Merge tag 'nfsd-6.11' of git://git.kernel.org/pub/scm/linux/kernel/git/cel/linux

Pull nfsd updates from Chuck Lever:
 "This is a light release containing optimizations, code clean-ups, and
  minor bug fixes.

  This development cycle focused on work outside of upstream kernel
  development:

   - Continuing to build upstream CI for NFSD based on kdevops

   - Continuing to focus on the quality of NFSD in LTS kernels

   - Participation in IETF nfsv4 WG discussions about NFSv4 ACLs,
     directory delegation, and NFSv4.2 COPY offload

  Notable features for v6.11 that do not come through the NFSD tree
  include NFS server-side support for the new pNFS NVMe layout type
  [RFC9561]. Functional testing for pNFS block layouts like this one has
  been introduced to our kdevops CI harness. Work on improving the
  resolution of file attribute time stamps in local filesystems is also
  ongoing tree-wide.

  As always I am grateful to NFSD contributors, reviewers, testers, and
  bug reporters who participated during this cycle"

* tag 'nfsd-6.11' of git://git.kernel.org/pub/scm/linux/kernel/git/cel/linux:
  nfsd: nfsd_file_lease_notifier_call gets a file_lease as an argument
  gss_krb5: Fix the error handling path for crypto_sync_skcipher_setkey
  MAINTAINERS: Add a bugzilla link for NFSD
  nfsd: new netlink ops to get/set server pool_mode
  sunrpc: refactor pool_mode setting code
  nfsd: allow passing in array of thread counts via netlink
  nfsd: make nfsd_svc take an array of thread counts
  sunrpc: fix up the special handling of sv_nrpools == 1
  SUNRPC: Add a trace point in svc_xprt_deferred_close
  NFSD: Support write delegations in LAYOUTGET
  lockd: Use *-y instead of *-objs in Makefile
  NFSD: Fix nfsdcld warning
  svcrdma: Handle ADDR_CHANGE CM event properly
  svcrdma: Refactor the creation of listener CMA ID
  NFSD: remove unused structs 'nfsd3_voidargs'
  NFSD: harden svcxdr_dupstr() and svcxdr_tmpalloc() against integer overflows

13 files changed, 168 insertions, 57 deletions

diff --git a/fs/lockd/Makefile b/fs/lockd/Makefile
index ac9f9d84510e..fe3e23dd29c3 100644
--- a/fs/lockd/Makefile
+++ b/fs/lockd/Makefile
@@ -7,8 +7,7 @@ ccflags-y += -I$(src)			# needed for trace events
 
 obj-$(CONFIG_LOCKD) += lockd.o
 
-lockd-objs-y += clntlock.o clntproc.o clntxdr.o host.o svc.o svclock.o \
-	        svcshare.o svcproc.o svcsubs.o mon.o trace.o xdr.o
-lockd-objs-$(CONFIG_LOCKD_V4) += clnt4xdr.o xdr4.o svc4proc.o
-lockd-objs-$(CONFIG_PROC_FS) += procfs.o
-lockd-objs		      := $(lockd-objs-y)
+lockd-y := clntlock.o clntproc.o clntxdr.o host.o svc.o svclock.o \
+	   svcshare.o svcproc.o svcsubs.o mon.o trace.o xdr.o
+lockd-$(CONFIG_LOCKD_V4) += clnt4xdr.o xdr4.o svc4proc.o
+lockd-$(CONFIG_PROC_FS) += procfs.o
diff --git a/fs/nfsd/Kconfig b/fs/nfsd/Kconfig
index 272ab8d5c4d7..ec2ab6429e00 100644
--- a/fs/nfsd/Kconfig
+++ b/fs/nfsd/Kconfig
@@ -162,7 +162,7 @@ config NFSD_V4_SECURITY_LABEL
 config NFSD_LEGACY_CLIENT_TRACKING
 	bool "Support legacy NFSv4 client tracking methods (DEPRECATED)"
 	depends on NFSD_V4
-	default n
+	default y
 	help
 	  The NFSv4 server needs to store a small amount of information on
 	  stable storage in order to handle state recovery after reboot. Most
diff --git a/fs/nfsd/filecache.c b/fs/nfsd/filecache.c
index ad9083ca144b..f4704f5d4086 100644
--- a/fs/nfsd/filecache.c
+++ b/fs/nfsd/filecache.c
@@ -664,7 +664,7 @@ static int
 nfsd_file_lease_notifier_call(struct notifier_block *nb, unsigned long arg,
 			    void *data)
 {
-	struct file_lock *fl = data;
+	struct file_lease *fl = data;
 
