-#include "ceph_debug.h"
+#include <linux/ceph/ceph_debug.h>
+#include <linux/fs.h>
#include <linux/wait.h>
#include <linux/slab.h>
#include <linux/sched.h>
+#include <linux/debugfs.h>
+#include <linux/seq_file.h>
-#include "mds_client.h"
-#include "mon_client.h"
#include "super.h"
-#include "messenger.h"
-#include "decode.h"
-#include "auth.h"
-#include "pagelist.h"
+#include "mds_client.h"
+
+#include <linux/ceph/messenger.h>
+#include <linux/ceph/decode.h>
+#include <linux/ceph/pagelist.h>
+#include <linux/ceph/auth.h>
+#include <linux/ceph/debugfs.h>
/*
* A cluster of MDS (metadata server) daemons is responsible for
* are no longer valid.
*/
+struct ceph_reconnect_state {
+ struct ceph_pagelist *pagelist;
+ bool flock;
+};
+
static void __wake_requests(struct ceph_mds_client *mdsc,
struct list_head *head);
* parse individual inode info
*/
static int parse_reply_info_in(void **p, void *end,
- struct ceph_mds_reply_info_in *info)
+ struct ceph_mds_reply_info_in *info,
+ int features)
{
int err = -EIO;
info->symlink = *p;
*p += info->symlink_len;
+ if (features & CEPH_FEATURE_DIRLAYOUTHASH)
+ ceph_decode_copy_safe(p, end, &info->dir_layout,
+ sizeof(info->dir_layout), bad);
+ else
+ memset(&info->dir_layout, 0, sizeof(info->dir_layout));
+
ceph_decode_32_safe(p, end, info->xattr_len, bad);
ceph_decode_need(p, end, info->xattr_len, bad);
info->xattr_data = *p;
* target inode.
*/
static int parse_reply_info_trace(void **p, void *end,
- struct ceph_mds_reply_info_parsed *info)
+ struct ceph_mds_reply_info_parsed *info,
+ int features)
{
int err;
if (info->head->is_dentry) {
- err = parse_reply_info_in(p, end, &info->diri);
+ err = parse_reply_info_in(p, end, &info->diri, features);
if (err < 0)
goto out_bad;
}
if (info->head->is_target) {
- err = parse_reply_info_in(p, end, &info->targeti);
+ err = parse_reply_info_in(p, end, &info->targeti, features);
if (err < 0)
goto out_bad;
}
* parse readdir results
*/
static int parse_reply_info_dir(void **p, void *end,
- struct ceph_mds_reply_info_parsed *info)
+ struct ceph_mds_reply_info_parsed *info,
+ int features)
{
u32 num, i = 0;
int err;
*p += sizeof(struct ceph_mds_reply_lease);
/* inode */
- err = parse_reply_info_in(p, end, &info->dir_in[i]);
+ err = parse_reply_info_in(p, end, &info->dir_in[i], features);
if (err < 0)
goto out_bad;
i++;
return err;
}
+/*
+ * parse fcntl F_GETLK results
+ */
+static int parse_reply_info_filelock(void **p, void *end,
+ struct ceph_mds_reply_info_parsed *info,
+ int features)
+{
+ if (*p + sizeof(*info->filelock_reply) > end)
+ goto bad;
+
+ info->filelock_reply = *p;
+ *p += sizeof(*info->filelock_reply);
+
+ if (unlikely(*p != end))
+ goto bad;
+ return 0;
+
+bad:
+ return -EIO;
+}
+
+/*
+ * parse extra results
+ */
+static int parse_reply_info_extra(void **p, void *end,
+ struct ceph_mds_reply_info_parsed *info,
+ int features)
+{
+ if (info->head->op == CEPH_MDS_OP_GETFILELOCK)
+ return parse_reply_info_filelock(p, end, info, features);
+ else
+ return parse_reply_info_dir(p, end, info, features);
+}
+
/*
* parse entire mds reply
*/
static int parse_reply_info(struct ceph_msg *msg,
- struct ceph_mds_reply_info_parsed *info)
+ struct ceph_mds_reply_info_parsed *info,
+ int features)
{
void *p, *end;
u32 len;
/* trace */
ceph_decode_32_safe(&p, end, len, bad);
if (len > 0) {
- err = parse_reply_info_trace(&p, p+len, info);
+ err = parse_reply_info_trace(&p, p+len, info, features);
if (err < 0)
goto out_bad;
}
- /* dir content */
+ /* extra */
ceph_decode_32_safe(&p, end, len, bad);
if (len > 0) {
