ktime_add(op_metrics->om_rtt, req->rq_rtt);
 
-	delta = ktime_sub(ktime_get(), task->tk_start);
+	delta = ktime_sub(now, task->tk_start);
 	op_metrics->om_execute = ktime_add(op_metrics->om_execute, delta);
+
+	spin_unlock(&op_metrics->om_lock);
 }
 EXPORT_SYMBOL_GPL(rpc_count_iostats);
 

net/sunrpc/sunrpc_syms.c

@@ -97,13 +97,20 @@ init_sunrpc(void)
 	err = register_rpc_pipefs();
 	if (err)
 		goto out4;
-#ifdef RPC_DEBUG
+
+	err = sunrpc_debugfs_init();
+	if (err)
+		goto out5;
+
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 	rpc_register_sysctl();
 #endif
 	svc_init_xprt_sock();	/* svc sock transport */
 	init_socket_xprt();	/* clnt sock transport */
 	return 0;
 
+out5:
+	unregister_rpc_pipefs();
 out4:
 	unregister_pernet_subsys(&sunrpc_net_ops);
 out3:
@@ -120,10 +127,11 @@ cleanup_sunrpc(void)
 	rpcauth_remove_module();
 	cleanup_socket_xprt();
 	svc_cleanup_xprt_sock();
+	sunrpc_debugfs_exit();
 	unregister_rpc_pipefs();
 	rpc_destroy_mempool();
 	unregister_pernet_subsys(&sunrpc_net_ops);
-#ifdef RPC_DEBUG
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 	rpc_unregister_sysctl();
 #endif
 	rcu_barrier(); /* Wait for completion of call_rcu()'s */

net/sunrpc/svc.c

@@ -28,6 +28,8 @@
 #include <linux/sunrpc/clnt.h>
 #include <linux/sunrpc/bc_xprt.h>
 
+#include <trace/events/sunrpc.h>
+
 #define RPCDBG_FACILITY	RPCDBG_SVCDSP
 
 static void svc_unregister(const struct svc_serv *serv, struct net *net);
@@ -1040,7 +1042,7 @@ static void svc_unregister(const struct svc_serv *serv, struct net *net)
 /*
  * dprintk the given error with the address of the client that caused it.
  */
-#ifdef RPC_DEBUG
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 static __printf(2, 3)
 void svc_printk(struct svc_rqst *rqstp, const char *fmt, ...)
 {
@@ -1314,24 +1316,25 @@ svc_process(struct svc_rqst *rqstp)
 	rqstp->rq_res.tail[0].iov_base = NULL;
 	rqstp->rq_res.tail[0].iov_len = 0;
 
-	rqstp->rq_xid = svc_getu32(argv);
-
 	dir  = svc_getnl(argv);
 	if (dir != 0) {
 		/* direction != CALL */
 		svc_printk(rqstp, "bad direction %d, dropping request\n", dir);
 		serv->sv_stats->rpcbadfmt++;
-		svc_drop(rqstp);
+		goto out_drop;
-		return 0;
 	}
 
 	/* Returns 1 for send, 0 for drop */
-	if (svc_process_common(rqstp, argv, resv))
+	if (likely(svc_process_common(rqstp, argv, resv))) {
-		return svc_send(rqstp);
+		int ret = svc_send(rqstp);
-	else {
+
-		svc_drop(rqstp);
+		trace_svc_process(rqstp, ret);
-		return 0;
+		return ret;
 	}
+out_drop:
+	trace_svc_process(rqstp, 0);
+	svc_drop(rqstp);
+	return 0;
 }
 
 #if defined(CONFIG_SUNRPC_BACKCHANNEL)

net/sunrpc/svc_xprt.c

@@ -15,6 +15,7 @@
 #include <linux/sunrpc/svcsock.h>
 #include <linux/sunrpc/xprt.h>
 #include <linux/module.h>
+#include <trace/events/sunrpc.h>
 