 	/* Only close files for F_SETLEASE leases */
 	if (fl->c.flc_flags & FL_LEASE)
diff --git a/fs/nfsd/netlink.c b/fs/nfsd/netlink.c
index 529a75ecf22e..ca54aa583530 100644
--- a/fs/nfsd/netlink.c
+++ b/fs/nfsd/netlink.c
@@ -40,6 +40,11 @@ static const struct nla_policy nfsd_listener_set_nl_policy[NFSD_A_SERVER_SOCK_AD
 	[NFSD_A_SERVER_SOCK_ADDR] = NLA_POLICY_NESTED(nfsd_sock_nl_policy),
 };
 
+/* NFSD_CMD_POOL_MODE_SET - do */
+static const struct nla_policy nfsd_pool_mode_set_nl_policy[NFSD_A_POOL_MODE_MODE + 1] = {
+	[NFSD_A_POOL_MODE_MODE] = { .type = NLA_NUL_STRING, },
+};
+
 /* Ops table for nfsd */
 static const struct genl_split_ops nfsd_nl_ops[] = {
 	{
@@ -83,6 +88,18 @@ static const struct genl_split_ops nfsd_nl_ops[] = {
 		.doit	= nfsd_nl_listener_get_doit,
 		.flags	= GENL_CMD_CAP_DO,
 	},
+	{
+		.cmd		= NFSD_CMD_POOL_MODE_SET,
+		.doit		= nfsd_nl_pool_mode_set_doit,
+		.policy		= nfsd_pool_mode_set_nl_policy,
+		.maxattr	= NFSD_A_POOL_MODE_MODE,
+		.flags		= GENL_ADMIN_PERM | GENL_CMD_CAP_DO,
+	},
+	{
+		.cmd	= NFSD_CMD_POOL_MODE_GET,
+		.doit	= nfsd_nl_pool_mode_get_doit,
+		.flags	= GENL_CMD_CAP_DO,
+	},
 };
 
 struct genl_family nfsd_nl_family __ro_after_init = {
diff --git a/fs/nfsd/netlink.h b/fs/nfsd/netlink.h
index 2e132ef328f8..8eb903f24c41 100644
--- a/fs/nfsd/netlink.h
+++ b/fs/nfsd/netlink.h
@@ -23,6 +23,8 @@ int nfsd_nl_version_set_doit(struct sk_buff *skb, struct genl_info *info);
 int nfsd_nl_version_get_doit(struct sk_buff *skb, struct genl_info *info);
 int nfsd_nl_listener_set_doit(struct sk_buff *skb, struct genl_info *info);
 int nfsd_nl_listener_get_doit(struct sk_buff *skb, struct genl_info *info);
+int nfsd_nl_pool_mode_set_doit(struct sk_buff *skb, struct genl_info *info);
+int nfsd_nl_pool_mode_get_doit(struct sk_buff *skb, struct genl_info *info);
 
 extern struct genl_family nfsd_nl_family;
 
diff --git a/fs/nfsd/nfs2acl.c b/fs/nfsd/nfs2acl.c
index 12b2b9bc07bf..4e3be7201b1c 100644
--- a/fs/nfsd/nfs2acl.c
+++ b/fs/nfsd/nfs2acl.c
@@ -308,8 +308,6 @@ static void nfsaclsvc_release_access(struct svc_rqst *rqstp)
 	fh_put(&resp->fh);
 }
 
