Merge tag 'trace-v7.3-2' of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-tracegrafted

Pull tracing fixes from Steven Rostedt:

 - Fix error output of boot instance creation failure

   Currently if a boot instance creation fails, instead of printing out
   the name of the instance that failed, it prints "(null)". That is
   because it prints "cur_str" that had already been processed by
   strsep(). Print the saved name instead.

   While at it, print the error code of the failure.

 - Fix use-after-free for same named historgrams

   Histograms can be named so that they can be used in multiple events.
   But if the named histogram has a variable attached, the second event
   that uses the named histogram which duplicates it and needs to free
   the original after duplication leaves the old variable in place and
   still visible. If another histogram uses than variable, it will use
   the stale one which will try to reference the freed duplicate
   histogram and crash the kernel.

   Free the duplicate variables along with the duplicated histogram
   data.

 - Check return value of kthread_run() in event self test

   The events self tests uses a kthread for testing but does not check
   if it succeeded in creating a kthread. If the kthread creation were
   to fail, the code will still try to call kthread_stop() on the error
   returned.

 - Fix race between reading trace_pipe and updating subbuffer size

   If a user is reading the trace_pipe file at the same time they update
   the ring buffer sub-buffer size, can cause the trace_pipe read to
   read stale data. Add trace_access_lock() around updating the ring
   buffer sub-buffer size.

 - Fix eventfs_inode on failure path in creation of the events directory

   In the creation of the "events" directory, if after allocating the
   eventfs_inode a failure is detected, it calls cleanup_ei() which
   calls free_ei(). The free_ei() will test if eventfs_inode being freed
   has no children. It is a bug if it does. But on the failure case of
   the creation of the "events" directory, the children lists have not
   yet been initialized and the free will trigger a warning because
   list_empty() on an uninitialized list returns false.

   Move the initialization into init_ei() where it makes more sense and
   makes sure that a created eventfs_inode has its lists initialized
   upon creation.

 - Check return value of kthread_run() in ftrace direct sample code

   The sample code that shows how to use the ftrace direct calls does
   not test the return of kthread_run() to see if it succeeds. Return a
   failure if the kthread_run() doesn't succeed.

 - Clear user events state on fork in case of alloc failure

   On fork, the child gets a pointer to the parent's user events state.
   It makes a copy of it then updates the child's pointer to it. But if
   the allocation fails, the duplication function leaves the child with
   a pointer to its parent's descriptor. When the child cleans up its
   data, it will free the parent's descriptor while the parent is still
   using it.

   In the duplication function, set the child's user_event_mm to NULL
   before testing if the allocation succeeded, and when it exits it will
   not free the parent's descriptor.

 - Fix retry exhaustion in simple ring buffer reader swap

   simple_ring_buffer_swap_reader_page() starts with retry set to 8 and
   post-decrements it only after a failed link replacement. On the final
   attempt, a successful replacement leaves retry at zero, while a
   failed replacement leaves it at -1.

   But the check for success expects the retry value to be non-zero and
   exits with an error on zero. This is the opposite result. Fix it.

 - Fail nicely when the remote swap_reader_page() returns an error

   Currently, if the swap_reader_page() of a remote buffer fails, it
   triggers a WARN_ON_ONCE() and continues normally. Instead, have it
   exit with an error and a pr_warn() print instead of a full WARNING.

* tag 'trace-v7.3-2' of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace:
  ring-buffer: Stop remote reader update when page swap fails
  tracing: Fix retry exhaustion in simple ring buffer reader swap
  tracing/user_events: Clear copied tracing state before fork duplication
  samples/ftrace: Fix kthread_stop() on ERR_PTR in ftrace-direct-multi-modify
  samples/ftrace: Fix kthread_stop() on ERR_PTR in ftrace-direct-modify
  eventfs: Initialize ei->children and ei->list in init_ei()
  tracing: Fix use-after-free in trace_pipe read on sub-buffer order change
  tracing: Fix crash passing ERR_PTR to kthread_stop()
  tracing: Fix use-after-free with same-name named triggers
  tracing: Fix logged instance name on creation failure