 #define RPCDBG_FACILITY	RPCDBG_SVCXPRT
 
@@ -773,35 +774,43 @@ int svc_recv(struct svc_rqst *rqstp, long timeout)
 
 	err = svc_alloc_arg(rqstp);
 	if (err)
-		return err;
+		goto out;
 
 	try_to_freeze();
 	cond_resched();
+	err = -EINTR;
 	if (signalled() || kthread_should_stop())
-		return -EINTR;
+		goto out;
 
 	xprt = svc_get_next_xprt(rqstp, timeout);
-	if (IS_ERR(xprt))
+	if (IS_ERR(xprt)) {
-		return PTR_ERR(xprt);
+		err = PTR_ERR(xprt);
+		goto out;
+	}
 
 	len = svc_handle_xprt(rqstp, xprt);
 
 	/* No data, incomplete (TCP) read, or accept() */
+	err = -EAGAIN;
 	if (len <= 0)
-		goto out;
+		goto out_release;
 
 	clear_bit(XPT_OLD, &xprt->xpt_flags);
 
 	rqstp->rq_secure = xprt->xpt_ops->xpo_secure_port(rqstp);
 	rqstp->rq_chandle.defer = svc_defer;
+	rqstp->rq_xid = svc_getu32(&rqstp->rq_arg.head[0]);
 
 	if (serv->sv_stats)
 		serv->sv_stats->netcnt++;
+	trace_svc_recv(rqstp, len);
 	return len;
-out:
+out_release:
 	rqstp->rq_res.len = 0;
 	svc_xprt_release(rqstp);
-	return -EAGAIN;
+out:
+	trace_svc_recv(rqstp, err);
+	return err;
 }
 EXPORT_SYMBOL_GPL(svc_recv);
 
@@ -821,12 +830,12 @@ EXPORT_SYMBOL_GPL(svc_drop);
 int svc_send(struct svc_rqst *rqstp)
 {
 	struct svc_xprt	*xprt;
-	int		len;
+	int		len = -EFAULT;
 	struct xdr_buf	*xb;
 
 	xprt = rqstp->rq_xprt;
 	if (!xprt)
-		return -EFAULT;
+		goto out;
 
 	/* release the receive skb before sending the reply */
 	rqstp->rq_xprt->xpt_ops->xpo_release_rqst(rqstp);
@@ -849,7 +858,9 @@ int svc_send(struct svc_rqst *rqstp)
 	svc_xprt_release(rqstp);
 
 	if (len == -ECONNREFUSED || len == -ENOTCONN || len == -EAGAIN)
-		return 0;
+		len = 0;
+out:
+	trace_svc_send(rqstp, len);
 	return len;
 }
 

net/sunrpc/sysctl.c

@@ -37,7 +37,7 @@ EXPORT_SYMBOL_GPL(nfsd_debug);
 unsigned int	nlm_debug;
 EXPORT_SYMBOL_GPL(nlm_debug);
 
-#ifdef RPC_DEBUG
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 
 static struct ctl_table_header *sunrpc_table_header;
 static struct ctl_table sunrpc_table[];

net/sunrpc/xprt.c

@@ -49,13 +49,15 @@
 #include <linux/sunrpc/metrics.h>
 #include <linux/sunrpc/bc_xprt.h>
 
+#include <trace/events/sunrpc.h>
+
 #include "sunrpc.h"
 
 /*
  * Local variables
  */
 
-#ifdef RPC_DEBUG
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 # define RPCDBG_FACILITY	RPCDBG_XPRT
 #endif
 
@@ -772,11 +774,14 @@ struct rpc_rqst *xprt_lookup_rqst(struct rpc_xprt *xprt, __be32 xid)
 	struct rpc_rqst *entry;
 
 	list_for_each_entry(entry, &xprt->recv, rq_list)
-		if (entry->rq_xid == xid)
+		if (entry->rq_xid == xid) {
+			trace_xprt_lookup_rqst(xprt, xid, 0);
 			return entry;
+		}
 
 	dprintk("RPC:       xprt_lookup_rqst did not find xid %08x\n",
 			ntohl(xid));
+	trace_xprt_lookup_rqst(xprt, xid, -ENOENT);
 	xprt->stat.bad_xids++;
 	return NULL;
 }
@@ -810,6 +815,7 @@ void xprt_complete_rqst(struct rpc_task *task, int copied)
 
