summaryrefslogtreecommitdiffstats
path: root/fs/ceph
diff options
context:
space:
mode:
Diffstat (limited to 'fs/ceph')
-rw-r--r--fs/ceph/Kconfig2
-rw-r--r--fs/ceph/auth_x.c3
-rw-r--r--fs/ceph/caps.c15
-rw-r--r--fs/ceph/dir.c13
-rw-r--r--fs/ceph/file.c2
-rw-r--r--fs/ceph/inode.c6
-rw-r--r--fs/ceph/mds_client.c45
-rw-r--r--fs/ceph/mds_client.h1
-rw-r--r--fs/ceph/messenger.c71
-rw-r--r--fs/ceph/mon_client.c6
-rw-r--r--fs/ceph/osd_client.c6
-rw-r--r--fs/ceph/osdmap.c27
12 files changed, 122 insertions, 75 deletions
diff --git a/fs/ceph/Kconfig b/fs/ceph/Kconfig
index 04b8280582a..bc87b9c1d27 100644
--- a/fs/ceph/Kconfig
+++ b/fs/ceph/Kconfig
@@ -2,7 +2,7 @@ config CEPH_FS
tristate "Ceph distributed file system (EXPERIMENTAL)"
depends on INET && EXPERIMENTAL
select LIBCRC32C
- select CONFIG_CRYPTO_AES
+ select CRYPTO_AES
help
Choose Y or M here to include support for mounting the
experimental Ceph distributed file system. Ceph is an extremely
diff --git a/fs/ceph/auth_x.c b/fs/ceph/auth_x.c
index 3fe49042d8a..6d44053ecff 100644
--- a/fs/ceph/auth_x.c
+++ b/fs/ceph/auth_x.c
@@ -613,6 +613,9 @@ static void ceph_x_destroy(struct ceph_auth_client *ac)
remove_ticket_handler(ac, th);
}
+ if (xi->auth_authorizer.buf)
+ ceph_buffer_put(xi->auth_authorizer.buf);
+
kfree(ac->private);
ac->private = NULL;
}
diff --git a/fs/ceph/caps.c b/fs/ceph/caps.c
index 74144d6389f..b81be9a5648 100644
--- a/fs/ceph/caps.c
+++ b/fs/ceph/caps.c
@@ -627,7 +627,7 @@ retry:
if (fmode >= 0)
__ceph_get_fmode(ci, fmode);
spin_unlock(&inode->i_lock);
- wake_up(&ci->i_cap_wq);
+ wake_up_all(&ci->i_cap_wq);
return 0;
}
@@ -1181,7 +1181,7 @@ static int __send_cap(struct ceph_mds_client *mdsc, struct ceph_cap *cap,
}
if (wake)
- wake_up(&ci->i_cap_wq);
+ wake_up_all(&ci->i_cap_wq);
return delayed;
}
@@ -2153,7 +2153,7 @@ void ceph_put_cap_refs(struct ceph_inode_info *ci, int had)
else if (flushsnaps)
ceph_flush_snaps(ci);
if (wake)
- wake_up(&ci->i_cap_wq);
+ wake_up_all(&ci->i_cap_wq);
if (put)
iput(inode);
}
@@ -2229,7 +2229,7 @@ void ceph_put_wrbuffer_cap_refs(struct ceph_inode_info *ci, int nr,
iput(inode);
} else if (complete_capsnap) {
ceph_flush_snaps(ci);
- wake_up(&ci->i_cap_wq);
+ wake_up_all(&ci->i_cap_wq);
}
if (drop_capsnap)
iput(inode);
@@ -2405,7 +2405,7 @@ static void handle_cap_grant(struct inode *inode, struct ceph_mds_caps *grant,
if (queue_invalidate)
ceph_queue_invalidate(inode);
if (wake)
- wake_up(&ci->i_cap_wq);
+ wake_up_all(&ci->i_cap_wq);
if (check_caps == 1)
ceph_check_caps(ci, CHECK_CAPS_NODELAY|CHECK_CAPS_AUTHONLY,
@@ -2460,7 +2460,7 @@ static void handle_cap_flush_ack(struct inode *inode, u64 flush_tid,
struct ceph_inode_info,
i_flushing_item)->vfs_inode);
mdsc->num_cap_flushing--;
- wake_up(&mdsc->cap_flushing_wq);
+ wake_up_all(&mdsc->cap_flushing_wq);
dout(" inode %p now !flushing\n", inode);
if (ci->i_dirty_caps == 0) {
