/*
* Copyright (c) 2004 The Regents of the University of Michigan.
+* Copyright (c) 2012 Jeff Layton <jlayton@redhat.com>
* All rights reserved.
*
* Andy Adamson <andros@citi.umich.edu>
#include <linux/namei.h>
#include <linux/crypto.h>
#include <linux/sched.h>
+#include <linux/fs.h>
+#include <linux/module.h>
+#include <net/net_namespace.h>
+#include <linux/sunrpc/rpc_pipe_fs.h>
+#include <linux/sunrpc/clnt.h>
+#include <linux/nfsd/cld.h>
#include "nfsd.h"
#include "state.h"
#include "vfs.h"
+#include "netns.h"
#define NFSDDBG_FACILITY NFSDDBG_PROC
+/* Declarations */
+struct nfsd4_client_tracking_ops {
+ int (*init)(struct net *);
+ void (*exit)(struct net *);
+ void (*create)(struct nfs4_client *);
+ void (*remove)(struct nfs4_client *);
+ int (*check)(struct nfs4_client *);
+ void (*grace_done)(struct net *, time_t);
+};
+
/* Globals */
static struct file *rec_file;
static char user_recovery_dirname[PATH_MAX] = "/var/lib/nfs/v4recovery";
+static struct nfsd4_client_tracking_ops *client_tracking_ops;
static int
nfs4_save_creds(const struct cred **original_creds)
return status;
}
-void nfsd4_create_clid_dir(struct nfs4_client *clp)
+static void
+nfsd4_create_clid_dir(struct nfs4_client *clp)
{
const struct cred *original_cred;
char *dname = clp->cl_recdir;
dprintk("NFSD: nfsd4_create_clid_dir for \"%s\"\n", dname);
- if (clp->cl_firststate)
+ if (test_and_set_bit(NFSD4_CLIENT_STABLE, &clp->cl_flags))
return;
- clp->cl_firststate = 1;
if (!rec_file)
return;
status = nfs4_save_creds(&original_cred);
return status;
}
-void
+static void
nfsd4_remove_clid_dir(struct nfs4_client *clp)
{
const struct cred *original_cred;
int status;
- if (!rec_file || !clp->cl_firststate)
+ if (!rec_file || !test_bit(NFSD4_CLIENT_STABLE, &clp->cl_flags))
return;
status = mnt_want_write_file(rec_file);
if (status)
goto out;
- clp->cl_firststate = 0;
+ clear_bit(NFSD4_CLIENT_STABLE, &clp->cl_flags);
status = nfs4_save_creds(&original_cred);
if (status < 0)
if (status)
printk("NFSD: Failed to remove expired client state directory"
" %.*s\n", HEXDIR_LEN, clp->cl_recdir);
- return;
}
static int
return 0;
}
-void
-nfsd4_recdir_purge_old(void) {
+static void
+nfsd4_recdir_purge_old(struct net *net, time_t boot_time)
+{
int status;
if (!rec_file)
return 0;
}
-int
+static int
nfsd4_recdir_load(void) {
int status;
* Hold reference to the recovery directory.
*/
-void
-nfsd4_init_recdir()
+static int
+nfsd4_init_recdir(void)
{
const struct cred *original_cred;
int status;
printk("NFSD: Unable to change credentials to find recovery"
" directory: error %d\n",
status);
- return;
+ return status;
}
rec_file = filp_open(user_recovery_dirname, O_RDONLY | O_DIRECTORY, 0);
if (IS_ERR(rec_file)) {
printk("NFSD: unable to find recovery directory %s\n",
user_recovery_dirname);
+ status = PTR_ERR(rec_file);
rec_file = NULL;
}
nfs4_reset_creds(original_cred);
+ return status;
}
-void
+static int
+nfsd4_load_reboot_recovery_data(struct net *net)
+{
+ int status;
+
+ /* XXX: The legacy code won't work in a container */
+ if (net != &init_net) {
+ WARN(1, KERN_ERR "NFSD: attempt to initialize legacy client "
+ "tracking in a container!\n");
+ return -EINVAL;
+ }
+
+ nfs4_lock_state();
+ status = nfsd4_init_recdir();
+ if (!status)
+ status = nfsd4_recdir_load();
+ nfs4_unlock_state();
+ if (status)
+ printk(KERN_ERR "NFSD: Failure reading reboot recovery data\n");
+ return status;
+}
+
+static void
nfsd4_shutdown_recdir(void)
{
if (!rec_file)
rec_file = NULL;
}
+static void
+nfsd4_legacy_tracking_exit(struct net *net)
+{
+ nfs4_release_reclaim();
+ nfsd4_shutdown_recdir();
+}
+
/*
* Change the NFSv4 recovery directory to recdir.