1 files changed, 892 insertions, 0 deletions

diff --git a/fs/lockd/clntproc.c b/fs/lockd/clntproc.c
new file mode 100644
index 000000000..f8018bfe9
--- /dev/null
+++ b/fs/lockd/clntproc.c
@@ -0,0 +1,892 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * linux/fs/lockd/clntproc.c
+ *
+ * RPC procedures for the client side NLM implementation
+ *
+ * Copyright (C) 1996, Olaf Kirch <okir@monad.swb.de>
+ */
+
+#include <linux/module.h>
+#include <linux/slab.h>
+#include <linux/types.h>
+#include <linux/errno.h>
+#include <linux/fs.h>
+#include <linux/filelock.h>
+#include <linux/nfs_fs.h>
+#include <linux/utsname.h>
+#include <linux/freezer.h>
+#include <linux/sunrpc/clnt.h>
+#include <linux/sunrpc/svc.h>
+
+#include "lockd.h"
+#include "trace.h"
+
+#define NLMDBG_FACILITY		NLMDBG_CLIENT
+#define NLMCLNT_GRACE_WAIT	(5*HZ)
+#define NLMCLNT_POLL_TIMEOUT	(30*HZ)
+#define NLMCLNT_MAX_RETRIES	3
+
+static int	nlmclnt_test(struct nlm_rqst *, struct file_lock *);
+static int	nlmclnt_lock(struct nlm_rqst *, struct file_lock *);
+static int	nlmclnt_unlock(struct nlm_rqst *, struct file_lock *);
+static int	nlm_stat_to_errno(__be32 stat);
+static void	nlmclnt_locks_init_private(struct file_lock *fl, struct nlm_host *host);
+static int	nlmclnt_cancel(struct nlm_host *, int , struct file_lock *);
+
+static const struct rpc_call_ops nlmclnt_unlock_ops;
+static const struct rpc_call_ops nlmclnt_cancel_ops;
+
+/*
+ * Cookie counter for NLM requests
+ */
+static atomic_t	nlm_cookie = ATOMIC_INIT(0x1234);
+
+void nlmclnt_next_cookie(struct lockd_cookie *c)
+{
+	u32	cookie = atomic_inc_return(&nlm_cookie);
+
+	memcpy(c->data, &cookie, 4);
+	c->len=4;
+}
+
+static struct nlm_lockowner *
+nlmclnt_get_lockowner(struct nlm_lockowner *lockowner)
+{
+	refcount_inc(&lockowner->count);
+	return lockowner;
+}
+
+static void nlmclnt_put_lockowner(struct nlm_lockowner *lockowner)
+{
+	if (!refcount_dec_and_lock(&lockowner->count, &lockowner->host->h_lock))
+		return;
+	list_del(&lockowner->list);
+	spin_unlock(&lockowner->host->h_lock);
+	nlmclnt_release_host(lockowner->host);
+	kfree(lockowner);
+}
+
+static inline int nlm_pidbusy(struct nlm_host *host, uint32_t pid)
+{
+	struct nlm_lockowner *lockowner;
+	list_for_each_entry(lockowner, &host->h_lockowners, list) {
+		if (lockowner->pid == pid)
+			return -EBUSY;
+	}
+	return 0;
+}
+
+static inline uint32_t __nlm_alloc_pid(struct nlm_host *host)
+{
+	uint32_t res;
+	do {
+		res = host->h_pidcount++;
+	} while (nlm_pidbusy(host, res) < 0);
+	return res;
+}
+
+static struct nlm_lockowner *__nlmclnt_find_lockowner(struct nlm_host *host, fl_owner_t owner)
+{
+	struct nlm_lockowner *lockowner;
+	list_for_each_entry(lockowner, &host->h_lockowners, list) {
+		if (lockowner->owner != owner)
+			continue;
+		return nlmclnt_get_lockowner(lockowner);
+	}
+	return NULL;
+}
+
+static struct nlm_lockowner *nlmclnt_find_lockowner(struct nlm_host *host, fl_owner_t owner)
+{
+	struct nlm_lockowner *res, *new = NULL;
+
+	spin_lock(&host->h_lock);
+	res = __nlmclnt_find_lockowner(host, owner);
+	if (res == NULL) {
+		spin_unlock(&host->h_lock);
+		new = kmalloc_obj(*new);
+		spin_lock(&host->h_lock);
+		res = __nlmclnt_find_lockowner(host, owner);
+		if (res == NULL && new != NULL) {
+			res = new;
+			refcount_set(&new->count, 1);
+			new->owner = owner;
+			new->pid = __nlm_alloc_pid(host);
+			new->host = nlm_get_host(host);
+			list_add(&new->list, &host->h_lockowners);
+			new = NULL;
+		}
+	}