-struct nfsd3_voidargs { int dummy; };
-
 #define ST 1		/* status*/
 #define AT 21		/* attributes */
 #define pAT (1+AT)	/* post attributes - conditional */
diff --git a/fs/nfsd/nfs3acl.c b/fs/nfsd/nfs3acl.c
index 73adca47d373..5e34e98db969 100644
--- a/fs/nfsd/nfs3acl.c
+++ b/fs/nfsd/nfs3acl.c
@@ -221,8 +221,6 @@ static void nfs3svc_release_getacl(struct svc_rqst *rqstp)
 	posix_acl_release(resp->acl_default);
 }
 
-struct nfsd3_voidargs { int dummy; };
-
 #define ST 1		/* status*/
 #define AT 21		/* attributes */
 #define pAT (1+AT)	/* post attributes - conditional */
diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c
index 46bd20fe5c0f..2e39cf2e502a 100644
--- a/fs/nfsd/nfs4proc.c
+++ b/fs/nfsd/nfs4proc.c
@@ -2269,7 +2269,7 @@ nfsd4_layoutget(struct svc_rqst *rqstp,
 	const struct nfsd4_layout_ops *ops;
 	struct nfs4_layout_stateid *ls;
 	__be32 nfserr;
-	int accmode = NFSD_MAY_READ_IF_EXEC;
+	int accmode = NFSD_MAY_READ_IF_EXEC | NFSD_MAY_OWNER_OVERRIDE;
 
 	switch (lgp->lg_seg.iomode) {
 	case IOMODE_READ:
@@ -2359,7 +2359,8 @@ nfsd4_layoutcommit(struct svc_rqst *rqstp,
 	struct nfs4_layout_stateid *ls;
 	__be32 nfserr;
 
-	nfserr = fh_verify(rqstp, current_fh, 0, NFSD_MAY_WRITE);
+	nfserr = fh_verify(rqstp, current_fh, 0,
+			   NFSD_MAY_WRITE | NFSD_MAY_OWNER_OVERRIDE);
 	if (nfserr)
 		goto out;
 
diff --git a/fs/nfsd/nfs4recover.c b/fs/nfsd/nfs4recover.c
index 2c060e0b1604..67d8673a9391 100644
--- a/fs/nfsd/nfs4recover.c
+++ b/fs/nfsd/nfs4recover.c
@@ -2086,8 +2086,8 @@ do_init:
 	status = nn->client_tracking_ops->init(net);
 out:
 	if (status) {
-		printk(KERN_WARNING "NFSD: Unable to initialize client "
-				    "recovery tracking! (%d)\n", status);
+		pr_warn("NFSD: Unable to initialize client recovery tracking! (%d)\n", status);
+		pr_warn("NFSD: Is nfsdcld running? If not, enable CONFIG_NFSD_LEGACY_CLIENT_TRACKING.\n");
 		nn->client_tracking_ops = NULL;
 	}
 	return status;
diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c
index c7bfd2180e3f..42b41d55d4ed 100644
--- a/fs/nfsd/nfs4xdr.c
+++ b/fs/nfsd/nfs4xdr.c
@@ -118,11 +118,11 @@ static int zero_clientid(clientid_t *clid)
  * operation described in @argp finishes.
  */
 static void *
-svcxdr_tmpalloc(struct nfsd4_compoundargs *argp, u32 len)
+svcxdr_tmpalloc(struct nfsd4_compoundargs *argp, size_t len)
 {
 	struct svcxdr_tmpbuf *tb;
 
-	tb = kmalloc(sizeof(*tb) + len, GFP_KERNEL);
+	tb = kmalloc(struct_size(tb, buf, len), GFP_KERNEL);
 	if (!tb)
 		return NULL;
 	tb->next = argp->to_free;
@@ -138,9 +138,9 @@ svcxdr_tmpalloc(struct nfsd4_compoundargs *argp, u32 len)
  * buffer might end on a page boundary.
  */
 static char *
-svcxdr_dupstr(struct nfsd4_compoundargs *argp, void *buf, u32 len)
+svcxdr_dupstr(struct nfsd4_compoundargs *argp, void *buf, size_t len)
 {
-	char *p = svcxdr_tmpalloc(argp, len + 1);
+	char *p = svcxdr_tmpalloc(argp, size_add(len, 1));
 
 	if (!p)
 		return NULL;
@@ -150,7 +150,7 @@ svcxdr_dupstr(struct nfsd4_compoundargs *argp, void *buf, u32 len)
 }
 
 static void *
-svcxdr_savemem(struct nfsd4_compoundargs *argp, __be32 *p, u32 len)
+svcxdr_savemem(struct nfsd4_compoundargs *argp, __be32 *p, size_t len)
 {
 	__be32 *tmp;
 