*/
{
return user_recovery_dirname;
}
+
+static int
+nfsd4_check_legacy_client(struct nfs4_client *clp)
+{
+ /* did we already find that this client is stable? */
+ if (test_bit(NFSD4_CLIENT_STABLE, &clp->cl_flags))
+ return 0;
+
+ /* look for it in the reclaim hashtable otherwise */
+ if (nfsd4_find_reclaim_client(clp)) {
+ set_bit(NFSD4_CLIENT_STABLE, &clp->cl_flags);
+ return 0;
+ }
+
+ return -ENOENT;
+}
+
+static struct nfsd4_client_tracking_ops nfsd4_legacy_tracking_ops = {
+ .init = nfsd4_load_reboot_recovery_data,
+ .exit = nfsd4_legacy_tracking_exit,
+ .create = nfsd4_create_clid_dir,
+ .remove = nfsd4_remove_clid_dir,
+ .check = nfsd4_check_legacy_client,
+ .grace_done = nfsd4_recdir_purge_old,
+};
+
+/* Globals */
+#define NFSD_PIPE_DIR "nfsd"
+#define NFSD_CLD_PIPE "cld"
+
+/* per-net-ns structure for holding cld upcall info */
+struct cld_net {
+ struct rpc_pipe *cn_pipe;
+ spinlock_t cn_lock;
+ struct list_head cn_list;
+ unsigned int cn_xid;
+};
+
+struct cld_upcall {
+ struct list_head cu_list;
+ struct cld_net *cu_net;
+ struct task_struct *cu_task;
+ struct cld_msg cu_msg;
+};
+
+static int
+__cld_pipe_upcall(struct rpc_pipe *pipe, struct cld_msg *cmsg)
+{
+ int ret;
+ struct rpc_pipe_msg msg;
+
+ memset(&msg, 0, sizeof(msg));
+ msg.data = cmsg;
+ msg.len = sizeof(*cmsg);
+
+ /*
+ * Set task state before we queue the upcall. That prevents
+ * wake_up_process in the downcall from racing with schedule.
+ */
+ set_current_state(TASK_UNINTERRUPTIBLE);
+ ret = rpc_queue_upcall(pipe, &msg);
+ if (ret < 0) {
+ set_current_state(TASK_RUNNING);
+ goto out;
+ }
+
+ schedule();
+ set_current_state(TASK_RUNNING);
+
+ if (msg.errno < 0)
+ ret = msg.errno;
+out:
+ return ret;
+}
+
+static int
+cld_pipe_upcall(struct rpc_pipe *pipe, struct cld_msg *cmsg)
+{
+ int ret;
+
+ /*
+ * -EAGAIN occurs when pipe is closed and reopened while there are
+ * upcalls queued.