+	spin_unlock(&host->h_lock);
+	kfree(new);
+	return res;
+}
+
+/*
+ * Initialize arguments for TEST/LOCK/UNLOCK/CANCEL calls
+ */
+static void nlmclnt_setlockargs(struct nlm_rqst *req, struct file_lock *fl)
+{
+	struct lockd_args *argp = &req->a_args;
+	struct lockd_lock *lock = &argp->lock;
+	char *nodename = req->a_host->h_rpcclnt->cl_nodename;
+
+	nlmclnt_next_cookie(&argp->cookie);
+	memcpy(&lock->fh, NFS_FH(file_inode(fl->c.flc_file)),
+	       sizeof(struct nfs_fh));
+	lock->caller  = nodename;
+	lock->oh.data = req->a_owner;
+	lock->oh.len  = snprintf(req->a_owner, sizeof(req->a_owner), "%u@%s",
+				(unsigned int)fl->fl_u.nfs_fl.owner->pid,
+				nodename);
+	lock->svid = fl->fl_u.nfs_fl.owner->pid;
+	lock->fl.fl_start = fl->fl_start;
+	lock->fl.fl_end = fl->fl_end;
+	lock->fl.c.flc_type = fl->c.flc_type;
+}
+
+static void nlmclnt_release_lockargs(struct nlm_rqst *req)
+{
+	WARN_ON_ONCE(req->a_args.lock.fl.fl_ops != NULL);
+}
+
+/**
+ * nlmclnt_proc - Perform a single client-side lock request
+ * @host: address of a valid nlm_host context representing the NLM server
+ * @cmd: fcntl-style file lock operation to perform
+ * @fl: address of arguments for the lock operation
+ * @data: address of data to be sent to callback operations
+ *
+ */
+int nlmclnt_proc(struct nlm_host *host, int cmd, struct file_lock *fl, void *data)
+{
+	struct nlm_rqst		*call;
+	int			status;
+	const struct nlmclnt_operations *nlmclnt_ops = host->h_nlmclnt_ops;
+
+	call = nlm_alloc_call(host);
+	if (call == NULL)
+		return -ENOMEM;
+
+	if (nlmclnt_ops && nlmclnt_ops->nlmclnt_alloc_call)
+		nlmclnt_ops->nlmclnt_alloc_call(data);
+
+	nlmclnt_locks_init_private(fl, host);
+	if (!fl->fl_u.nfs_fl.owner) {
+		/* lockowner allocation has failed */
+		nlmclnt_release_call(call);
+		return -ENOMEM;
+	}
+	/* Set up the argument struct */
+	nlmclnt_setlockargs(call, fl);
+	call->a_callback_data = data;
+
+	if (IS_SETLK(cmd) || IS_SETLKW(cmd)) {
+		if (fl->c.flc_type != F_UNLCK) {
+			call->a_args.block = IS_SETLKW(cmd) ? 1 : 0;
+			status = nlmclnt_lock(call, fl);
+		} else
+			status = nlmclnt_unlock(call, fl);
+	} else if (IS_GETLK(cmd))
+		status = nlmclnt_test(call, fl);
+	else
+		status = -EINVAL;
+	fl->fl_ops->fl_release_private(fl);
+	fl->fl_ops = NULL;
+
+	dprintk("lockd: clnt proc returns %d\n", status);
+	return status;
+}
+EXPORT_SYMBOL_GPL(nlmclnt_proc);
+
+/*
+ * Allocate an NLM RPC call struct
+ */
+struct nlm_rqst *nlm_alloc_call(struct nlm_host *host)
+{
+	struct nlm_rqst	*call;
+
+	for(;;) {
+		call = kzalloc_obj(*call);
+		if (call != NULL) {
+			refcount_set(&call->a_count, 1);
+			locks_init_lock(&call->a_args.lock.fl);
+			locks_init_lock(&call->a_res.lock.fl);
+			call->a_host = nlm_get_host(host);
+			return call;
+		}
+		if (signalled())
+			break;
+		printk("nlm_alloc_call: failed, waiting for memory\n");
+		schedule_timeout_interruptible(5*HZ);
+	}
+	return NULL;
+}
+
+void nlmclnt_release_call(struct nlm_rqst *call)
+{
+	const struct nlmclnt_operations *nlmclnt_ops = call->a_host->h_nlmclnt_ops;
+
+	if (!refcount_dec_and_test(&call->a_count))
+		return;
+	if (nlmclnt_ops && nlmclnt_ops->nlmclnt_release_call)
+		nlmclnt_ops->nlmclnt_release_call(call->a_callback_data);
+	nlmclnt_release_host(call->a_host);
+	nlmclnt_release_lockargs(call);
+	kfree(call);
+}
+
+static void nlmclnt_rpc_release(void *data)
+{
+	nlmclnt_release_call(data);
+}
+
+static int nlm_wait_on_grace(wait_queue_head_t *queue)
+{
+	DEFINE_WAIT(wait);
