summaryrefslogtreecommitdiffstats
path: root/fs/ceph
diff options
context:
space:
mode:
Diffstat (limited to 'fs/ceph')
-rw-r--r--fs/ceph/auth_x.c5
-rw-r--r--fs/ceph/caps.c21
-rw-r--r--fs/ceph/crush/mapper.c41
-rw-r--r--fs/ceph/debugfs.c2
-rw-r--r--fs/ceph/inode.c19
-rw-r--r--fs/ceph/mds_client.c41
-rw-r--r--fs/ceph/mds_client.h1
-rw-r--r--fs/ceph/messenger.c75
-rw-r--r--fs/ceph/mon_client.c3
-rw-r--r--fs/ceph/osd_client.c3
-rw-r--r--fs/ceph/osdmap.c2
11 files changed, 141 insertions, 72 deletions
diff --git a/fs/ceph/auth_x.c b/fs/ceph/auth_x.c
index 83d4d2785ffe..6d44053ecff1 100644
--- a/fs/ceph/auth_x.c
+++ b/fs/ceph/auth_x.c
@@ -493,7 +493,7 @@ static int ceph_x_handle_reply(struct ceph_auth_client *ac, int result,
return -EAGAIN;
}
- op = le32_to_cpu(head->op);
+ op = le16_to_cpu(head->op);
result = le32_to_cpu(head->result);
dout("handle_reply op %d result %d\n", op, result);
switch (op) {
@@ -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 619b61655ee5..74144d6389f0 100644
--- a/fs/ceph/caps.c
+++ b/fs/ceph/caps.c
@@ -244,8 +244,14 @@ static struct ceph_cap *get_cap(struct ceph_cap_reservation *ctx)
struct ceph_cap *cap = NULL;
/* temporary, until we do something about cap import/export */
- if (!ctx)
- return kmem_cache_alloc(ceph_cap_cachep, GFP_NOFS);
+ if (!ctx) {
+ cap = kmem_cache_alloc(ceph_cap_cachep, GFP_NOFS);
+ if (cap) {
+ caps_use_count++;
+ caps_total_count++;
+ }
+ return cap;
+ }
spin_lock(&caps_list_lock);
dout("get_cap ctx=%p (%d) %d = %d used + %d resv + %d avail\n",
@@ -2886,18 +2892,19 @@ int ceph_encode_inode_release(void **p, struct inode *inode,
struct ceph_inode_info *ci = ceph_inode(inode);
struct ceph_cap *cap;
struct ceph_mds_request_release *rel = *p;
+ int used, dirty;
int ret = 0;
- int used = 0;
spin_lock(&inode->i_lock);
used = __ceph_caps_used(ci);
+ dirty = __ceph_caps_dirty(ci);
- dout("encode_inode_release %p mds%d used %s drop %s unless %s\n", inode,
- mds, ceph_cap_string(used), ceph_cap_string(drop),
+ dout("encode_inode_release %p mds%d used|dirty %s drop %s unless %s\n",
+ inode, mds, ceph_cap_string(used|dirty), ceph_cap_string(drop),
ceph_cap_string(unless));
- /* only drop unused caps */
- drop &= ~used;
+ /* only drop unused, clean caps */
+ drop &= ~(used | dirty);
cap = __get_cap_for_mds(ci, mds);
if (cap && __cap_is_valid(cap)) {
diff --git a/fs/ceph/crush/mapper.c b/fs/ceph/crush/mapper.c
index 9ba54efb6543..a4eec133258e 100644
--- a/fs/ceph/crush/mapper.c
+++ b/fs/ceph/crush/mapper.c
@@ -238,7 +238,7 @@ static int bucket_straw_choose(struct crush_bucket_straw *bucket,
static int crush_bucket_choose(struct crush_bucket *in, int x, int r)
{
- dprintk("choose %d x=%d r=%d\n", in->id, x, r);
+ dprintk(" crush_bucket_choose %d x=%d r=%d\n", in->id, x, r);
switch (in->alg) {
case CRUSH_BUCKET_UNIFORM:
return bucket_uniform_choose((struct crush_bucket_uniform *)in,
@@ -264,7 +264,7 @@ static int crush_bucket_choose(struct crush_bucket *in, int x, int r)
*/
static int is_out(struct crush_map *map, __u32 *weight, int item, int x)
{