 	dprintk("RPC: %5u xid %08x complete (%d bytes received)\n",
 			task->tk_pid, ntohl(req->rq_xid), copied);
+	trace_xprt_complete_rqst(xprt, req->rq_xid, copied);
 
 	xprt->stat.recvs++;
 	req->rq_rtt = ktime_sub(ktime_get(), req->rq_xtime);
@@ -926,6 +932,7 @@ void xprt_transmit(struct rpc_task *task)
 
 	req->rq_xtime = ktime_get();
 	status = xprt->ops->send_request(task);
+	trace_xprt_transmit(xprt, req->rq_xid, status);
 	if (status != 0) {
 		task->tk_status = status;
 		return;
@@ -1296,6 +1303,7 @@ static void xprt_init(struct rpc_xprt *xprt, struct net *net)
  */
 struct rpc_xprt *xprt_create_transport(struct xprt_create *args)
 {
+	int err;
 	struct rpc_xprt	*xprt;
 	struct xprt_class *t;
 
@@ -1336,6 +1344,12 @@ found:
 		return ERR_PTR(-ENOMEM);
 	}
 
+	err = rpc_xprt_debugfs_register(xprt);
+	if (err) {
+		xprt_destroy(xprt);
+		return ERR_PTR(err);
+	}
+
 	dprintk("RPC:       created transport %p with %u slots\n", xprt,
 			xprt->max_reqs);
 out:
@@ -1352,6 +1366,7 @@ static void xprt_destroy(struct rpc_xprt *xprt)
 	dprintk("RPC:       destroying transport %p\n", xprt);
 	del_timer_sync(&xprt->timer);
 
+	rpc_xprt_debugfs_unregister(xprt);
 	rpc_destroy_wait_queue(&xprt->binding);
 	rpc_destroy_wait_queue(&xprt->pending);
 	rpc_destroy_wait_queue(&xprt->sending);

net/sunrpc/xprtrdma/rpc_rdma.c

@@ -49,11 +49,11 @@
 
 #include <linux/highmem.h>
 
-#ifdef RPC_DEBUG
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 # define RPCDBG_FACILITY	RPCDBG_TRANS
 #endif
 
-#ifdef RPC_DEBUG
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 static const char transfertypes[][12] = {
 	"pure inline",	/* no chunks */
 	" read chunk",	/* some argument via rdma read */

net/sunrpc/xprtrdma/transport.c

@@ -55,7 +55,7 @@
 
 #include "xprt_rdma.h"
 
-#ifdef RPC_DEBUG
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 # define RPCDBG_FACILITY	RPCDBG_TRANS
 #endif
 
@@ -73,9 +73,9 @@ static unsigned int xprt_rdma_max_inline_read = RPCRDMA_DEF_INLINE;
 static unsigned int xprt_rdma_max_inline_write = RPCRDMA_DEF_INLINE;
 static unsigned int xprt_rdma_inline_write_padding;
 static unsigned int xprt_rdma_memreg_strategy = RPCRDMA_FRMR;
-                int xprt_rdma_pad_optimize = 0;
+		int xprt_rdma_pad_optimize = 1;
 
-#ifdef RPC_DEBUG
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 
 static unsigned int min_slot_table_size = RPCRDMA_MIN_SLOT_TABLE;
 static unsigned int max_slot_table_size = RPCRDMA_MAX_SLOT_TABLE;
@@ -599,7 +599,7 @@ xprt_rdma_send_request(struct rpc_task *task)
 
 	if (req->rl_niovs == 0)
 		rc = rpcrdma_marshal_req(rqst);
-	else if (r_xprt->rx_ia.ri_memreg_strategy == RPCRDMA_FRMR)
+	else if (r_xprt->rx_ia.ri_memreg_strategy != RPCRDMA_ALLPHYSICAL)
 		rc = rpcrdma_marshal_chunks(rqst, 0);
 	if (rc < 0)
 		goto failed_marshal;
@@ -705,7 +705,7 @@ static void __exit xprt_rdma_cleanup(void)
 	int rc;
 