@@ -2472,7 +2472,7 @@ static void handle_cap_flush_ack(struct inode *inode, u64 flush_tid,
}
}
spin_unlock(&mdsc->cap_dirty_lock);
- wake_up(&ci->i_cap_wq);
+ wake_up_all(&ci->i_cap_wq);
out:
spin_unlock(&inode->i_lock);
@@ -2984,6 +2984,7 @@ int ceph_encode_dentry_release(void **p, struct dentry *dentry,
memcpy(*p, dentry->d_name.name, dentry->d_name.len);
*p += dentry->d_name.len;
rel->dname_seq = cpu_to_le32(di->lease_seq);
+ __ceph_mdsc_drop_dentry_lease(dentry);
}
spin_unlock(&dentry->d_lock);
return ret;
diff --git a/fs/ceph/dir.c b/fs/ceph/dir.c
index f85719310db..f94ed3c7f6a 100644
--- a/fs/ceph/dir.c
+++ b/fs/ceph/dir.c
@@ -266,6 +266,7 @@ static int ceph_readdir(struct file *filp, void *dirent, filldir_t filldir)
spin_lock(&inode->i_lock);
if ((filp->f_pos == 2 || fi->dentry) &&
!ceph_test_opt(client, NOASYNCREADDIR) &&
+ ceph_snap(inode) != CEPH_SNAPDIR &&
(ci->i_ceph_flags & CEPH_I_COMPLETE) &&
__ceph_caps_issued_mask(ci, CEPH_CAP_FILE_SHARED, 1)) {
err = __dcache_readdir(filp, dirent, filldir);
@@ -1013,18 +1014,22 @@ out_touch:
/*
* When a dentry is released, clear the dir I_COMPLETE if it was part
- * of the current dir gen.
+ * of the current dir gen or if this is in the snapshot namespace.
*/
static void ceph_dentry_release(struct dentry *dentry)
{
struct ceph_dentry_info *di = ceph_dentry(dentry);
struct inode *parent_inode = dentry->d_parent->d_inode;
+ u64 snapid = ceph_snap(parent_inode);
- if (parent_inode) {
+ dout("dentry_release %p parent %p\n", dentry, parent_inode);
+
+ if (parent_inode && snapid != CEPH_SNAPDIR) {
struct ceph_inode_info *ci = ceph_inode(parent_inode);
spin_lock(&parent_inode->i_lock);
- if (ci->i_shared_gen == di->lease_shared_gen) {
+ if (ci->i_shared_gen == di->lease_shared_gen ||
+ snapid <= CEPH_MAXSNAP) {
dout(" clearing %p complete (d_release)\n",
parent_inode);
ci->i_ceph_flags &= ~CEPH_I_COMPLETE;
@@ -1241,7 +1246,9 @@ struct dentry_operations ceph_dentry_ops = {
struct dentry_operations ceph_snapdir_dentry_ops = {
.d_revalidate = ceph_snapdir_d_revalidate,
+ .d_release = ceph_dentry_release,
};
struct dentry_operations ceph_snap_dentry_ops = {
+ .d_release = ceph_dentry_release,
};
diff --git a/fs/ceph/file.c b/fs/ceph/file.c
index 6251a1574b9..7c08698fad3 100644
--- a/fs/ceph/file.c
+++ b/fs/ceph/file.c
@@ -265,7 +265,7 @@ int ceph_release(struct inode *inode, struct file *file)
kmem_cache_free(ceph_file_cachep, cf);
/* wake up anyone waiting for caps on this inode */
- wake_up(&ci->i_cap_wq);
+ wake_up_all(&ci->i_cap_wq);
return 0;
}
diff --git a/fs/ceph/inode.c b/fs/ceph/inode.c
index 8f9b9fe8ef9..389f9dbd994 100644
--- a/fs/ceph/inode.c
+++ b/fs/ceph/inode.c
@@ -1199,8 +1199,10 @@ retry_lookup:
goto out;
}
err = ceph_init_dentry(dn);
- if (err < 0)
+ if (err < 0) {
+ dput(dn);
goto out;
+ }
} else if (dn->d_inode &&
(ceph_ino(dn->d_inode) != vino.ino ||
ceph_snap(dn->d_inode) != vino.snap)) {