- err = parse_reply_info_dir(&p, p+len, info);
+ err = parse_reply_info_extra(&p, p+len, info, features);
if (err < 0)
goto out_bad;
}
atomic_read(&s->s_ref), atomic_read(&s->s_ref)-1);
if (atomic_dec_and_test(&s->s_ref)) {
if (s->s_authorizer)
- s->s_mdsc->client->monc.auth->ops->destroy_authorizer(
- s->s_mdsc->client->monc.auth, s->s_authorizer);
+ s->s_mdsc->fsc->client->monc.auth->ops->destroy_authorizer(
+ s->s_mdsc->fsc->client->monc.auth,
+ s->s_authorizer);
kfree(s);
}
}
s->s_seq = 0;
mutex_init(&s->s_mutex);
- ceph_con_init(mdsc->client->msgr, &s->s_con);
+ ceph_con_init(mdsc->fsc->client->msgr, &s->s_con);
s->s_con.private = s;
s->s_con.ops = &mds_con_ops;
s->s_con.peer_name.type = CEPH_ENTITY_TYPE_MDS;
destroy_reply_info(&req->r_reply_info);
}
if (req->r_inode) {
- ceph_put_cap_refs(ceph_inode(req->r_inode),
- CEPH_CAP_PIN);
+ ceph_put_cap_refs(ceph_inode(req->r_inode), CEPH_CAP_PIN);
iput(req->r_inode);
}
if (req->r_locked_dir)
- ceph_put_cap_refs(ceph_inode(req->r_locked_dir),
- CEPH_CAP_PIN);
+ ceph_put_cap_refs(ceph_inode(req->r_locked_dir), CEPH_CAP_PIN);
if (req->r_target_inode)
iput(req->r_target_inode);
if (req->r_dentry)
dput(req->r_dentry);
if (req->r_old_dentry) {
- ceph_put_cap_refs(
- ceph_inode(req->r_old_dentry->d_parent->d_inode),
- CEPH_CAP_PIN);
+ /*
+ * track (and drop pins for) r_old_dentry_dir
+ * separately, since r_old_dentry's d_parent may have
+ * changed between the dir mutex being dropped and
+ * this request being freed.
+ */
+ ceph_put_cap_refs(ceph_inode(req->r_old_dentry_dir),
+ CEPH_CAP_PIN);
dput(req->r_old_dentry);
+ iput(req->r_old_dentry_dir);
}
kfree(req->r_path1);
kfree(req->r_path2);
ceph_mdsc_get_request(req);
__insert_request(mdsc, req);
+ req->r_uid = current_fsuid();
+ req->r_gid = current_fsgid();
+
if (dir) {
struct ceph_inode_info *ci = ceph_inode(dir);
+ ihold(dir);
spin_lock(&ci->i_unsafe_lock);
req->r_unsafe_dir = dir;
list_add_tail(&req->r_unsafe_dir_item, &ci->i_unsafe_dirops);
spin_lock(&ci->i_unsafe_lock);
list_del_init(&req->r_unsafe_dir_item);
spin_unlock(&ci->i_unsafe_lock);
+
+ iput(req->r_unsafe_dir);
+ req->r_unsafe_dir = NULL;
}
ceph_mdsc_put_request(req);
*
* Called under mdsc->mutex.
*/
+struct dentry *get_nonsnap_parent(struct dentry *dentry)
+{
+ /*
+ * we don't need to worry about protecting the d_parent access
+ * here because we never renaming inside the snapped namespace
+ * except to resplice to another snapdir, and either the old or new
+ * result is a valid result.
+ */
+ while (!IS_ROOT(dentry) && ceph_snap(dentry->d_inode) != CEPH_NOSNAP)
+ dentry = dentry->d_parent;
+ return dentry;
+}
+
static int __choose_mds(struct ceph_mds_client *mdsc,
struct ceph_mds_request *req)
{
if (req->r_inode) {
inode = req->r_inode;
} else if (req->r_dentry) {
- if (req->r_dentry->d_inode) {
+ /* ignore race with rename; old or new d_parent is okay */
+ struct dentry *parent = req->r_dentry->d_parent;
+ struct inode *dir = parent->d_inode;
+
+ if (dir->i_sb != mdsc->fsc->sb) {
+ /* not this fs! */
+ inode = req->r_dentry->d_inode;
+ } else if (ceph_snap(dir) != CEPH_NOSNAP) {
+ /* direct snapped/virtual snapdir requests
+ * based on parent dir inode */
+ struct dentry *dn = get_nonsnap_parent(parent);
+ inode = dn->d_inode;
+ dout("__choose_mds using nonsnap parent %p\n", inode);
+ } else if (req->r_dentry->d_inode) {
+ /* dentry target */
inode = req->r_dentry->d_inode;
} else {
- inode = req->r_dentry->d_parent->d_inode;
- hash = req->r_dentry->d_name.hash;
+ /* dir + name */
+ inode = dir;
+ hash = ceph_dentry_hash(dir, req->r_dentry);
is_hash = true;