+	int status = -EINTR;
+
+	prepare_to_wait(queue, &wait, TASK_INTERRUPTIBLE);
+	if (!signalled ()) {
+		schedule_timeout(NLMCLNT_GRACE_WAIT);
+		try_to_freeze();
+		if (!signalled ())
+			status = 0;
+	}
+	finish_wait(queue, &wait);
+	return status;
+}
+
+/*
+ * Generic NLM call
+ */
+static int
+nlmclnt_call(const struct cred *cred, struct nlm_rqst *req, u32 proc)
+{
+	struct nlm_host	*host = req->a_host;
+	struct rpc_clnt	*clnt;
+	struct lockd_args *argp = &req->a_args;
+	struct lockd_res *resp = &req->a_res;
+	struct rpc_message msg = {
+		.rpc_argp	= argp,
+		.rpc_resp	= resp,
+		.rpc_cred	= cred,
+	};
+	int		status;
+
+	dprintk("lockd: call procedure %d on %s\n",
+			(int)proc, host->h_name);
+
+	do {
+		if (host->h_reclaiming && !argp->reclaim)
+			goto in_grace_period;
+
+		/* If we have no RPC client yet, create one. */
+		if ((clnt = nlm_bind_host(host)) == NULL)
+			return -ENOLCK;
+		msg.rpc_proc = &clnt->cl_procinfo[proc];
+
+		/* Perform the RPC call. If an error occurs, try again */
+		if ((status = rpc_call_sync(clnt, &msg, 0)) < 0) {
+			dprintk("lockd: rpc_call returned error %d\n", -status);
+			switch (status) {
+			case -EPROTONOSUPPORT:
+				status = -EINVAL;
+				break;
+			case -ECONNREFUSED:
+			case -ETIMEDOUT:
+			case -ENOTCONN:
+				nlm_rebind_host(host);
+				status = -EAGAIN;
+				break;
+			case -ERESTARTSYS:
+				return signalled () ? -EINTR : status;
+			default:
+				break;
+			}
+			break;
+		} else
+		if (resp->status == nlm_lck_denied_grace_period) {
+			dprintk("lockd: server in grace period\n");
+			if (argp->reclaim) {
+				printk(KERN_WARNING
+				     "lockd: spurious grace period reject?!\n");
+				return -ENOLCK;
+			}
+		} else {
+			if (!argp->reclaim) {
+				/* We appear to be out of the grace period */
+				wake_up_all(&host->h_gracewait);
+			}
+			dprintk("lockd: server returns status %d\n",
+				ntohl(resp->status));
+			return 0;	/* Okay, call complete */
+		}
+
+in_grace_period:
+		/*
+		 * The server has rebooted and appears to be in the grace
+		 * period during which locks are only allowed to be
+		 * reclaimed.
+		 * We can only back off and try again later.
+		 */
+		status = nlm_wait_on_grace(&host->h_gracewait);
+	} while (status == 0);
+
+	return status;
+}
+
+/*
+ * Generic NLM call, async version.
+ */
+static struct rpc_task *__nlm_async_call(struct nlm_rqst *req, u32 proc, struct rpc_message *msg, const struct rpc_call_ops *tk_ops)
+{
+	struct nlm_host	*host = req->a_host;
+	struct rpc_clnt	*clnt;
+	struct rpc_task_setup task_setup_data = {
+		.rpc_message = msg,
+		.callback_ops = tk_ops,
+		.callback_data = req,
+		.flags = RPC_TASK_ASYNC,
+	};
+
+	dprintk("lockd: call procedure %d on %s (async)\n",
+			(int)proc, host->h_name);
+
+	/* If we have no RPC client yet, create one. */
+	clnt = nlm_bind_host(host);
+	if (clnt == NULL)
+		goto out_err;
+	msg->rpc_proc = &clnt->cl_procinfo[proc];
+	task_setup_data.rpc_client = clnt;
+
+        /* bootstrap and kick off the async RPC call */
+	return rpc_run_task(&task_setup_data);
+out_err:
+	tk_ops->rpc_release(req);
+	return ERR_PTR(-ENOLCK);
+}
+
+static int nlm_do_async_call(struct nlm_rqst *req, u32 proc, struct rpc_message *msg, const struct rpc_call_ops *tk_ops)
+{
+	struct rpc_task *task;
+
+	task = __nlm_async_call(req, proc, msg, tk_ops);
+	if (IS_ERR(task))
+		return PTR_ERR(task);
+	rpc_put_task(task);
+	return 0;
+}
+
+/*
+ * NLM asynchronous call.
+ */