@@ -1499,7 +1501,7 @@ retry:
if (wrbuffer_refs == 0)
ceph_check_caps(ci, CHECK_CAPS_AUTHONLY, NULL);
if (wake)
- wake_up(&ci->i_cap_wq);
+ wake_up_all(&ci->i_cap_wq);
}
diff --git a/fs/ceph/mds_client.c b/fs/ceph/mds_client.c
index 3ab79f6c4ce..dd440bd438a 100644
--- a/fs/ceph/mds_client.c
+++ b/fs/ceph/mds_client.c
@@ -868,7 +868,7 @@ static int wake_up_session_cb(struct inode *inode, struct ceph_cap *cap,
{
struct ceph_inode_info *ci = ceph_inode(inode);
- wake_up(&ci->i_cap_wq);
+ wake_up_all(&ci->i_cap_wq);
if (arg) {
spin_lock(&inode->i_lock);
ci->i_wanted_max_size = 0;
@@ -1514,6 +1514,9 @@ static struct ceph_msg *create_request_message(struct ceph_mds_client *mdsc,
ceph_encode_filepath(&p, end, ino1, path1);
ceph_encode_filepath(&p, end, ino2, path2);
+ /* make note of release offset, in case we need to replay */
+ req->r_request_release_offset = p - msg->front.iov_base;
+
/* cap releases */
releases = 0;
if (req->r_inode_drop)
@@ -1561,7 +1564,7 @@ static void complete_request(struct ceph_mds_client *mdsc,
if (req->r_callback)
req->r_callback(mdsc, req);
else
- complete(&req->r_completion);
+ complete_all(&req->r_completion);
}
/*
@@ -1580,6 +1583,32 @@ static int __prepare_send_request(struct ceph_mds_client *mdsc,
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);
+ if (req->r_got_unsafe) {
+ /*
+ * Replay. Do not regenerate message (and rebuild
+ * paths, etc.); just use the original message.
+ * Rebuilding paths will break for renames because
+ * d_move mangles the src name.
+ */
+ msg = req->r_request;
+ rhead = msg->front.iov_base;
+
+ flags = le32_to_cpu(rhead->flags);
+ flags |= CEPH_MDS_FLAG_REPLAY;
+ rhead->flags = cpu_to_le32(flags);
+
+ if (req->r_target_inode)
+ rhead->ino = cpu_to_le64(ceph_ino(req->r_target_inode));
+
+ rhead->num_retry = req->r_attempts - 1;
+
+ /* remove cap/dentry releases from message */
+ rhead->num_releases = 0;
+ msg->hdr.front_len = cpu_to_le32(req->r_request_release_offset);
+ msg->front.iov_len = req->r_request_release_offset;
+ return 0;
+ }
+
if (req->r_request) {
ceph_msg_put(req->r_request);
req->r_request = NULL;
@@ -1601,13 +1630,9 @@ static int __prepare_send_request(struct ceph_mds_client *mdsc,
rhead->flags = cpu_to_le32(flags);
rhead->num_fwd = req->r_num_fwd;
rhead->num_retry = req->r_attempts - 1;
+ rhead->ino = 0;
dout(" r_locked_dir = %p\n", req->r_locked_dir);
-
- if (req->r_target_inode && req->r_got_unsafe)
- rhead->ino = cpu_to_le64(ceph_ino(req->r_target_inode));
- else
- rhead->ino = 0;
return 0;
}
@@ -1907,7 +1932,7 @@ static void handle_reply(struct ceph_mds_session *session, struct ceph_msg *msg)
if (head->safe) {
req->r_got_safe = true;
__unregister_request(mdsc, req);
- complete(&req->r_safe_completion);
+ complete_all(&req->r_safe_completion);
if (req->r_got_unsafe) {
/*
@@ -1922,7 +1947,7 @@ static void handle_reply(struct ceph_mds_session *session, struct ceph_msg *msg)