@@ -2146,7 +2146,7 @@ nfsd4_decode_clone(struct nfsd4_compoundargs *argp, union nfsd4_op_u *u)
  */
 static __be32
 nfsd4_vbuf_from_vector(struct nfsd4_compoundargs *argp, struct xdr_buf *xdr,
-		       char **bufp, u32 buflen)
+		       char **bufp, size_t buflen)
 {
 	struct page **pages = xdr->pages;
 	struct kvec *head = xdr->head;
diff --git a/fs/nfsd/nfsctl.c b/fs/nfsd/nfsctl.c
index c848ebe5d08f..9e0ea6fc2aa3 100644
--- a/fs/nfsd/nfsctl.c
+++ b/fs/nfsd/nfsctl.c
@@ -406,7 +406,7 @@ static ssize_t write_threads(struct file *file, char *buf, size_t size)
 			return -EINVAL;
 		trace_nfsd_ctl_threads(net, newthreads);
 		mutex_lock(&nfsd_mutex);
-		rv = nfsd_svc(newthreads, net, file->f_cred, NULL);
+		rv = nfsd_svc(1, &newthreads, net, file->f_cred, NULL);
 		mutex_unlock(&nfsd_mutex);
 		if (rv < 0)
 			return rv;
@@ -481,6 +481,14 @@ static ssize_t write_pool_threads(struct file *file, char *buf, size_t size)
 				goto out_free;
 			trace_nfsd_ctl_pool_threads(net, i, nthreads[i]);
 		}
+
+		/*
+		 * There must always be a thread in pool 0; the admin
+		 * can't shut down NFS completely using pool_threads.
+		 */
+		if (nthreads[0] == 0)
+			nthreads[0] = 1;
+
 		rv = nfsd_set_nrthreads(i, nthreads, net);
 		if (rv)
 			goto out_free;
@@ -1637,7 +1645,7 @@ out_unlock:
  */
 int nfsd_nl_threads_set_doit(struct sk_buff *skb, struct genl_info *info)
 {
-	int nthreads = 0, count = 0, nrpools, ret = -EOPNOTSUPP, rem;
+	int *nthreads, count = 0, nrpools, i, ret = -EOPNOTSUPP, rem;
 	struct net *net = genl_info_net(info);
 	struct nfsd_net *nn = net_generic(net, nfsd_net_id);
 	const struct nlattr *attr;
@@ -1654,15 +1662,22 @@ int nfsd_nl_threads_set_doit(struct sk_buff *skb, struct genl_info *info)
 
 	mutex_lock(&nfsd_mutex);
 
-	nrpools = nfsd_nrpools(net);
-	if (nrpools && count > nrpools)
-		count = nrpools;
-
-	/* XXX: make this handle non-global pool-modes */
-	if (count > 1)
+	nrpools = max(count, nfsd_nrpools(net));
+	nthreads = kcalloc(nrpools, sizeof(int), GFP_KERNEL);
+	if (!nthreads) {
+		ret = -ENOMEM;
 		goto out_unlock;
+	}
+
+	i = 0;
+	nlmsg_for_each_attr(attr, info->nlhdr, GENL_HDRLEN, rem) {
+		if (nla_type(attr) == NFSD_A_SERVER_THREADS) {
+			nthreads[i++] = nla_get_u32(attr);
+			if (i >= nrpools)
+				break;
+		}
+	}
 
-	nthreads = nla_get_u32(info->attrs[NFSD_A_SERVER_THREADS]);
 	if (info->attrs[NFSD_A_SERVER_GRACETIME] ||
 	    info->attrs[NFSD_A_SERVER_LEASETIME] ||
 	    info->attrs[NFSD_A_SERVER_SCOPE]) {
@@ -1696,12 +1711,13 @@ int nfsd_nl_threads_set_doit(struct sk_buff *skb, struct genl_info *info)
 			scope = nla_data(attr);
 	}
 