}
}
+
dout("__choose_mds %p is_hash=%d (%d) mode %d\n", inode, (int)is_hash,
(int)hash, mode);
if (!inode)
dout("choose_mds %p %llx.%llx "
"frag %u mds%d (%d/%d)\n",
inode, ceph_vinop(inode),
- frag.frag, frag.mds,
+ frag.frag, mds,
(int)r, frag.ndist);
- return mds;
+ if (ceph_mdsmap_get_state(mdsc->mdsmap, mds) >=
+ CEPH_MDS_STATE_ACTIVE)
+ return mds;
}
/* since this file/dir wasn't known to be
dout("choose_mds %p %llx.%llx "
"frag %u mds%d (auth)\n",
inode, ceph_vinop(inode), frag.frag, mds);
- return mds;
+ if (ceph_mdsmap_get_state(mdsc->mdsmap, mds) >=
+ CEPH_MDS_STATE_ACTIVE)
+ return mds;
}
}
}
struct ceph_msg *msg;
struct ceph_mds_session_head *h;
- msg = ceph_msg_new(CEPH_MSG_CLIENT_SESSION, sizeof(*h), GFP_NOFS);
+ msg = ceph_msg_new(CEPH_MSG_CLIENT_SESSION, sizeof(*h), GFP_NOFS,
+ false);
if (!msg) {
pr_err("create_session_msg ENOMEM creating msg\n");
return NULL;
__ceph_remove_cap(cap);
if (!__ceph_is_any_real_caps(ci)) {
struct ceph_mds_client *mdsc =
- &ceph_sb_to_client(inode->i_sb)->mdsc;
+ ceph_sb_to_client(inode->i_sb)->mdsc;
spin_lock(&mdsc->cap_dirty_lock);
if (!list_empty(&ci->i_dirty_item)) {
struct ceph_msg *msg, *partial = NULL;
struct ceph_mds_cap_release *head;
int err = -ENOMEM;
- int extra = mdsc->client->mount_args->cap_release_safety;
+ int extra = mdsc->fsc->mount_options->cap_release_safety;
int num;
dout("add_cap_releases %p mds%d extra %d\n", session, session->s_mds,
while (session->s_num_cap_releases < session->s_nr_caps + extra) {
spin_unlock(&session->s_cap_lock);
msg = ceph_msg_new(CEPH_MSG_CLIENT_CAPRELEASE, PAGE_CACHE_SIZE,
- GFP_NOFS);
+ GFP_NOFS, false);
if (!msg)
goto out_unlocked;
dout("add_cap_releases %p msg %p now %d\n", session, msg,
struct dentry *temp;
char *path;
int len, pos;
+ unsigned seq;
if (dentry == NULL)
return ERR_PTR(-EINVAL);
retry:
len = 0;
+ seq = read_seqbegin(&rename_lock);
+ rcu_read_lock();
for (temp = dentry; !IS_ROOT(temp);) {
struct inode *inode = temp->d_inode;
if (inode && ceph_snap(inode) == CEPH_SNAPDIR)
len += 1 + temp->d_name.len;
temp = temp->d_parent;
if (temp == NULL) {
+ rcu_read_unlock();
pr_err("build_path corrupt dentry %p\n", dentry);
return ERR_PTR(-EINVAL);
}
}
+ rcu_read_unlock();
if (len)
len--; /* no leading '/' */
return ERR_PTR(-ENOMEM);
pos = len;
path[pos] = 0; /* trailing null */
+ rcu_read_lock();
for (temp = dentry; !IS_ROOT(temp) && pos != 0; ) {
- struct inode *inode = temp->d_inode;
+ struct inode *inode;
+ spin_lock(&temp->d_lock);
+ inode = temp->d_inode;
if (inode && ceph_snap(inode) == CEPH_SNAPDIR) {
dout("build_path path+%d: %p SNAPDIR\n",
pos, temp);
break;
} else {
pos -= temp->d_name.len;
- if (pos < 0)
+ if (pos < 0) {
+ spin_unlock(&temp->d_lock);
break;
+ }
strncpy(path + pos, temp->d_name.name,
temp->d_name.len);
}
+ spin_unlock(&temp->d_lock);
if (pos)
path[--pos] = '/';
temp = temp->d_parent;
if (temp == NULL) {
+ rcu_read_unlock();
pr_err("build_path corrupt dentry\n");
kfree(path);
return ERR_PTR(-EINVAL);
}
}
- if (pos != 0) {
+ rcu_read_unlock();
+ if (pos != 0 || read_seqretry(&rename_lock, seq)) {
pr_err("build_path did not end path lookup where "
"expected, namelen is %d, pos is %d\n", len, pos);
/* presumably this is only possible if racing with a
*base = ceph_ino(temp->d_inode);
*plen = len;
dout("build_path on %p %d built %llx '%.*s'\n",
- dentry, atomic_read(&dentry->d_count), *base, len, path);
+ dentry, dentry->d_count, *base, len, path);
return path;
}
r = build_dentry_path(rdentry, ppath, pathlen, ino, freepath);
dout(" dentry %p %llx/%.*s\n", rdentry, *ino, *pathlen,