+ */
+ do {
+ ret = __cld_pipe_upcall(pipe, cmsg);
+ } while (ret == -EAGAIN);
+
+ return ret;
+}
+
+static ssize_t
+cld_pipe_downcall(struct file *filp, const char __user *src, size_t mlen)
+{
+ struct cld_upcall *tmp, *cup;
+ struct cld_msg *cmsg = (struct cld_msg *)src;
+ uint32_t xid;
+ struct nfsd_net *nn = net_generic(filp->f_dentry->d_sb->s_fs_info,
+ nfsd_net_id);
+ struct cld_net *cn = nn->cld_net;
+
+ if (mlen != sizeof(*cmsg)) {
+ dprintk("%s: got %lu bytes, expected %lu\n", __func__, mlen,
+ sizeof(*cmsg));
+ return -EINVAL;
+ }
+
+ /* copy just the xid so we can try to find that */
+ if (copy_from_user(&xid, &cmsg->cm_xid, sizeof(xid)) != 0) {
+ dprintk("%s: error when copying xid from userspace", __func__);
+ return -EFAULT;
+ }
+
+ /* walk the list and find corresponding xid */
+ cup = NULL;
+ spin_lock(&cn->cn_lock);
+ list_for_each_entry(tmp, &cn->cn_list, cu_list) {
+ if (get_unaligned(&tmp->cu_msg.cm_xid) == xid) {
+ cup = tmp;
+ list_del_init(&cup->cu_list);
+ break;
+ }
+ }
+ spin_unlock(&cn->cn_lock);
+
+ /* couldn't find upcall? */
+ if (!cup) {
+ dprintk("%s: couldn't find upcall -- xid=%u\n", __func__, xid);
+ return -EINVAL;
+ }
+
+ if (copy_from_user(&cup->cu_msg, src, mlen) != 0)
+ return -EFAULT;
+
+ wake_up_process(cup->cu_task);
+ return mlen;
+}
+
+static void
+cld_pipe_destroy_msg(struct rpc_pipe_msg *msg)
+{
+ struct cld_msg *cmsg = msg->data;
+ struct cld_upcall *cup = container_of(cmsg, struct cld_upcall,
+ cu_msg);
+
+ /* errno >= 0 means we got a downcall */
+ if (msg->errno >= 0)
+ return;
+
+ wake_up_process(cup->cu_task);
+}
+
+static const struct rpc_pipe_ops cld_upcall_ops = {
+ .upcall = rpc_pipe_generic_upcall,
+ .downcall = cld_pipe_downcall,
+ .destroy_msg = cld_pipe_destroy_msg,
+};
+
+static struct dentry *
+nfsd4_cld_register_sb(struct super_block *sb, struct rpc_pipe *pipe)
+{
+ struct dentry *dir, *dentry;
+
+ dir = rpc_d_lookup_sb(sb, NFSD_PIPE_DIR);
+ if (dir == NULL)
+ return ERR_PTR(-ENOENT);
+ dentry = rpc_mkpipe_dentry(dir, NFSD_CLD_PIPE, NULL, pipe);
+ dput(dir);
+ return dentry;
+}
+
+static void
+nfsd4_cld_unregister_sb(struct rpc_pipe *pipe)
+{
+ if (pipe->dentry)
+ rpc_unlink(pipe->dentry);
+}
+
+static struct dentry *
+nfsd4_cld_register_net(struct net *net, struct rpc_pipe *pipe)
+{
+ struct super_block *sb;
+ struct dentry *dentry;
+
+ sb = rpc_get_sb_net(net);
+ if (!sb)
+ return NULL;
+ dentry = nfsd4_cld_register_sb(sb, pipe);
+ rpc_put_sb_net(net);
+ return dentry;
+}
+
+static void
+nfsd4_cld_unregister_net(struct net *net, struct rpc_pipe *pipe)
+{
+ struct super_block *sb;
+
+ sb = rpc_get_sb_net(net);
+ if (sb) {
+ nfsd4_cld_unregister_sb(pipe);
+ rpc_put_sb_net(net);
+ }
+}
+
+/* Initialize rpc_pipefs pipe for communication with client tracking daemon */