+int nlm_async_call(struct nlm_rqst *req, u32 proc, const struct rpc_call_ops *tk_ops)
+{
+	struct rpc_message msg = {
+		.rpc_argp	= &req->a_args,
+		.rpc_resp	= &req->a_res,
+	};
+	return nlm_do_async_call(req, proc, &msg, tk_ops);
+}
+
+int nlm_async_reply(struct nlm_rqst *req, u32 proc, const struct rpc_call_ops *tk_ops)
+{
+	struct rpc_message msg = {
+		.rpc_argp	= &req->a_res,
+	};
+	return nlm_do_async_call(req, proc, &msg, tk_ops);
+}
+
+/*
+ * NLM client asynchronous call.
+ *
+ * Note that although the calls are asynchronous, and are therefore
+ *      guaranteed to complete, we still always attempt to wait for
+ *      completion in order to be able to correctly track the lock
+ *      state.
+ */
+static int nlmclnt_async_call(const struct cred *cred, struct nlm_rqst *req, u32 proc, const struct rpc_call_ops *tk_ops)
+{
+	struct rpc_message msg = {
+		.rpc_argp	= &req->a_args,
+		.rpc_resp	= &req->a_res,
+		.rpc_cred	= cred,
+	};
+	struct rpc_task *task;
+	int err;
+
+	task = __nlm_async_call(req, proc, &msg, tk_ops);
+	if (IS_ERR(task))
+		return PTR_ERR(task);
+	err = rpc_wait_for_completion_task(task);
+	rpc_put_task(task);
+	return err;
+}
+
+/*
+ * TEST for the presence of a conflicting lock
+ */
+static int
+nlmclnt_test(struct nlm_rqst *req, struct file_lock *fl)
+{
+	int	status;
+
+	status = nlmclnt_call(nfs_file_cred(fl->c.flc_file), req,
+			      NLMPROC_TEST);
+	if (status < 0)
+		goto out;
+
+	switch (req->a_res.status) {
+		case nlm_granted:
+			fl->c.flc_type = F_UNLCK;
+			break;
+		case nlm_lck_denied:
+			/*
+			 * Report the conflicting lock back to the application.
+			 */
+			fl->fl_start = req->a_res.lock.fl.fl_start;
+			fl->fl_end = req->a_res.lock.fl.fl_end;
+			fl->c.flc_type = req->a_res.lock.fl.c.flc_type;
+			fl->c.flc_pid = -req->a_res.lock.fl.c.flc_pid;
+			break;
+		default:
+			status = nlm_stat_to_errno(req->a_res.status);
+	}
+out:
+	trace_nlmclnt_test(&req->a_args.lock,
+			   (const struct sockaddr *)&req->a_host->h_addr,
+			   req->a_host->h_addrlen, req->a_res.status);
+	nlmclnt_release_call(req);
+	return status;
+}
+
+static void nlmclnt_locks_copy_lock(struct file_lock *new, struct file_lock *fl)
+{
+	spin_lock(&fl->fl_u.nfs_fl.owner->host->h_lock);
+	new->fl_u.nfs_fl.state = fl->fl_u.nfs_fl.state;
+	new->fl_u.nfs_fl.owner = nlmclnt_get_lockowner(fl->fl_u.nfs_fl.owner);
+	list_add_tail(&new->fl_u.nfs_fl.list, &fl->fl_u.nfs_fl.owner->host->h_granted);
+	spin_unlock(&fl->fl_u.nfs_fl.owner->host->h_lock);
+}
+
+static void nlmclnt_locks_release_private(struct file_lock *fl)
+{
+	spin_lock(&fl->fl_u.nfs_fl.owner->host->h_lock);
+	list_del(&fl->fl_u.nfs_fl.list);
+	spin_unlock(&fl->fl_u.nfs_fl.owner->host->h_lock);
+	nlmclnt_put_lockowner(fl->fl_u.nfs_fl.owner);
+}
+
+static const struct file_lock_operations nlmclnt_lock_ops = {
+	.fl_copy_lock = nlmclnt_locks_copy_lock,
+	.fl_release_private = nlmclnt_locks_release_private,
+};
+
+static void nlmclnt_locks_init_private(struct file_lock *fl, struct nlm_host *host)
+{
+	fl->fl_u.nfs_fl.state = 0;
+	fl->fl_ops = NULL;
+	fl->fl_u.nfs_fl.owner = nlmclnt_find_lockowner(host,
+						       fl->c.flc_owner);
+	INIT_LIST_HEAD(&fl->fl_u.nfs_fl.list);
+	if (!fl->fl_u.nfs_fl.owner)
+		return;
+	fl->fl_ops = &nlmclnt_lock_ops;
+}
+
+static int do_vfs_lock(struct file_lock *fl)
+{
+	return locks_lock_file_wait(fl->c.flc_file, fl);
+}
+
+/*
+ * LOCK: Try to create a lock
+ *
+ *			Programmer Harassment Alert
+ *