*ppath);
- } else if (rpath) {
+ } else if (rpath || rino) {
*ino = rino;
*ppath = rpath;
*pathlen = strlen(rpath);
if (req->r_old_dentry_drop)
len += req->r_old_dentry->d_name.len;
- msg = ceph_msg_new(CEPH_MSG_CLIENT_REQUEST, len, GFP_NOFS);
+ msg = ceph_msg_new(CEPH_MSG_CLIENT_REQUEST, len, GFP_NOFS, false);
if (!msg) {
msg = ERR_PTR(-ENOMEM);
goto out_free2;
head->mdsmap_epoch = cpu_to_le32(mdsc->mdsmap->m_epoch);
head->op = cpu_to_le32(req->r_op);
- head->caller_uid = cpu_to_le32(current_fsuid());
- head->caller_gid = cpu_to_le32(current_fsgid());
+ head->caller_uid = cpu_to_le32(req->r_uid);
+ head->caller_gid = cpu_to_le32(req->r_gid);
head->args = req->r_args;
ceph_encode_filepath(&p, end, ino1, path1);
struct ceph_msg *msg;
int flags = 0;
- req->r_mds = mds;
req->r_attempts++;
+ if (req->r_inode) {
+ struct ceph_cap *cap =
+ ceph_get_cap_for_mds(ceph_inode(req->r_inode), mds);
+
+ if (cap)
+ req->r_sent_on_mseq = cap->mseq;
+ else
+ req->r_sent_on_mseq = -1;
+ }
dout("prepare_send_request %p tid %lld %s (attempt %d)\n", req,
req->r_tid, ceph_mds_op_name(req->r_op), req->r_attempts);
goto finish;
}
+ put_request_session(req);
+
mds = __choose_mds(mdsc, req);
if (mds < 0 ||
ceph_mdsmap_get_state(mdsc->mdsmap, mds) < CEPH_MDS_STATE_ACTIVE) {
goto finish;
}
}
+ req->r_session = get_session(session);
+
dout("do_request mds%d session %p state %s\n", mds, session,
session_state_name(session->s_state));
if (session->s_state != CEPH_MDS_SESSION_OPEN &&
}
/* send request */
- req->r_session = get_session(session);
req->r_resend_mds = -1; /* forget any previous mds hint */
if (req->r_request_started == 0) /* note request start time */
if (req->r_session &&
req->r_session->s_mds == mds) {
dout(" kicking tid %llu\n", req->r_tid);
- put_request_session(req);
__do_request(mdsc, req);
}
}
if (req->r_locked_dir)
ceph_get_cap_refs(ceph_inode(req->r_locked_dir), CEPH_CAP_PIN);
if (req->r_old_dentry)
- ceph_get_cap_refs(
- ceph_inode(req->r_old_dentry->d_parent->d_inode),
- CEPH_CAP_PIN);
+ ceph_get_cap_refs(ceph_inode(req->r_old_dentry_dir),
+ CEPH_CAP_PIN);
/* issue */
mutex_lock(&mdsc->mutex);
}
/*
- * Invalidate dir I_COMPLETE, dentry lease state on an aborted MDS
+ * Invalidate dir D_COMPLETE, dentry lease state on an aborted MDS
* namespace request.
*/
void ceph_invalidate_dir_request(struct ceph_mds_request *req)
struct inode *inode = req->r_locked_dir;
struct ceph_inode_info *ci = ceph_inode(inode);
- dout("invalidate_dir_request %p (I_COMPLETE, lease(s))\n", inode);
+ dout("invalidate_dir_request %p (D_COMPLETE, lease(s))\n", inode);
spin_lock(&inode->i_lock);
- ci->i_ceph_flags &= ~CEPH_I_COMPLETE;
+ ceph_dir_clear_complete(inode);
ci->i_release_count++;
spin_unlock(&inode->i_lock);
result = le32_to_cpu(head->result);
/*
- * Tolerate 2 consecutive ESTALEs from the same mds.
- * FIXME: we should be looking at the cap migrate_seq.
+ * Handle an ESTALE
+ * if we're not talking to the authority, send to them
+ * if the authority has changed while we weren't looking,
+ * send to new authority
+ * Otherwise we just have to return an ESTALE
*/
if (result == -ESTALE) {
- req->r_direct_mode = USE_AUTH_MDS;
- req->r_num_stale++;
- if (req->r_num_stale <= 2) {
+ dout("got ESTALE on request %llu", req->r_tid);
+ if (!req->r_inode) {
+ /* do nothing; not an authority problem */
+ } else if (req->r_direct_mode != USE_AUTH_MDS) {
+ dout("not using auth, setting for that now");
+ req->r_direct_mode = USE_AUTH_MDS;
__do_request(mdsc, req);
mutex_unlock(&mdsc->mutex);
goto out;
+ } else {