-	ret = nfsd_svc(nthreads, net, get_current_cred(), scope);
-
+	ret = nfsd_svc(nrpools, nthreads, net, get_current_cred(), scope);
+	if (ret > 0)
+		ret = 0;
 out_unlock:
 	mutex_unlock(&nfsd_mutex);
-
-	return ret == nthreads ? 0 : ret;
+	kfree(nthreads);
+	return ret;
 }
 
 /**
@@ -2141,6 +2157,63 @@ err_free_msg:
 }
 
 /**
+ * nfsd_nl_pool_mode_set_doit - set the number of running threads
+ * @skb: reply buffer
+ * @info: netlink metadata and command arguments
+ *
+ * Return 0 on success or a negative errno.
+ */
+int nfsd_nl_pool_mode_set_doit(struct sk_buff *skb, struct genl_info *info)
+{
+	const struct nlattr *attr;
+
+	if (GENL_REQ_ATTR_CHECK(info, NFSD_A_POOL_MODE_MODE))
+		return -EINVAL;
+
+	attr = info->attrs[NFSD_A_POOL_MODE_MODE];
+	return sunrpc_set_pool_mode(nla_data(attr));
+}
+
+/**
+ * nfsd_nl_pool_mode_get_doit - get info about pool_mode
+ * @skb: reply buffer
+ * @info: netlink metadata and command arguments
+ *
+ * Return 0 on success or a negative errno.
+ */
+int nfsd_nl_pool_mode_get_doit(struct sk_buff *skb, struct genl_info *info)
+{
+	struct net *net = genl_info_net(info);
+	char buf[16];
+	void *hdr;
+	int err;
+
+	if (sunrpc_get_pool_mode(buf, ARRAY_SIZE(buf)) >= ARRAY_SIZE(buf))
+		return -ERANGE;
+
+	skb = genlmsg_new(GENLMSG_DEFAULT_SIZE, GFP_KERNEL);
+	if (!skb)
+		return -ENOMEM;
+
+	err = -EMSGSIZE;
+	hdr = genlmsg_iput(skb, info);
+	if (!hdr)
+		goto err_free_msg;
+
+	err = nla_put_string(skb, NFSD_A_POOL_MODE_MODE, buf) |
+	      nla_put_u32(skb, NFSD_A_POOL_MODE_NPOOLS, nfsd_nrpools(net));
+	if (err)
+		goto err_free_msg;
+
+	genlmsg_end(skb, hdr);
+	return genlmsg_reply(skb, info);
+
+err_free_msg:
+	nlmsg_free(skb);
+	return err;
+}
+
+/**
  * nfsd_net_init - Prepare the nfsd_net portion of a new net namespace
  * @net: a freshly-created network namespace
  *
diff --git a/fs/nfsd/nfsd.h b/fs/nfsd/nfsd.h
index 8f4f239d9f8a..cec8697b1cd6 100644
--- a/fs/nfsd/nfsd.h
+++ b/fs/nfsd/nfsd.h
@@ -103,7 +103,8 @@ bool		nfssvc_encode_voidres(struct svc_rqst *rqstp,
 /*
  * Function prototypes.
  */
-int		nfsd_svc(int nrservs, struct net *net, const struct cred *cred, const char *scope);
+int		nfsd_svc(int n, int *nservers, struct net *net,
+			 const struct cred *cred, const char *scope);
 int		nfsd_dispatch(struct svc_rqst *rqstp);
 
 int		nfsd_nrthreads(struct net *);
diff --git a/fs/nfsd/nfssvc.c b/fs/nfsd/nfssvc.c
index 89d7918de7b1..0bc8eaa5e009 100644
--- a/fs/nfsd/nfssvc.c
+++ b/fs/nfsd/nfssvc.c
@@ -709,6 +709,19 @@ int nfsd_get_nrthreads(int n, int *nthreads, struct net *net)
 	return 0;
 }
 