+static int
+nfsd4_init_cld_pipe(struct net *net)
+{
+ int ret;
+ struct dentry *dentry;
+ struct nfsd_net *nn = net_generic(net, nfsd_net_id);
+ struct cld_net *cn;
+
+ if (nn->cld_net)
+ return 0;
+
+ cn = kzalloc(sizeof(*cn), GFP_KERNEL);
+ if (!cn) {
+ ret = -ENOMEM;
+ goto err;
+ }
+
+ cn->cn_pipe = rpc_mkpipe_data(&cld_upcall_ops, RPC_PIPE_WAIT_FOR_OPEN);
+ if (IS_ERR(cn->cn_pipe)) {
+ ret = PTR_ERR(cn->cn_pipe);
+ goto err;
+ }
+ spin_lock_init(&cn->cn_lock);
+ INIT_LIST_HEAD(&cn->cn_list);
+
+ dentry = nfsd4_cld_register_net(net, cn->cn_pipe);
+ if (IS_ERR(dentry)) {
+ ret = PTR_ERR(dentry);
+ goto err_destroy_data;
+ }
+
+ cn->cn_pipe->dentry = dentry;
+ nn->cld_net = cn;
+ return 0;
+
+err_destroy_data:
+ rpc_destroy_pipe_data(cn->cn_pipe);
+err:
+ kfree(cn);
+ printk(KERN_ERR "NFSD: unable to create nfsdcld upcall pipe (%d)\n",
+ ret);
+ return ret;
+}
+
+static void
+nfsd4_remove_cld_pipe(struct net *net)
+{
+ struct nfsd_net *nn = net_generic(net, nfsd_net_id);
+ struct cld_net *cn = nn->cld_net;
+
+ nfsd4_cld_unregister_net(net, cn->cn_pipe);
+ rpc_destroy_pipe_data(cn->cn_pipe);
+ kfree(nn->cld_net);
+ nn->cld_net = NULL;
+}
+
+static struct cld_upcall *
+alloc_cld_upcall(struct cld_net *cn)
+{
+ struct cld_upcall *new, *tmp;
+
+ new = kzalloc(sizeof(*new), GFP_KERNEL);
+ if (!new)
+ return new;
+
+ /* FIXME: hard cap on number in flight? */
+restart_search:
+ spin_lock(&cn->cn_lock);
+ list_for_each_entry(tmp, &cn->cn_list, cu_list) {
+ if (tmp->cu_msg.cm_xid == cn->cn_xid) {
+ cn->cn_xid++;
+ spin_unlock(&cn->cn_lock);
+ goto restart_search;
+ }
+ }
+ new->cu_task = current;
+ new->cu_msg.cm_vers = CLD_UPCALL_VERSION;
+ put_unaligned(cn->cn_xid++, &new->cu_msg.cm_xid);
+ new->cu_net = cn;
+ list_add(&new->cu_list, &cn->cn_list);
+ spin_unlock(&cn->cn_lock);
+
+ dprintk("%s: allocated xid %u\n", __func__, new->cu_msg.cm_xid);
+
+ return new;
+}
+
+static void
+free_cld_upcall(struct cld_upcall *victim)
+{
+ struct cld_net *cn = victim->cu_net;
+
+ spin_lock(&cn->cn_lock);
+ list_del(&victim->cu_list);
+ spin_unlock(&cn->cn_lock);
+ kfree(victim);
+}
+
+/* Ask daemon to create a new record */
+static void
+nfsd4_cld_create(struct nfs4_client *clp)
+{
+ int ret;
+ struct cld_upcall *cup;
+ /* FIXME: determine net from clp */
+ struct nfsd_net *nn = net_generic(&init_net, nfsd_net_id);
+ struct cld_net *cn = nn->cld_net;
+
+ /* Don't upcall if it's already stored */
+ if (test_bit(NFSD4_CLIENT_STABLE, &clp->cl_flags))
+ return;
+
+ cup = alloc_cld_upcall(cn);
+ if (!cup) {
+ ret = -ENOMEM;
+ goto out_err;
+ }
+
+ cup->cu_msg.cm_cmd = Cld_Create;
+ cup->cu_msg.cm_u.cm_name.cn_len = clp->cl_name.len;
+ memcpy(cup->cu_msg.cm_u.cm_name.cn_id, clp->cl_name.data,