+ * When given a blocking lock request in a sync RPC call, the HPUX lockd
+ * will faithfully return LCK_BLOCKED but never cares to notify us when
+ * the lock could be granted. This way, our local process could hang
+ * around forever waiting for the callback.
+ *
+ *  Solution A:	Implement busy-waiting
+ *  Solution B: Use the async version of the call (NLM_LOCK_{MSG,RES})
+ *
+ * For now I am implementing solution A, because I hate the idea of
+ * re-implementing lockd for a third time in two months. The async
+ * calls shouldn't be too hard to do, however.
+ *
+ * This is one of the lovely things about standards in the NFS area:
+ * they're so soft and squishy you can't really blame HP for doing this.
+ */
+static int
+nlmclnt_lock(struct nlm_rqst *req, struct file_lock *fl)
+{
+	const struct cred *cred = nfs_file_cred(fl->c.flc_file);
+	struct nlm_host	*host = req->a_host;
+	struct lockd_res *resp = &req->a_res;
+	struct nlm_wait block;
+	unsigned char flags = fl->c.flc_flags;
+	unsigned char type;
+	__be32 b_status;
+	int status = -ENOLCK;
+
+	if (nsm_monitor(host) < 0)
+		goto out;
+	req->a_args.state = nsm_local_state;
+
+	fl->c.flc_flags |= FL_ACCESS;
+	status = do_vfs_lock(fl);
+	fl->c.flc_flags = flags;
+	if (status < 0)
+		goto out;
+
+	nlmclnt_prepare_block(&block, host, fl);
+again:
+	/*
+	 * Initialise resp->status to a valid non-zero value,
+	 * since 0 == nlm_lck_granted
+	 */
+	resp->status = nlm_lck_blocked;
+
+	/*
+	 * A GRANTED callback can come at any time -- even before the reply
+	 * to the LOCK request arrives, so we queue the wait before
+	 * requesting the lock.
+	 */
+	nlmclnt_queue_block(&block);
+	for (;;) {
+		/* Reboot protection */
+		fl->fl_u.nfs_fl.state = host->h_state;
+		status = nlmclnt_call(cred, req, NLMPROC_LOCK);
+		if (status < 0)
+			break;
+		/* Did a reclaimer thread notify us of a server reboot? */
+		if (resp->status == nlm_lck_denied_grace_period)
+			continue;
+		if (resp->status != nlm_lck_blocked)
+			break;
+		/* Wait on an NLM blocking lock */
+		status = nlmclnt_wait(&block, req, NLMCLNT_POLL_TIMEOUT);
+		if (status < 0)
+			break;
+		if (block.b_status != nlm_lck_blocked)
+			break;
+	}
+	b_status = nlmclnt_dequeue_block(&block);
+	if (resp->status == nlm_lck_blocked)
+		resp->status = b_status;
+
+	/* if we were interrupted while blocking, then cancel the lock request
+	 * and exit
+	 */
+	if (resp->status == nlm_lck_blocked) {
+		if (!req->a_args.block)
+			goto out_unlock;
+		if (nlmclnt_cancel(host, req->a_args.block, fl) == 0)
+			goto out;
+	}
+
+	if (resp->status == nlm_granted) {
+		down_read(&host->h_rwsem);
+		/* Check whether or not the server has rebooted */
+		if (fl->fl_u.nfs_fl.state != host->h_state) {
+			up_read(&host->h_rwsem);
+			goto again;
+		}
+		/* Ensure the resulting lock will get added to granted list */
+		fl->c.flc_flags |= FL_SLEEP;
+		if (do_vfs_lock(fl) < 0)
+			printk(KERN_WARNING "%s: VFS is out of sync with lock manager!\n", __func__);
+		up_read(&host->h_rwsem);
+		fl->c.flc_flags = flags;
+		status = 0;
+	}
+	if (status < 0)
+		goto out_unlock;
+	/*
+	 * EAGAIN doesn't make sense for sleeping locks, and in some
+	 * cases NLM_LCK_DENIED is returned for a permanent error.  So
+	 * turn it into an ENOLCK.
+	 */
+	if (resp->status == nlm_lck_denied && (flags & FL_SLEEP))
+		status = -ENOLCK;
+	else
+		status = nlm_stat_to_errno(resp->status);
+out:
+	trace_nlmclnt_lock(&req->a_args.lock,
+			   (const struct sockaddr *)&req->a_host->h_addr,