+/**
+ * nfsd_set_nrthreads - set the number of running threads in the net's service
+ * @n: number of array members in @nthreads
+ * @nthreads: array of thread counts for each pool
+ * @net: network namespace to operate within
+ *
+ * This function alters the number of running threads for the given network
+ * namespace in each pool. If passed an array longer then the number of pools
+ * the extra pool settings are ignored. If passed an array shorter than the
+ * number of pools, the missing values are interpreted as 0's.
+ *
+ * Returns 0 on success or a negative errno on error.
+ */
 int nfsd_set_nrthreads(int n, int *nthreads, struct net *net)
 {
 	int i = 0;
@@ -716,11 +729,18 @@ int nfsd_set_nrthreads(int n, int *nthreads, struct net *net)
 	int err = 0;
 	struct nfsd_net *nn = net_generic(net, nfsd_net_id);
 
-	WARN_ON(!mutex_is_locked(&nfsd_mutex));
+	lockdep_assert_held(&nfsd_mutex);
 
 	if (nn->nfsd_serv == NULL || n <= 0)
 		return 0;
 
+	/*
+	 * Special case: When n == 1, pass in NULL for the pool, so that the
+	 * change is distributed equally among them.
+	 */
+	if (n == 1)
+		return svc_set_num_threads(nn->nfsd_serv, NULL, nthreads[0]);
+
 	if (n > nn->nfsd_serv->sv_nrpools)
 		n = nn->nfsd_serv->sv_nrpools;
 
@@ -743,31 +763,40 @@ int nfsd_set_nrthreads(int n, int *nthreads, struct net *net)
 		}
 	}
 
-	/*
-	 * There must always be a thread in pool 0; the admin
-	 * can't shut down NFS completely using pool_threads.
-	 */
-	if (nthreads[0] == 0)
-		nthreads[0] = 1;
-
 	/* apply the new numbers */
 	for (i = 0; i < n; i++) {
 		err = svc_set_num_threads(nn->nfsd_serv,
 					  &nn->nfsd_serv->sv_pools[i],
 					  nthreads[i]);
 		if (err)
-			break;
+			goto out;
 	}
+
+	/* Anything undefined in array is considered to be 0 */
+	for (i = n; i < nn->nfsd_serv->sv_nrpools; ++i) {
+		err = svc_set_num_threads(nn->nfsd_serv,
+					  &nn->nfsd_serv->sv_pools[i],
+					  0);
+		if (err)
+			goto out;
+	}
+out:
 	return err;
 }
 
-/*
- * Adjust the number of threads and return the new number of threads.
- * This is also the function that starts the server if necessary, if
- * this is the first time nrservs is nonzero.
+/**
+ * nfsd_svc: start up or shut down the nfsd server
+ * @n: number of array members in @nthreads
+ * @nthreads: array of thread counts for each pool
+ * @net: network namespace to operate within
+ * @cred: credentials to use for xprt creation
+ * @scope: server scope value (defaults to nodename)
+ *
+ * Adjust the number of threads in each pool and return the new
+ * total number of threads in the service.
  */
 int
-nfsd_svc(int nrservs, struct net *net, const struct cred *cred, const char *scope)
+nfsd_svc(int n, int *nthreads, struct net *net, const struct cred *cred, const char *scope)
 {
 	int	error;
 	struct nfsd_net *nn = net_generic(net, nfsd_net_id);
@@ -777,13 +806,6 @@ nfsd_svc(int nrservs, struct net *net, const struct cred *cred, const char *scop
 
 	dprintk("nfsd: creating service\n");
 
-	nrservs = max(nrservs, 0);
-	nrservs = min(nrservs, NFSD_MAXSERVS);
-	error = 0;
-
-	if (nrservs == 0 && nn->nfsd_serv == NULL)
-		goto out;
-
 	strscpy(nn->nfsd_name, scope ? scope : utsname()->nodename,
 		sizeof(nn->nfsd_name));
 
@@ -795,7 +817,7 @@ nfsd_svc(int nrservs, struct net *net, const struct cred *cred, const char *scop
 	error = nfsd_startup_net(net, cred);
 	if (error)
 		goto out_put;
-	error = svc_set_num_threads(serv, NULL, nrservs);
+	error = nfsd_set_nrthreads(n, nthreads, net);
 	if (error)
 		goto out_put;
 	error = serv->sv_nrthreads;