+ struct ceph_inode_info *ci = ceph_inode(req->r_inode);
+ struct ceph_cap *cap = NULL;
+
+ if (req->r_session)
+ cap = ceph_get_cap_for_mds(ci,
+ req->r_session->s_mds);
+
+ dout("already using auth");
+ if ((!cap || cap != ci->i_auth_cap) ||
+ (cap->mseq != req->r_sent_on_mseq)) {
+ dout("but cap changed, so resending");
+ __do_request(mdsc, req);
+ mutex_unlock(&mdsc->mutex);
+ goto out;
+ }
}
- } else {
- req->r_num_stale = 0;
+ dout("have to return ESTALE on request %llu", req->r_tid);
}
+
if (head->safe) {
req->r_got_safe = true;
__unregister_request(mdsc, req);
dout("handle_reply tid %lld result %d\n", tid, result);
rinfo = &req->r_reply_info;
- err = parse_reply_info(msg, rinfo);
+ err = parse_reply_info(msg, rinfo, session->s_con.peer_features);
mutex_unlock(&mdsc->mutex);
mutex_lock(&session->s_mutex);
if (err < 0) {
- pr_err("mdsc_handle_reply got corrupt reply mds%d\n", mds);
+ pr_err("mdsc_handle_reply got corrupt reply mds%d(tid:%lld)\n", mds, tid);
ceph_msg_dump(msg);
goto out_err;
}
/* insert trace into our cache */
mutex_lock(&req->r_fill_mutex);
- err = ceph_fill_trace(mdsc->client->sb, req, req->r_session);
+ err = ceph_fill_trace(mdsc->fsc->sb, req, req->r_session);
if (err == 0) {
- if (result == 0 && rinfo->dir_nr)
+ if (result == 0 && req->r_op != CEPH_MDS_OP_GETFILELOCK &&
+ rinfo->dir_nr)
ceph_readdir_prepopulate(req, req->r_session);
ceph_unreserve_caps(mdsc, &req->r_caps_reservation);
}
pr_info("mds%d reconnect denied\n", session->s_mds);
remove_session_caps(session);
wake = 1; /* for good measure */
- complete_all(&mdsc->session_close_waiters);
+ wake_up_all(&mdsc->session_close_wq);
kick_requests(mdsc, mds);
break;
static int encode_caps_cb(struct inode *inode, struct ceph_cap *cap,
void *arg)
{
- struct ceph_mds_cap_reconnect rec;
+ union {
+ struct ceph_mds_cap_reconnect v2;
+ struct ceph_mds_cap_reconnect_v1 v1;
+ } rec;
+ size_t reclen;
struct ceph_inode_info *ci;
- struct ceph_pagelist *pagelist = arg;
+ struct ceph_reconnect_state *recon_state = arg;
+ struct ceph_pagelist *pagelist = recon_state->pagelist;
char *path;
int pathlen, err;
u64 pathbase;
path = ceph_mdsc_build_path(dentry, &pathlen, &pathbase, 0);
if (IS_ERR(path)) {
err = PTR_ERR(path);
- BUG_ON(err);
+ goto out_dput;
}
} else {
path = NULL;
}
err = ceph_pagelist_encode_string(pagelist, path, pathlen);
if (err)
- goto out;
+ goto out_free;
spin_lock(&inode->i_lock);
cap->seq = 0; /* reset cap seq */
cap->issue_seq = 0; /* and issue_seq */
- rec.cap_id = cpu_to_le64(cap->cap_id);
- rec.pathbase = cpu_to_le64(pathbase);
- rec.wanted = cpu_to_le32(__ceph_caps_wanted(ci));
- rec.issued = cpu_to_le32(cap->issued);
- rec.size = cpu_to_le64(inode->i_size);
- ceph_encode_timespec(&rec.mtime, &inode->i_mtime);
- ceph_encode_timespec(&rec.atime, &inode->i_atime);
- rec.snaprealm = cpu_to_le64(ci->i_snap_realm->ino);
+
+ if (recon_state->flock) {
+ rec.v2.cap_id = cpu_to_le64(cap->cap_id);
+ rec.v2.wanted = cpu_to_le32(__ceph_caps_wanted(ci));
+ rec.v2.issued = cpu_to_le32(cap->issued);
+ rec.v2.snaprealm = cpu_to_le64(ci->i_snap_realm->ino);
+ rec.v2.pathbase = cpu_to_le64(pathbase);
+ rec.v2.flock_len = 0;
+ reclen = sizeof(rec.v2);
+ } else {
+ rec.v1.cap_id = cpu_to_le64(cap->cap_id);
+ rec.v1.wanted = cpu_to_le32(__ceph_caps_wanted(ci));
+ rec.v1.issued = cpu_to_le32(cap->issued);
+ rec.v1.size = cpu_to_le64(inode->i_size);
+ ceph_encode_timespec(&rec.v1.mtime, &inode->i_mtime);
+ ceph_encode_timespec(&rec.v1.atime, &inode->i_atime);
+ rec.v1.snaprealm = cpu_to_le64(ci->i_snap_realm->ino);
+ rec.v1.pathbase = cpu_to_le64(pathbase);