 	dprintk("RPCRDMA Module Removed, deregister RPC RDMA transport\n");
-#ifdef RPC_DEBUG
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 	if (sunrpc_table_header) {
 		unregister_sysctl_table(sunrpc_table_header);
 		sunrpc_table_header = NULL;
@@ -736,7 +736,7 @@ static int __init xprt_rdma_init(void)
 	dprintk("\tPadding %d\n\tMemreg %d\n",
 		xprt_rdma_inline_write_padding, xprt_rdma_memreg_strategy);
 
-#ifdef RPC_DEBUG
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 	if (!sunrpc_table_header)
 		sunrpc_table_header = register_sysctl_table(sunrpc_table);
 #endif

net/sunrpc/xprtrdma/verbs.c

@@ -57,11 +57,12 @@
  * Globals/Macros
  */
 
-#ifdef RPC_DEBUG
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 # define RPCDBG_FACILITY	RPCDBG_TRANS
 #endif
 
 static void rpcrdma_reset_frmrs(struct rpcrdma_ia *);
+static void rpcrdma_reset_fmrs(struct rpcrdma_ia *);
 
 /*
  * internal functions
@@ -105,13 +106,51 @@ rpcrdma_run_tasklet(unsigned long data)
 
 static DECLARE_TASKLET(rpcrdma_tasklet_g, rpcrdma_run_tasklet, 0UL);
 
+static const char * const async_event[] = {
+	"CQ error",
+	"QP fatal error",
+	"QP request error",
+	"QP access error",
+	"communication established",
+	"send queue drained",
+	"path migration successful",
+	"path mig error",
+	"device fatal error",
+	"port active",
+	"port error",
+	"LID change",
+	"P_key change",
+	"SM change",
+	"SRQ error",
+	"SRQ limit reached",
+	"last WQE reached",
+	"client reregister",
+	"GID change",
+};
+
+#define ASYNC_MSG(status)					\
+	((status) < ARRAY_SIZE(async_event) ?			\
+		async_event[(status)] : "unknown async error")
+
+static void
+rpcrdma_schedule_tasklet(struct list_head *sched_list)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&rpcrdma_tk_lock_g, flags);
+	list_splice_tail(sched_list, &rpcrdma_tasklets_g);
+	spin_unlock_irqrestore(&rpcrdma_tk_lock_g, flags);
+	tasklet_schedule(&rpcrdma_tasklet_g);
+}
+
 static void
 rpcrdma_qp_async_error_upcall(struct ib_event *event, void *context)
 {
 	struct rpcrdma_ep *ep = context;
 
-	dprintk("RPC:       %s: QP error %X on device %s ep %p\n",
+	pr_err("RPC:       %s: %s on device %s ep %p\n",
-		__func__, event->event, event->device->name, context);
+	       __func__, ASYNC_MSG(event->event),
+		event->device->name, context);
 	if (ep->rep_connected == 1) {
 		ep->rep_connected = -EIO;
 		ep->rep_func(ep);
@@ -124,8 +163,9 @@ rpcrdma_cq_async_error_upcall(struct ib_event *event, void *context)
 {
 	struct rpcrdma_ep *ep = context;
 
-	dprintk("RPC:       %s: CQ error %X on device %s ep %p\n",
+	pr_err("RPC:       %s: %s on device %s ep %p\n",
-		__func__, event->event, event->device->name, context);
+	       __func__, ASYNC_MSG(event->event),
+		event->device->name, context);
 	if (ep->rep_connected == 1) {
 		ep->rep_connected = -EIO;
 		ep->rep_func(ep);
@@ -243,7 +283,6 @@ rpcrdma_recvcq_poll(struct ib_cq *cq, struct rpcrdma_ep *ep)
 	struct list_head sched_list;
 	struct ib_wc *wcs;
 	int budget, count, rc;
-	unsigned long flags;
 