+ clp->cl_name.len);
+
+ ret = cld_pipe_upcall(cn->cn_pipe, &cup->cu_msg);
+ if (!ret) {
+ ret = cup->cu_msg.cm_status;
+ set_bit(NFSD4_CLIENT_STABLE, &clp->cl_flags);
+ }
+
+ free_cld_upcall(cup);
+out_err:
+ if (ret)
+ printk(KERN_ERR "NFSD: Unable to create client "
+ "record on stable storage: %d\n", ret);
+}
+
+/* Ask daemon to create a new record */
+static void
+nfsd4_cld_remove(struct nfs4_client *clp)
+{
+ int ret;
+ struct cld_upcall *cup;
+ /* FIXME: determine net from clp */
+ struct nfsd_net *nn = net_generic(&init_net, nfsd_net_id);
+ struct cld_net *cn = nn->cld_net;
+
+ /* Don't upcall if it's already removed */
+ if (!test_bit(NFSD4_CLIENT_STABLE, &clp->cl_flags))
+ return;
+
+ cup = alloc_cld_upcall(cn);
+ if (!cup) {
+ ret = -ENOMEM;
+ goto out_err;
+ }
+
+ cup->cu_msg.cm_cmd = Cld_Remove;
+ cup->cu_msg.cm_u.cm_name.cn_len = clp->cl_name.len;
+ memcpy(cup->cu_msg.cm_u.cm_name.cn_id, clp->cl_name.data,
+ clp->cl_name.len);
+
+ ret = cld_pipe_upcall(cn->cn_pipe, &cup->cu_msg);
+ if (!ret) {
+ ret = cup->cu_msg.cm_status;
+ clear_bit(NFSD4_CLIENT_STABLE, &clp->cl_flags);
+ }
+
+ free_cld_upcall(cup);
+out_err:
+ if (ret)
+ printk(KERN_ERR "NFSD: Unable to remove client "
+ "record from stable storage: %d\n", ret);
+}
+
+/* Check for presence of a record, and update its timestamp */
+static int
+nfsd4_cld_check(struct nfs4_client *clp)
+{
+ int ret;
+ struct cld_upcall *cup;
+ /* FIXME: determine net from clp */
+ struct nfsd_net *nn = net_generic(&init_net, nfsd_net_id);
+ struct cld_net *cn = nn->cld_net;
+
+ /* Don't upcall if one was already stored during this grace pd */
+ if (test_bit(NFSD4_CLIENT_STABLE, &clp->cl_flags))
+ return 0;
+
+ cup = alloc_cld_upcall(cn);
+ if (!cup) {
+ printk(KERN_ERR "NFSD: Unable to check client record on "
+ "stable storage: %d\n", -ENOMEM);
+ return -ENOMEM;
+ }
+
+ cup->cu_msg.cm_cmd = Cld_Check;
+ cup->cu_msg.cm_u.cm_name.cn_len = clp->cl_name.len;
+ memcpy(cup->cu_msg.cm_u.cm_name.cn_id, clp->cl_name.data,
+ clp->cl_name.len);
+
+ ret = cld_pipe_upcall(cn->cn_pipe, &cup->cu_msg);
+ if (!ret) {
+ ret = cup->cu_msg.cm_status;
+ set_bit(NFSD4_CLIENT_STABLE, &clp->cl_flags);
+ }
+
+ free_cld_upcall(cup);
+ return ret;
+}
+
+static void
+nfsd4_cld_grace_done(struct net *net, time_t boot_time)
+{
+ int ret;
+ struct cld_upcall *cup;
+ struct nfsd_net *nn = net_generic(net, nfsd_net_id);
+ struct cld_net *cn = nn->cld_net;
+
+ cup = alloc_cld_upcall(cn);
+ if (!cup) {
+ ret = -ENOMEM;
+ goto out_err;
+ }
+
+ cup->cu_msg.cm_cmd = Cld_GraceDone;
+ cup->cu_msg.cm_u.cm_gracetime = (int64_t)boot_time;
+ ret = cld_pipe_upcall(cn->cn_pipe, &cup->cu_msg);
+ if (!ret)
+ ret = cup->cu_msg.cm_status;
+
+ free_cld_upcall(cup);
+out_err:
+ if (ret)