+ reclen = sizeof(rec.v1);
+ }
spin_unlock(&inode->i_lock);
- err = ceph_pagelist_append(pagelist, &rec, sizeof(rec));
+ if (recon_state->flock) {
+ int num_fcntl_locks, num_flock_locks;
+ struct ceph_pagelist_cursor trunc_point;
+
+ ceph_pagelist_set_cursor(pagelist, &trunc_point);
+ do {
+ lock_flocks();
+ ceph_count_locks(inode, &num_fcntl_locks,
+ &num_flock_locks);
+ rec.v2.flock_len = (2*sizeof(u32) +
+ (num_fcntl_locks+num_flock_locks) *
+ sizeof(struct ceph_filelock));
+ unlock_flocks();
+
+ /* pre-alloc pagelist */
+ ceph_pagelist_truncate(pagelist, &trunc_point);
+ err = ceph_pagelist_append(pagelist, &rec, reclen);
+ if (!err)
+ err = ceph_pagelist_reserve(pagelist,
+ rec.v2.flock_len);
+
+ /* encode locks */
+ if (!err) {
+ lock_flocks();
+ err = ceph_encode_locks(inode,
+ pagelist,
+ num_fcntl_locks,
+ num_flock_locks);
+ unlock_flocks();
+ }
+ } while (err == -ENOSPC);
+ } else {
+ err = ceph_pagelist_append(pagelist, &rec, reclen);
+ }
-out:
+out_free:
kfree(path);
+out_dput:
dput(dentry);
return err;
}
int mds = session->s_mds;
int err = -ENOMEM;
struct ceph_pagelist *pagelist;
+ struct ceph_reconnect_state recon_state;
pr_info("mds%d reconnect start\n", mds);
goto fail_nopagelist;
ceph_pagelist_init(pagelist);
- reply = ceph_msg_new(CEPH_MSG_CLIENT_RECONNECT, 0, GFP_NOFS);
+ reply = ceph_msg_new(CEPH_MSG_CLIENT_RECONNECT, 0, GFP_NOFS, false);
if (!reply)
goto fail_nomsg;
err = ceph_pagelist_encode_32(pagelist, session->s_nr_caps);
if (err)
goto fail;
- err = iterate_session_caps(session, encode_caps_cb, pagelist);
+
+ recon_state.pagelist = pagelist;
+ recon_state.flock = session->s_con.peer_features & CEPH_FEATURE_FLOCK;
+ err = iterate_session_caps(session, encode_caps_cb, &recon_state);
if (err < 0)
goto fail;
}
reply->pagelist = pagelist;
+ if (recon_state.flock)
+ reply->hdr.version = cpu_to_le16(2);
reply->hdr.data_len = cpu_to_le32(pagelist->length);
reply->nr_pages = calc_pages_for(0, pagelist->length);
ceph_con_send(&session->s_con, reply);
struct ceph_mds_session *session,
struct ceph_msg *msg)
{
- struct super_block *sb = mdsc->client->sb;
+ struct super_block *sb = mdsc->fsc->sb;
struct inode *inode;
- struct ceph_inode_info *ci;
struct dentry *parent, *dentry;
struct ceph_dentry_info *di;
int mds = session->s_mds;
struct ceph_mds_lease *h = msg->front.iov_base;
u32 seq;
struct ceph_vino vino;
- int mask;
struct qstr dname;
int release = 0;
goto bad;
vino.ino = le64_to_cpu(h->ino);
vino.snap = CEPH_NOSNAP;
- mask = le16_to_cpu(h->mask);
seq = le32_to_cpu(h->seq);
dname.name = (void *)h + sizeof(*h) + sizeof(u32);
dname.len = msg->front.iov_len - sizeof(*h) - sizeof(u32);
/* lookup inode */
inode = ceph_find_inode(sb, vino);
- dout("handle_lease %s, mask %d, ino %llx %p %.*s\n",
- ceph_lease_op_name(h->action), mask, vino.ino, inode,
+ dout("handle_lease %s, ino %llx %p %.*s\n",
+ ceph_lease_op_name(h->action), vino.ino, inode,
dname.len, dname.name);
if (inode == NULL) {
dout("handle_lease no inode %llx\n", vino.ino);
goto release;
}
- ci = ceph_inode(inode);
/* dentry */
parent = d_find_alias(inode);
dnamelen = dentry->d_name.len;
len += dnamelen;
- msg = ceph_msg_new(CEPH_MSG_CLIENT_LEASE, len, GFP_NOFS);
+ msg = ceph_msg_new(CEPH_MSG_CLIENT_LEASE, len, GFP_NOFS, false);
if (!msg)
return;
lease = msg->front.iov_base;
lease->action = action;
- lease->mask = cpu_to_le16(1);
lease->ino = cpu_to_le64(ceph_vino(inode).ino);
lease->first = lease->last = cpu_to_le64(ceph_vino(inode).snap);
lease->seq = cpu_to_le32(seq);
* Pass @inode always, @dentry is optional.