 	INIT_LIST_HEAD(&sched_list);
 	budget = RPCRDMA_WC_BUDGET / RPCRDMA_POLLSIZE;
@@ -261,10 +300,7 @@ rpcrdma_recvcq_poll(struct ib_cq *cq, struct rpcrdma_ep *ep)
 	rc = 0;
 
 out_schedule:
-	spin_lock_irqsave(&rpcrdma_tk_lock_g, flags);
+	rpcrdma_schedule_tasklet(&sched_list);
-	list_splice_tail(&sched_list, &rpcrdma_tasklets_g);
-	spin_unlock_irqrestore(&rpcrdma_tk_lock_g, flags);
-	tasklet_schedule(&rpcrdma_tasklet_g);
 	return rc;
 }
 
@@ -309,11 +345,18 @@ rpcrdma_recvcq_upcall(struct ib_cq *cq, void *cq_context)
 static void
 rpcrdma_flush_cqs(struct rpcrdma_ep *ep)
 {
-	rpcrdma_recvcq_upcall(ep->rep_attr.recv_cq, ep);
+	struct ib_wc wc;
-	rpcrdma_sendcq_upcall(ep->rep_attr.send_cq, ep);
+	LIST_HEAD(sched_list);
+
+	while (ib_poll_cq(ep->rep_attr.recv_cq, 1, &wc) > 0)
+		rpcrdma_recvcq_process_wc(&wc, &sched_list);
+	if (!list_empty(&sched_list))
+		rpcrdma_schedule_tasklet(&sched_list);
+	while (ib_poll_cq(ep->rep_attr.send_cq, 1, &wc) > 0)
+		rpcrdma_sendcq_process_wc(&wc);
 }
 
-#ifdef RPC_DEBUG
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 static const char * const conn[] = {
 	"address resolved",
 	"address error",
@@ -344,7 +387,7 @@ rpcrdma_conn_upcall(struct rdma_cm_id *id, struct rdma_cm_event *event)
 	struct rpcrdma_xprt *xprt = id->context;
 	struct rpcrdma_ia *ia = &xprt->rx_ia;
 	struct rpcrdma_ep *ep = &xprt->rx_ep;
-#ifdef RPC_DEBUG
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 	struct sockaddr_in *addr = (struct sockaddr_in *) &ep->rep_remote_addr;
 #endif
 	struct ib_qp_attr attr;
@@ -408,7 +451,7 @@ connected:
 		break;
 	}
 
-#ifdef RPC_DEBUG
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 	if (connstate == 1) {
 		int ird = attr.max_dest_rd_atomic;
 		int tird = ep->rep_remote_cma.responder_resources;
@@ -733,7 +776,9 @@ rpcrdma_ep_create(struct rpcrdma_ep *ep, struct rpcrdma_ia *ia,
 
 	/* set trigger for requesting send completion */
 	ep->rep_cqinit = ep->rep_attr.cap.max_send_wr/2 - 1;
-	if (ep->rep_cqinit <= 2)
+	if (ep->rep_cqinit > RPCRDMA_MAX_UNSIGNALED_SENDS)
+		ep->rep_cqinit = RPCRDMA_MAX_UNSIGNALED_SENDS;
+	else if (ep->rep_cqinit <= 2)
 		ep->rep_cqinit = 0;
 	INIT_CQCOUNT(ep);
 	ep->rep_ia = ia;
@@ -866,8 +911,19 @@ retry:
 		rpcrdma_ep_disconnect(ep, ia);
 		rpcrdma_flush_cqs(ep);
 
-		if (ia->ri_memreg_strategy == RPCRDMA_FRMR)
+		switch (ia->ri_memreg_strategy) {
+		case RPCRDMA_FRMR:
 			rpcrdma_reset_frmrs(ia);
+			break;
+		case RPCRDMA_MTHCAFMR:
+			rpcrdma_reset_fmrs(ia);
+			break;
+		case RPCRDMA_ALLPHYSICAL:
+			break;
+		default:
+			rc = -EIO;
+			goto out;
+		}
 
 		xprt = container_of(ia, struct rpcrdma_xprt, rx_ia);
 		id = rpcrdma_create_id(xprt, ia,
@@ -1287,6 +1343,34 @@ rpcrdma_buffer_destroy(struct rpcrdma_buffer *buf)
 	kfree(buf->rb_pool);
 }
 