+			   req->a_host->h_addrlen, req->a_res.status);
+	nlmclnt_release_call(req);
+	return status;
+out_unlock:
+	/* Fatal error: ensure that we remove the lock altogether */
+	trace_nlmclnt_lock(&req->a_args.lock,
+			   (const struct sockaddr *)&req->a_host->h_addr,
+			   req->a_host->h_addrlen, req->a_res.status);
+	dprintk("lockd: lock attempt ended in fatal error.\n"
+		"       Attempting to unlock.\n");
+	type = fl->c.flc_type;
+	fl->c.flc_type = F_UNLCK;
+	down_read(&host->h_rwsem);
+	do_vfs_lock(fl);
+	up_read(&host->h_rwsem);
+	fl->c.flc_type = type;
+	fl->c.flc_flags = flags;
+	nlmclnt_async_call(cred, req, NLMPROC_UNLOCK, &nlmclnt_unlock_ops);
+	return status;
+}
+
+/*
+ * RECLAIM: Try to reclaim a lock
+ */
+int
+nlmclnt_reclaim(struct nlm_host *host, struct file_lock *fl,
+		struct nlm_rqst *req)
+{
+	int		status;
+
+	memset(req, 0, sizeof(*req));
+	locks_init_lock(&req->a_args.lock.fl);
+	locks_init_lock(&req->a_res.lock.fl);
+	req->a_host  = host;
+
+	/* Set up the argument struct */
+	nlmclnt_setlockargs(req, fl);
+	req->a_args.reclaim = 1;
+
+	status = nlmclnt_call(nfs_file_cred(fl->c.flc_file), req,
+			      NLMPROC_LOCK);
+	if (status >= 0 && req->a_res.status == nlm_granted)
+		return 0;
+
+	printk(KERN_WARNING "lockd: failed to reclaim lock for pid %d "
+				"(errno %d, status %d)\n",
+				fl->c.flc_pid,
+				status, ntohl(req->a_res.status));
+
+	/*
+	 * FIXME: This is a serious failure. We can
+	 *
+	 *  a.	Ignore the problem
+	 *  b.	Send the owning process some signal (Linux doesn't have
+	 *	SIGLOST, though...)
+	 *  c.	Retry the operation
+	 *
+	 * Until someone comes up with a simple implementation
+	 * for b or c, I'll choose option a.
+	 */
+
+	return -ENOLCK;
+}
+
+/*
+ * UNLOCK: remove an existing lock
+ */
+static int
+nlmclnt_unlock(struct nlm_rqst *req, struct file_lock *fl)
+{
+	struct nlm_host	*host = req->a_host;
+	struct lockd_res *resp = &req->a_res;
+	int status;
+	unsigned char flags = fl->c.flc_flags;
+
+	/*
+	 * Note: the server is supposed to either grant us the unlock
+	 * request, or to deny it with NLM_LCK_DENIED_GRACE_PERIOD. In either
+	 * case, we want to unlock.
+	 */
+	fl->c.flc_flags |= FL_EXISTS;
+	down_read(&host->h_rwsem);
+	status = do_vfs_lock(fl);
+	up_read(&host->h_rwsem);
+	fl->c.flc_flags = flags;
+	if (status == -ENOENT) {
+		status = 0;
+		goto out;
+	}
+
+	refcount_inc(&req->a_count);
+	status = nlmclnt_async_call(nfs_file_cred(fl->c.flc_file), req,
+				    NLMPROC_UNLOCK, &nlmclnt_unlock_ops);
+	if (status < 0)
+		goto out;
+
+	if (resp->status == nlm_granted)
+		goto out;
+
+	if (resp->status != nlm_lck_denied_nolocks)
+		printk("lockd: unexpected unlock status: %d\n",
+			ntohl(resp->status));
+	/* What to do now? I'm out of my depth... */
+	status = -ENOLCK;
+out:
+	trace_nlmclnt_unlock(&req->a_args.lock,
+			     (const struct sockaddr *)&req->a_host->h_addr,
+			     req->a_host->h_addrlen, req->a_res.status);
+	nlmclnt_release_call(req);
+	return status;
+}
+
+static void nlmclnt_unlock_prepare(struct rpc_task *task, void *data)
+{
+	struct nlm_rqst	*req = data;
+	const struct nlmclnt_operations *nlmclnt_ops = req->a_host->h_nlmclnt_ops;
+	bool defer_call = false;
+
+	if (nlmclnt_ops && nlmclnt_ops->nlmclnt_unlock_prepare)
+		defer_call = nlmclnt_ops->nlmclnt_unlock_prepare(task, req->a_callback_data);
+
+	if (!defer_call)
+		rpc_call_start(task);
+}
+
+static void nlmclnt_unlock_callback(struct rpc_task *task, void *data)
+{