*/
void ceph_mdsc_lease_release(struct ceph_mds_client *mdsc, struct inode *inode,
- struct dentry *dentry, int mask)
+ struct dentry *dentry)
{
struct ceph_dentry_info *di;
struct ceph_mds_session *session;
BUG_ON(inode == NULL);
BUG_ON(dentry == NULL);
- BUG_ON(mask == 0);
/* is dentry lease valid? */
spin_lock(&dentry->d_lock);
di->lease_gen != di->lease_session->s_cap_gen ||
!time_before(jiffies, dentry->d_time)) {
dout("lease_release inode %p dentry %p -- "
- "no lease on %d\n",
- inode, dentry, mask);
+ "no lease\n",
+ inode, dentry);
spin_unlock(&dentry->d_lock);
return;
}
__ceph_mdsc_drop_dentry_lease(dentry);
spin_unlock(&dentry->d_lock);
- dout("lease_release inode %p dentry %p mask %d to mds%d\n",
- inode, dentry, mask, session->s_mds);
+ dout("lease_release inode %p dentry %p to mds%d\n",
+ inode, dentry, session->s_mds);
ceph_mdsc_lease_send_msg(session, inode, dentry,
CEPH_MDS_LEASE_RELEASE, seq);
ceph_put_mds_session(session);
schedule_delayed(mdsc);
}
+int ceph_mdsc_init(struct ceph_fs_client *fsc)
-int ceph_mdsc_init(struct ceph_mds_client *mdsc, struct ceph_client *client)
{
- mdsc->client = client;
+ struct ceph_mds_client *mdsc;
+
+ mdsc = kzalloc(sizeof(struct ceph_mds_client), GFP_NOFS);
+ if (!mdsc)
+ return -ENOMEM;
+ mdsc->fsc = fsc;
+ fsc->mdsc = mdsc;
mutex_init(&mdsc->mutex);
mdsc->mdsmap = kzalloc(sizeof(*mdsc->mdsmap), GFP_NOFS);
if (mdsc->mdsmap == NULL)
return -ENOMEM;
init_completion(&mdsc->safe_umount_waiters);
- init_completion(&mdsc->session_close_waiters);
+ init_waitqueue_head(&mdsc->session_close_wq);
INIT_LIST_HEAD(&mdsc->waiting_for_map);
mdsc->sessions = NULL;
mdsc->max_sessions = 0;
spin_lock_init(&mdsc->snap_flush_lock);
mdsc->cap_flush_seq = 0;
INIT_LIST_HEAD(&mdsc->cap_dirty);
+ INIT_LIST_HEAD(&mdsc->cap_dirty_migrating);
mdsc->num_cap_flushing = 0;
spin_lock_init(&mdsc->cap_dirty_lock);
init_waitqueue_head(&mdsc->cap_flushing_wq);
INIT_LIST_HEAD(&mdsc->dentry_lru);
ceph_caps_init(mdsc);
- ceph_adjust_min_caps(mdsc, client->min_caps);
+ ceph_adjust_min_caps(mdsc, fsc->min_caps);
return 0;
}
static void wait_requests(struct ceph_mds_client *mdsc)
{
struct ceph_mds_request *req;
- struct ceph_client *client = mdsc->client;
+ struct ceph_fs_client *fsc = mdsc->fsc;
mutex_lock(&mdsc->mutex);
if (__get_oldest_req(mdsc)) {
dout("wait_requests waiting for requests\n");
wait_for_completion_timeout(&mdsc->safe_umount_waiters,
- client->mount_args->mount_timeout * HZ);
+ fsc->client->options->mount_timeout * HZ);
/* tear down remaining requests */
mutex_lock(&mdsc->mutex);
{
u64 want_tid, want_flush;
- if (mdsc->client->mount_state == CEPH_MOUNT_SHUTDOWN)
+ if (mdsc->fsc->mount_state == CEPH_MOUNT_SHUTDOWN)
return;
dout("sync\n");
wait_event(mdsc->cap_flushing_wq, check_cap_flush(mdsc, want_flush));
}
+/*
+ * true if all sessions are closed, or we force unmount
+ */
+bool done_closing_sessions(struct ceph_mds_client *mdsc)
+{
+ int i, n = 0;
+
+ if (mdsc->fsc->mount_state == CEPH_MOUNT_SHUTDOWN)
+ return true;
+
+ mutex_lock(&mdsc->mutex);
+ for (i = 0; i < mdsc->max_sessions; i++)
+ if (mdsc->sessions[i])
+ n++;
+ mutex_unlock(&mdsc->mutex);
+ return n == 0;
+}
/*
* called after sb is ro.