+/* After a disconnect, unmap all FMRs.
+ *
+ * This is invoked only in the transport connect worker in order
+ * to serialize with rpcrdma_register_fmr_external().
+ */
+static void
+rpcrdma_reset_fmrs(struct rpcrdma_ia *ia)
+{
+	struct rpcrdma_xprt *r_xprt =
+				container_of(ia, struct rpcrdma_xprt, rx_ia);
+	struct rpcrdma_buffer *buf = &r_xprt->rx_buf;
+	struct list_head *pos;
+	struct rpcrdma_mw *r;
+	LIST_HEAD(l);
+	int rc;
+
+	list_for_each(pos, &buf->rb_all) {
+		r = list_entry(pos, struct rpcrdma_mw, mw_all);
+
+		INIT_LIST_HEAD(&l);
+		list_add(&r->r.fmr->list, &l);
+		rc = ib_unmap_fmr(&l);
+		if (rc)
+			dprintk("RPC:       %s: ib_unmap_fmr failed %i\n",
+				__func__, rc);
+	}
+}
+
 /* After a disconnect, a flushed FAST_REG_MR can leave an FRMR in
  * an unusable state. Find FRMRs in this state and dereg / reg
  * each.  FRMRs that are VALID and attached to an rpcrdma_req are
@@ -1918,10 +2002,10 @@ rpcrdma_register_external(struct rpcrdma_mr_seg *seg,
 		break;
 
 	default:
-		return -1;
+		return -EIO;
 	}
 	if (rc)
-		return -1;
+		return rc;
 
 	return nsegs;
 }

net/sunrpc/xprtrdma/xprt_rdma.h

@@ -97,6 +97,12 @@ struct rpcrdma_ep {
 	struct ib_wc		rep_recv_wcs[RPCRDMA_POLLSIZE];
 };
 
+/*
+ * Force a signaled SEND Work Request every so often,
+ * in case the provider needs to do some housekeeping.
+ */
+#define RPCRDMA_MAX_UNSIGNALED_SENDS	(32)
+
 #define INIT_CQCOUNT(ep) atomic_set(&(ep)->rep_cqcount, (ep)->rep_cqinit)
 #define DECR_CQCOUNT(ep) atomic_sub_return(1, &(ep)->rep_cqcount)
 

net/sunrpc/xprtsock.c

@@ -75,7 +75,7 @@ static unsigned int xs_tcp_fin_timeout __read_mostly = XS_TCP_LINGER_TO;
  * someone else's file names!
  */
 
-#ifdef RPC_DEBUG
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 
 static unsigned int min_slot_table_size = RPC_MIN_SLOT_TABLE;
 static unsigned int max_slot_table_size = RPC_MAX_SLOT_TABLE;
@@ -186,7 +186,7 @@ static struct ctl_table sunrpc_table[] = {
  */
 #define XS_IDLE_DISC_TO		(5U * 60 * HZ)
 
-#ifdef RPC_DEBUG
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 # undef  RPC_DEBUG_DATA
 # define RPCDBG_FACILITY	RPCDBG_TRANS
 #endif
@@ -216,65 +216,6 @@ static inline void xs_pktdump(char *msg, u32 *packet, unsigned int count)
 }
 #endif
 