/* last unsafe request during umount? */
if (mdsc->stopping && !__get_oldest_req(mdsc))
- complete(&mdsc->safe_umount_waiters);
+ complete_all(&mdsc->safe_umount_waiters);
mutex_unlock(&mdsc->mutex);
goto out;
}
@@ -2101,7 +2126,7 @@ static void handle_session(struct ceph_mds_session *session,
pr_info("mds%d reconnect denied\n", session->s_mds);
remove_session_caps(session);
wake = 1; /* for good measure */
- complete(&mdsc->session_close_waiters);
+ complete_all(&mdsc->session_close_waiters);
kick_requests(mdsc, mds);
break;
diff --git a/fs/ceph/mds_client.h b/fs/ceph/mds_client.h
index b292fa42a66..952410c60d0 100644
--- a/fs/ceph/mds_client.h
+++ b/fs/ceph/mds_client.h
@@ -188,6 +188,7 @@ struct ceph_mds_request {
int r_old_inode_drop, r_old_inode_unless;
struct ceph_msg *r_request; /* original request */
+ int r_request_release_offset;
struct ceph_msg *r_reply;
struct ceph_mds_reply_info_parsed r_reply_info;
int r_err;
diff --git a/fs/ceph/messenger.c b/fs/ceph/messenger.c
index 9ad43a310a4..15167b2daa5 100644
--- a/fs/ceph/messenger.c
+++ b/fs/ceph/messenger.c
@@ -43,7 +43,8 @@ static void ceph_fault(struct ceph_connection *con);
* nicely render a sockaddr as a string.
*/
#define MAX_ADDR_STR 20
-static char addr_str[MAX_ADDR_STR][40];
+#define MAX_ADDR_STR_LEN 60
+static char addr_str[MAX_ADDR_STR][MAX_ADDR_STR_LEN];
static DEFINE_SPINLOCK(addr_str_lock);
static int last_addr_str;
@@ -52,7 +53,6 @@ const char *pr_addr(const struct sockaddr_storage *ss)
int i;
char *s;
struct sockaddr_in *in4 = (void *)ss;
- unsigned char *quad = (void *)&in4->sin_addr.s_addr;
struct sockaddr_in6 *in6 = (void *)ss;
spin_lock(&addr_str_lock);
@@ -64,25 +64,13 @@ const char *pr_addr(const struct sockaddr_storage *ss)
switch (ss->ss_family) {
case AF_INET:
- sprintf(s, "%u.%u.%u.%u:%u",
- (unsigned int)quad[0],
- (unsigned int)quad[1],
- (unsigned int)quad[2],
- (unsigned int)quad[3],
- (unsigned int)ntohs(in4->sin_port));
+ snprintf(s, MAX_ADDR_STR_LEN, "%pI4:%u", &in4->sin_addr,
+ (unsigned int)ntohs(in4->sin_port));
break;
case AF_INET6:
- sprintf(s, "%04x:%04x:%04x:%04x:%04x:%04x:%04x:%04x:%u",
- in6->sin6_addr.s6_addr16[0],
- in6->sin6_addr.s6_addr16[1],
- in6->sin6_addr.s6_addr16[2],
- in6->sin6_addr.s6_addr16[3],
- in6->sin6_addr.s6_addr16[4],
- in6->sin6_addr.s6_addr16[5],
- in6->sin6_addr.s6_addr16[6],
- in6->sin6_addr.s6_addr16[7],
- (unsigned int)ntohs(in6->sin6_port));
+ snprintf(s, MAX_ADDR_STR_LEN, "[%pI6c]:%u", &in6->sin6_addr,
+ (unsigned int)ntohs(in6->sin6_port));
break;
default:
@@ -215,12 +203,13 @@ static void set_sock_callbacks(struct socket *sock,
*/
static struct socket *ceph_tcp_connect(struct ceph_connection *con)
{
- struct sockaddr *paddr = (struct sockaddr *)&con->peer_addr.in_addr;
+ struct sockaddr_storage *paddr = &con->peer_addr.in_addr;
struct socket *sock;
int ret;
BUG_ON(con->sock);