{
struct ceph_mds_session *session;
int i;
- int n;
- struct ceph_client *client = mdsc->client;
- unsigned long started, timeout = client->mount_args->mount_timeout * HZ;
+ struct ceph_fs_client *fsc = mdsc->fsc;
+ unsigned long timeout = fsc->client->options->mount_timeout * HZ;
dout("close_sessions\n");
- mutex_lock(&mdsc->mutex);
-
/* close sessions */
- started = jiffies;
- while (time_before(jiffies, started + timeout)) {
- dout("closing sessions\n");
- n = 0;
- for (i = 0; i < mdsc->max_sessions; i++) {
- session = __ceph_lookup_mds_session(mdsc, i);
- if (!session)
- continue;
- mutex_unlock(&mdsc->mutex);
- mutex_lock(&session->s_mutex);
- __close_session(mdsc, session);
- mutex_unlock(&session->s_mutex);
- ceph_put_mds_session(session);
- mutex_lock(&mdsc->mutex);
- n++;
- }
- if (n == 0)
- break;
-
- if (client->mount_state == CEPH_MOUNT_SHUTDOWN)
- break;
-
- dout("waiting for sessions to close\n");
+ mutex_lock(&mdsc->mutex);
+ for (i = 0; i < mdsc->max_sessions; i++) {
+ session = __ceph_lookup_mds_session(mdsc, i);
+ if (!session)
+ continue;
mutex_unlock(&mdsc->mutex);
- wait_for_completion_timeout(&mdsc->session_close_waiters,
- timeout);
+ mutex_lock(&session->s_mutex);
+ __close_session(mdsc, session);
+ mutex_unlock(&session->s_mutex);
+ ceph_put_mds_session(session);
mutex_lock(&mdsc->mutex);
}
+ mutex_unlock(&mdsc->mutex);
+
+ dout("waiting for sessions to close\n");
+ wait_event_timeout(mdsc->session_close_wq, done_closing_sessions(mdsc),
+ timeout);
/* tear down remaining sessions */
+ mutex_lock(&mdsc->mutex);
for (i = 0; i < mdsc->max_sessions; i++) {
if (mdsc->sessions[i]) {
session = get_session(mdsc->sessions[i]);
mutex_lock(&mdsc->mutex);
}
}
-
WARN_ON(!list_empty(&mdsc->cap_delay_list));
-
mutex_unlock(&mdsc->mutex);
ceph_cleanup_empty_realms(mdsc);
dout("stopped\n");
}
-void ceph_mdsc_stop(struct ceph_mds_client *mdsc)
+static void ceph_mdsc_stop(struct ceph_mds_client *mdsc)
{
dout("stop\n");
cancel_delayed_work_sync(&mdsc->delayed_work); /* cancel timer */
ceph_caps_finalize(mdsc);
}
+void ceph_mdsc_destroy(struct ceph_fs_client *fsc)
+{
+ struct ceph_mds_client *mdsc = fsc->mdsc;
+
+ dout("mdsc_destroy %p\n", mdsc);
+ ceph_mdsc_stop(mdsc);
+
+ /* flush out any connection work with references to us */
+ ceph_msgr_flush();
+
+ fsc->mdsc = NULL;
+ kfree(mdsc);
+ dout("mdsc_destroy %p done\n", mdsc);
+}
+
/*
* handle mds map update.
ceph_decode_need(&p, end, sizeof(fsid)+2*sizeof(u32), bad);
ceph_decode_copy(&p, &fsid, sizeof(fsid));
- if (ceph_check_fsid(mdsc->client, &fsid) < 0)
+ if (ceph_check_fsid(mdsc->fsc->client, &fsid) < 0)
return;
epoch = ceph_decode_32(&p);
maplen = ceph_decode_32(&p);
dout("handle_map epoch %u len %d\n", epoch, (int)maplen);
/* do we need it? */
- ceph_monc_got_mdsmap(&mdsc->client->monc, epoch);
+ ceph_monc_got_mdsmap(&mdsc->fsc->client->monc, epoch);
mutex_lock(&mdsc->mutex);
if (mdsc->mdsmap && epoch <= mdsc->mdsmap->m_epoch) {
dout("handle_map epoch %u <= our %u\n",
} else {
mdsc->mdsmap = newmap; /* first mds map */
}
- mdsc->client->sb->s_maxbytes = mdsc->mdsmap->m_max_file_size;
+ mdsc->fsc->sb->s_maxbytes = mdsc->mdsmap->m_max_file_size;
__wake_requests(mdsc, &mdsc->waiting_for_map);
{
struct ceph_mds_session *s = con->private;
+ dout("mdsc con_put %p (%d)\n", s, atomic_read(&s->s_ref) - 1);
ceph_put_mds_session(s);
- dout("mdsc con_put %p (%d)\n", s, atomic_read(&s->s_ref));
}
/*
{
struct ceph_mds_session *s = con->private;
struct ceph_mds_client *mdsc = s->s_mdsc;
- struct ceph_auth_client *ac = mdsc->client->monc.auth;
+ struct ceph_auth_client *ac = mdsc->fsc->client->monc.auth;
int ret = 0;
if (force_new && s->s_authorizer) {
{
struct ceph_mds_session *s = con->private;
struct ceph_mds_client *mdsc = s->s_mdsc;
- struct ceph_auth_client *ac = mdsc->client->monc.auth;
+ struct ceph_auth_client *ac = mdsc->fsc->client->monc.auth;
return ac->ops->verify_authorizer_reply(ac, s->s_authorizer, len);
}
{
struct ceph_mds_session *s = con->private;
struct ceph_mds_client *mdsc = s->s_mdsc;
- struct ceph_auth_client *ac = mdsc->client->monc.auth;
+ struct ceph_auth_client *ac = mdsc->fsc->client->monc.auth;
if (ac->ops->invalidate_authorizer)
ac->ops->invalidate_authorizer(ac, CEPH_ENTITY_TYPE_MDS);
- return ceph_monc_validate_auth(&mdsc->client->monc);
+ return ceph_monc_validate_auth(&mdsc->fsc->client->monc);
}
static const struct ceph_connection_operations mds_con_ops = {
.peer_reset = peer_reset,
};
-
-
-
/* eof */