-struct sock_xprt {
-	struct rpc_xprt		xprt;
-
-	/*
-	 * Network layer
-	 */
-	struct socket *		sock;
-	struct sock *		inet;
-
-	/*
-	 * State of TCP reply receive
-	 */
-	__be32			tcp_fraghdr,
-				tcp_xid,
-				tcp_calldir;
-
-	u32			tcp_offset,
-				tcp_reclen;
-
-	unsigned long		tcp_copied,
-				tcp_flags;
-
-	/*
-	 * Connection of transports
-	 */
-	struct delayed_work	connect_worker;
-	struct sockaddr_storage	srcaddr;
-	unsigned short		srcport;
-
-	/*
-	 * UDP socket buffer size parameters
-	 */
-	size_t			rcvsize,
-				sndsize;
-
-	/*
-	 * Saved socket callback addresses
-	 */
-	void			(*old_data_ready)(struct sock *);
-	void			(*old_state_change)(struct sock *);
-	void			(*old_write_space)(struct sock *);
-	void			(*old_error_report)(struct sock *);
-};
-
-/*
- * TCP receive state flags
- */
-#define TCP_RCV_LAST_FRAG	(1UL << 0)
-#define TCP_RCV_COPY_FRAGHDR	(1UL << 1)
-#define TCP_RCV_COPY_XID	(1UL << 2)
-#define TCP_RCV_COPY_DATA	(1UL << 3)
-#define TCP_RCV_READ_CALLDIR	(1UL << 4)
-#define TCP_RCV_COPY_CALLDIR	(1UL << 5)
-
-/*
- * TCP RPC flags
- */
-#define TCP_RPC_REPLY		(1UL << 6)
-
 static inline struct rpc_xprt *xprt_from_sock(struct sock *sk)
 {
 	return (struct rpc_xprt *) sk->sk_user_data;
@@ -1415,6 +1356,7 @@ static int xs_tcp_data_recv(read_descriptor_t *rd_desc, struct sk_buff *skb, uns
 
 	dprintk("RPC:       xs_tcp_data_recv started\n");
 	do {
+		trace_xs_tcp_data_recv(transport);
 		/* Read in a new fragment marker if necessary */
 		/* Can we ever really expect to get completely empty fragments? */
 		if (transport->tcp_flags & TCP_RCV_COPY_FRAGHDR) {
@@ -1439,6 +1381,7 @@ static int xs_tcp_data_recv(read_descriptor_t *rd_desc, struct sk_buff *skb, uns
 		/* Skip over any trailing bytes on short reads */
 		xs_tcp_read_discard(transport, &desc);
 	} while (desc.count);
+	trace_xs_tcp_data_recv(transport);
 	dprintk("RPC:       xs_tcp_data_recv done\n");
 	return len - desc.count;
 }
@@ -1454,12 +1397,15 @@ static void xs_tcp_data_ready(struct sock *sk)
 	struct rpc_xprt *xprt;
 	read_descriptor_t rd_desc;
 	int read;
+	unsigned long total = 0;
 
 	dprintk("RPC:       xs_tcp_data_ready...\n");
 
 	read_lock_bh(&sk->sk_callback_lock);
-	if (!(xprt = xprt_from_sock(sk)))
+	if (!(xprt = xprt_from_sock(sk))) {
+		read = 0;
 		goto out;
+	}
 	/* Any data means we had a useful conversation, so
 	 * the we don't need to delay the next reconnect
 	 */
@@ -1471,8 +1417,11 @@ static void xs_tcp_data_ready(struct sock *sk)
 	do {
 		rd_desc.count = 65536;
 		read = tcp_read_sock(sk, &rd_desc, xs_tcp_data_recv);
+		if (read > 0)
+			total += read;
 	} while (read > 0);
 out:
+	trace_xs_tcp_data_ready(xprt, read, total);
 	read_unlock_bh(&sk->sk_callback_lock);
 }
 
@@ -3042,7 +2991,7 @@ static struct xprt_class	xs_bc_tcp_transport = {
  */
 int init_socket_xprt(void)
 {
-#ifdef RPC_DEBUG
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 	if (!sunrpc_table_header)
 		sunrpc_table_header = register_sysctl_table(sunrpc_table);
 #endif
@@ -3061,7 +3010,7 @@ int init_socket_xprt(void)
  */
 void cleanup_socket_xprt(void)
 {
-#ifdef RPC_DEBUG
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 	if (sunrpc_table_header) {
 		unregister_sysctl_table(sunrpc_table_header);
 		sunrpc_table_header = NULL;