- ret = sock_create_kern(AF_INET, SOCK_STREAM, IPPROTO_TCP, &sock);
+ ret = sock_create_kern(con->peer_addr.in_addr.ss_family, SOCK_STREAM,
+ IPPROTO_TCP, &sock);
if (ret)
return ERR_PTR(ret);
con->sock = sock;
@@ -234,7 +223,8 @@ static struct socket *ceph_tcp_connect(struct ceph_connection *con)
dout("connect %s\n", pr_addr(&con->peer_addr.in_addr));
- ret = sock->ops->connect(sock, paddr, sizeof(*paddr), O_NONBLOCK);
+ ret = sock->ops->connect(sock, (struct sockaddr *)paddr, sizeof(*paddr),
+ O_NONBLOCK);
if (ret == -EINPROGRESS) {
dout("connect %s EINPROGRESS sk_state = %u\n",
pr_addr(&con->peer_addr.in_addr),
@@ -1009,19 +999,32 @@ int ceph_parse_ips(const char *c, const char *end,
struct sockaddr_in *in4 = (void *)ss;
struct sockaddr_in6 *in6 = (void *)ss;
int port;
+ char delim = ',';
+
+ if (*p == '[') {
+ delim = ']';
+ p++;
+ }
memset(ss, 0, sizeof(*ss));
if (in4_pton(p, end - p, (u8 *)&in4->sin_addr.s_addr,
- ',', &ipend)) {
+ delim, &ipend))
ss->ss_family = AF_INET;
- } else if (in6_pton(p, end - p, (u8 *)&in6->sin6_addr.s6_addr,
- ',', &ipend)) {
+ else if (in6_pton(p, end - p, (u8 *)&in6->sin6_addr.s6_addr,
+ delim, &ipend))
ss->ss_family = AF_INET6;
- } else {
+ else
goto bad;
- }
p = ipend;
+ if (delim == ']') {
+ if (*p != ']') {
+ dout("missing matching ']'\n");
+ goto bad;
+ }
+ p++;
+ }
+
/* port? */
if (p < end && *p == ':') {
port = 0;
@@ -1055,7 +1058,7 @@ int ceph_parse_ips(const char *c, const char *end,
return 0;
bad:
- pr_err("parse_ips bad ip '%s'\n", c);
+ pr_err("parse_ips bad ip '%.*s'\n", (int)(end - c), c);
return -EINVAL;
}
@@ -2015,20 +2018,20 @@ void ceph_con_revoke(struct ceph_connection *con, struct ceph_msg *msg)
{
mutex_lock(&con->mutex);
if (!list_empty(&msg->list_head)) {
- dout("con_revoke %p msg %p\n", con, msg);
+ dout("con_revoke %p msg %p - was on queue\n", con, msg);
list_del_init(&msg->list_head);
ceph_msg_put(msg);
msg->hdr.seq = 0;
- if (con->out_msg == msg) {
- ceph_msg_put(con->out_msg);
- con->out_msg = NULL;
- }
+ }
+ if (con->out_msg == msg) {
+ dout("con_revoke %p msg %p - was sending\n", con, msg);
+ con->out_msg = NULL;
if (con->out_kvec_is_msg) {
con->out_skip = con->out_kvec_bytes;
con->out_kvec_is_msg = false;
}
- } else {
- dout("con_revoke %p msg %p - not queued (sent?)\n", con, msg);
+ ceph_msg_put(msg);
+ msg->hdr.seq = 0;
}
mutex_unlock(&con->mutex);
}
diff --git a/fs/ceph/mon_client.c b/fs/ceph/mon_client.c
index cc115eafae1..54fe01c5070 100644
--- a/fs/ceph/mon_client.c
+++ b/fs/ceph/mon_client.c
@@ -345,7 +345,7 @@ static void ceph_monc_handle_map(struct ceph_mon_client *monc,
out:
mutex_unlock(&monc->mutex);
- wake_up(&client->auth_wq);
+ wake_up_all(&client->auth_wq);
}
/*
@@ -462,7 +462,7 @@ static void handle_statfs_reply(struct ceph_mon_client *monc,
}
mutex_unlock(&monc->mutex);
if (req) {
- complete(&req->completion);
+ complete_all(&req->completion);
put_generic_request(req);