+	struct nlm_rqst	*req = data;
+	u32 status = ntohl(req->a_res.status);
+
+	if (RPC_SIGNALLED(task))
+		goto die;
+
+	if (task->tk_status < 0) {
+		dprintk("lockd: unlock failed (err = %d)\n", -task->tk_status);
+		switch (task->tk_status) {
+		case -EACCES:
+		case -EIO:
+			goto die;
+		default:
+			goto retry_rebind;
+		}
+	}
+	if (status == NLM_LCK_DENIED_GRACE_PERIOD) {
+		rpc_delay(task, NLMCLNT_GRACE_WAIT);
+		goto retry_unlock;
+	}
+	if (status != NLM_LCK_GRANTED)
+		printk(KERN_WARNING "lockd: unexpected unlock status: %d\n", status);
+die:
+	return;
+ retry_rebind:
+	nlm_rebind_host(req->a_host);
+ retry_unlock:
+	rpc_restart_call(task);
+}
+
+static const struct rpc_call_ops nlmclnt_unlock_ops = {
+	.rpc_call_prepare = nlmclnt_unlock_prepare,
+	.rpc_call_done = nlmclnt_unlock_callback,
+	.rpc_release = nlmclnt_rpc_release,
+};
+
+/*
+ * Cancel a blocked lock request.
+ * We always use an async RPC call for this in order not to hang a
+ * process that has been Ctrl-C'ed.
+ */
+static int nlmclnt_cancel(struct nlm_host *host, int block, struct file_lock *fl)
+{
+	struct nlm_rqst	*req;
+	int status;
+
+	dprintk("lockd: blocking lock attempt was interrupted by a signal.\n"
+		"       Attempting to cancel lock.\n");
+
+	req = nlm_alloc_call(host);
+	if (!req)
+		return -ENOMEM;
+	req->a_flags = RPC_TASK_ASYNC;
+
+	nlmclnt_setlockargs(req, fl);
+	req->a_args.block = block;
+
+	refcount_inc(&req->a_count);
+	status = nlmclnt_async_call(nfs_file_cred(fl->c.flc_file), req,
+				    NLMPROC_CANCEL, &nlmclnt_cancel_ops);
+	if (status == 0 && req->a_res.status == nlm_lck_denied)
+		status = -ENOLCK;
+	nlmclnt_release_call(req);
+	return status;
+}
+
+static void nlmclnt_cancel_callback(struct rpc_task *task, void *data)
+{
+	struct nlm_rqst	*req = data;
+	u32 status = ntohl(req->a_res.status);
+
+	if (RPC_SIGNALLED(task))
+		goto die;
+
+	if (task->tk_status < 0) {
+		dprintk("lockd: CANCEL call error %d, retrying.\n",
+					task->tk_status);
+		goto retry_cancel;
+	}
+
+	switch (status) {
+	case NLM_LCK_GRANTED:
+	case NLM_LCK_DENIED_GRACE_PERIOD:
+	case NLM_LCK_DENIED:
+		/* Everything's good */
+		break;
+	case NLM_LCK_DENIED_NOLOCKS:
+		dprintk("lockd: CANCEL failed (server has no locks)\n");
+		goto retry_cancel;
+	default:
+		printk(KERN_NOTICE "lockd: weird return %d for CANCEL call\n",
+			status);
+	}
+
+die:
+	return;
+
+retry_cancel:
+	/* Don't ever retry more than 3 times */
+	if (req->a_retries++ >= NLMCLNT_MAX_RETRIES)
+		goto die;
+	nlm_rebind_host(req->a_host);
+	rpc_restart_call(task);
+	rpc_delay(task, 30 * HZ);
+}
+
+static const struct rpc_call_ops nlmclnt_cancel_ops = {
+	.rpc_call_done = nlmclnt_cancel_callback,
+	.rpc_release = nlmclnt_rpc_release,
+};
+
+/*
+ * Convert an NLM status code to a generic kernel errno
+ */
+static int
+nlm_stat_to_errno(__be32 status)
+{
+	switch(ntohl(status)) {
+	case NLM_LCK_GRANTED:
+		return 0;
+	case NLM_LCK_DENIED:
+		return -EAGAIN;
+	case NLM_LCK_DENIED_NOLOCKS:
+	case NLM_LCK_DENIED_GRACE_PERIOD:
+		return -ENOLCK;
+	case NLM_LCK_BLOCKED:
+		printk(KERN_NOTICE "lockd: unexpected status NLM_BLOCKED\n");
+		return -ENOLCK;
+#ifdef CONFIG_LOCKD_V4
+	case NLM_DEADLCK:
+		return -EDEADLK;
+	case NLM_ROFS:
+		return -EROFS;
+	case NLM_STALE_FH:
+		return -ESTALE;
+	case NLM_FBIG:
+		return -EOVERFLOW;
+	case NLM_FAILED:
+		return -ENOLCK;
+#endif
+	}
+	printk(KERN_NOTICE "lockd: unexpected server status %d\n",
+		 ntohl(status));
+	return -ENOLCK;
+}