+ printk(KERN_ERR "NFSD: Unable to end grace period: %d\n", ret);
+}
+
+static struct nfsd4_client_tracking_ops nfsd4_cld_tracking_ops = {
+ .init = nfsd4_init_cld_pipe,
+ .exit = nfsd4_remove_cld_pipe,
+ .create = nfsd4_cld_create,
+ .remove = nfsd4_cld_remove,
+ .check = nfsd4_cld_check,
+ .grace_done = nfsd4_cld_grace_done,
+};
+
+int
+nfsd4_client_tracking_init(struct net *net)
+{
+ int status;
+ struct path path;
+
+ if (!client_tracking_ops) {
+ client_tracking_ops = &nfsd4_cld_tracking_ops;
+ status = kern_path(nfs4_recoverydir(), LOOKUP_FOLLOW, &path);
+ if (!status) {
+ if (S_ISDIR(path.dentry->d_inode->i_mode))
+ client_tracking_ops =
+ &nfsd4_legacy_tracking_ops;
+ path_put(&path);
+ }
+ }
+
+ status = client_tracking_ops->init(net);
+ if (status) {
+ printk(KERN_WARNING "NFSD: Unable to initialize client "
+ "recovery tracking! (%d)\n", status);
+ client_tracking_ops = NULL;
+ }
+ return status;
+}
+
+void
+nfsd4_client_tracking_exit(struct net *net)
+{
+ if (client_tracking_ops) {
+ client_tracking_ops->exit(net);
+ client_tracking_ops = NULL;
+ }
+}
+
+void
+nfsd4_client_record_create(struct nfs4_client *clp)
+{
+ if (client_tracking_ops)
+ client_tracking_ops->create(clp);
+}
+
+void
+nfsd4_client_record_remove(struct nfs4_client *clp)
+{
+ if (client_tracking_ops)
+ client_tracking_ops->remove(clp);
+}
+
+int
+nfsd4_client_record_check(struct nfs4_client *clp)
+{
+ if (client_tracking_ops)
+ return client_tracking_ops->check(clp);
+
+ return -EOPNOTSUPP;
+}
+
+void
+nfsd4_record_grace_done(struct net *net, time_t boot_time)
+{
+ if (client_tracking_ops)
+ client_tracking_ops->grace_done(net, boot_time);
+}
+
+static int
+rpc_pipefs_event(struct notifier_block *nb, unsigned long event, void *ptr)
+{
+ struct super_block *sb = ptr;
+ struct net *net = sb->s_fs_info;
+ struct nfsd_net *nn = net_generic(net, nfsd_net_id);
+ struct cld_net *cn = nn->cld_net;
+ struct dentry *dentry;
+ int ret = 0;
+
+ if (!try_module_get(THIS_MODULE))
+ return 0;
+
+ if (!cn) {
+ module_put(THIS_MODULE);
+ return 0;
+ }
+
+ switch (event) {
+ case RPC_PIPEFS_MOUNT:
+ dentry = nfsd4_cld_register_sb(sb, cn->cn_pipe);
+ if (IS_ERR(dentry)) {
+ ret = PTR_ERR(dentry);
+ break;
+ }
+ cn->cn_pipe->dentry = dentry;
+ break;
+ case RPC_PIPEFS_UMOUNT:
+ if (cn->cn_pipe->dentry)
+ nfsd4_cld_unregister_sb(cn->cn_pipe);
+ break;
+ default:
+ ret = -ENOTSUPP;
+ break;
+ }
+ module_put(THIS_MODULE);
+ return ret;
+}
+
+struct notifier_block nfsd4_cld_block = {
+ .notifier_call = rpc_pipefs_event,
+};
+
+int
+register_cld_notifier(void)
+{
+ return rpc_pipefs_notifier_register(&nfsd4_cld_block);
+}
+
+void
+unregister_cld_notifier(void)
+{
+ rpc_pipefs_notifier_unregister(&nfsd4_cld_block);
+}
projects / cascardo / linux.git / blobdiff