}
return;
@@ -718,7 +718,7 @@ static void handle_auth_reply(struct ceph_mon_client *monc,
monc->m_auth->front_max);
if (ret < 0) {
monc->client->auth_err = ret;
- wake_up(&monc->client->auth_wq);
+ wake_up_all(&monc->client->auth_wq);
} else if (ret > 0) {
__send_prepared_auth_request(monc, ret);
} else if (!was_auth && monc->auth->ops->is_authenticated(monc->auth)) {
diff --git a/fs/ceph/osd_client.c b/fs/ceph/osd_client.c
index 92b7251a53f..e3852234789 100644
--- a/fs/ceph/osd_client.c
+++ b/fs/ceph/osd_client.c
@@ -862,12 +862,12 @@ static void handle_reply(struct ceph_osd_client *osdc, struct ceph_msg *msg,
if (req->r_callback)
req->r_callback(req, msg);
else
- complete(&req->r_completion);
+ complete_all(&req->r_completion);
if (flags & CEPH_OSD_FLAG_ONDISK) {
if (req->r_safe_callback)
req->r_safe_callback(req, msg);
- complete(&req->r_safe_completion); /* fsync waiter */
+ complete_all(&req->r_safe_completion); /* fsync waiter */
}
done:
@@ -1083,7 +1083,7 @@ done:
if (newmap)
kick_requests(osdc, NULL);
up_read(&osdc->map_sem);
- wake_up(&osdc->client->auth_wq);
+ wake_up_all(&osdc->client->auth_wq);
return;
bad:
diff --git a/fs/ceph/osdmap.c b/fs/ceph/osdmap.c
index 50ce64ebd33..416d46adbf8 100644
--- a/fs/ceph/osdmap.c
+++ b/fs/ceph/osdmap.c
@@ -568,6 +568,7 @@ struct ceph_osdmap *osdmap_decode(void **p, void *end)
if (ev > CEPH_PG_POOL_VERSION) {
pr_warning("got unknown v %d > %d of ceph_pg_pool\n",
ev, CEPH_PG_POOL_VERSION);
+ kfree(pi);
goto bad;
}
__decode_pool(p, pi);
@@ -830,12 +831,13 @@ struct ceph_osdmap *osdmap_apply_incremental(void **p, void *end,
/* remove any? */
while (rbp && pgid_cmp(rb_entry(rbp, struct ceph_pg_mapping,
node)->pgid, pgid) <= 0) {
- struct rb_node *cur = rbp;
+ struct ceph_pg_mapping *cur =
+ rb_entry(rbp, struct ceph_pg_mapping, node);
+
rbp = rb_next(rbp);
- dout(" removed pg_temp %llx\n",
- *(u64 *)&rb_entry(cur, struct ceph_pg_mapping,
- node)->pgid);
- rb_erase(cur, &map->pg_temp);
+ dout(" removed pg_temp %llx\n", *(u64 *)&cur->pgid);
+ rb_erase(&cur->node, &map->pg_temp);
+ kfree(cur);
}
if (pglen) {
@@ -851,19 +853,22 @@ struct ceph_osdmap *osdmap_apply_incremental(void **p, void *end,
for (j = 0; j < pglen; j++)
pg->osds[j] = ceph_decode_32(p);
err = __insert_pg_mapping(pg, &map->pg_temp);
- if (err)
+ if (err) {
+ kfree(pg);
goto bad;
+ }
dout(" added pg_temp %llx len %d\n", *(u64 *)&pgid,
pglen);
}
}
while (rbp) {
- struct rb_node *cur = rbp;
+ struct ceph_pg_mapping *cur =
+ rb_entry(rbp, struct ceph_pg_mapping, node);
+
rbp = rb_next(rbp);
- dout(" removed pg_temp %llx\n",
- *(u64 *)&rb_entry(cur, struct ceph_pg_mapping,
- node)->pgid);
- rb_erase(cur, &map->pg_temp);
+ dout(" removed pg_temp %llx\n", *(u64 *)&cur->pgid);
+ rb_erase(&cur->node, &map->pg_temp);
+ kfree(cur);
}
/* ignore the rest */
generated by cgit v1.2.3-70-g09d2 (git 2.46.0) at 2024-12-25 08:46:11 +0000