- if (weight[item] >= 0x1000)
+ if (weight[item] >= 0x10000)
return 0;
if (weight[item] == 0)
return 1;
@@ -305,7 +305,9 @@ static int crush_choose(struct crush_map *map,
int itemtype;
int collide, reject;
const int orig_tries = 5; /* attempts before we fall back to search */
- dprintk("choose bucket %d x %d outpos %d\n", bucket->id, x, outpos);
+
+ dprintk("CHOOSE%s bucket %d x %d outpos %d numrep %d\n", recurse_to_leaf ? "_LEAF" : "",
+ bucket->id, x, outpos, numrep);
for (rep = outpos; rep < numrep; rep++) {
/* keep trying until we get a non-out, non-colliding item */
@@ -366,6 +368,7 @@ static int crush_choose(struct crush_map *map,
BUG_ON(item >= 0 ||
(-1-item) >= map->max_buckets);
in = map->buckets[-1-item];
+ retry_bucket = 1;
continue;
}
@@ -377,15 +380,25 @@ static int crush_choose(struct crush_map *map,
}
}
- if (recurse_to_leaf &&
- item < 0 &&
- crush_choose(map, map->buckets[-1-item],
- weight,
- x, outpos+1, 0,
- out2, outpos,
- firstn, 0, NULL) <= outpos) {
- reject = 1;
- } else {
+ reject = 0;
+ if (recurse_to_leaf) {
+ if (item < 0) {
+ if (crush_choose(map,
+ map->buckets[-1-item],
+ weight,
+ x, outpos+1, 0,
+ out2, outpos,
+ firstn, 0,
+ NULL) <= outpos)
+ /* didn't get leaf */
+ reject = 1;
+ } else {
+ /* we already have a leaf! */
+ out2[outpos] = item;
+ }
+ }
+
+ if (!reject) {
/* out? */
if (itemtype == 0)
reject = is_out(map, weight,
@@ -424,12 +437,12 @@ reject:
continue;
}
- dprintk("choose got %d\n", item);
+ dprintk("CHOOSE got %d\n", item);
out[outpos] = item;
outpos++;
}
- dprintk("choose returns %d\n", outpos);
+ dprintk("CHOOSE returns %d\n", outpos);
return outpos;
}
diff --git a/fs/ceph/debugfs.c b/fs/ceph/debugfs.c
index 3be33fb066cc..f2f5332ddbba 100644
--- a/fs/ceph/debugfs.c
+++ b/fs/ceph/debugfs.c
@@ -261,7 +261,7 @@ static int osdc_show(struct seq_file *s, void *pp)
static int caps_show(struct seq_file *s, void *p)
{
- struct ceph_client *client = p;
+ struct ceph_client *client = s->private;
int total, avail, used, reserved, min;
ceph_reservation_status(client, &total, &avail, &used, &reserved, &min);
diff --git a/fs/ceph/inode.c b/fs/ceph/inode.c
index ab47f46ca282..8f9b9fe8ef9f 100644
--- a/fs/ceph/inode.c
+++ b/fs/ceph/inode.c
@@ -854,8 +854,8 @@ static struct dentry *splice_dentry(struct dentry *dn, struct inode *in,
d_drop(dn);
realdn = d_materialise_unique(dn, in);
if (IS_ERR(realdn)) {
- pr_err("splice_dentry error %p inode %p ino %llx.%llx\n",
- dn, in, ceph_vinop(in));
+ pr_err("splice_dentry error %ld %p inode %p ino %llx.%llx\n",
+ PTR_ERR(realdn), dn, in, ceph_vinop(in));
if (prehash)
*prehash = false; /* don't rehash on error */
dn = realdn; /* note realdn contains the error */
@@ -1234,18 +1234,23 @@ retry_lookup:
goto out;
}
dn = splice_dentry(dn, in, NULL);
+ if (IS_ERR(dn))
+ dn = NULL;
}
if (fill_inode(in, &rinfo->dir_in[i], NULL, session,
req->r_request_started, -1,
&req->r_caps_reservation) < 0) {
pr_err("fill_inode badness on %p\n", in);
- dput(dn);
- continue;
+ goto next_item;
}
- update_dentry_lease(dn, rinfo->dir_dlease[i],
- req->r_session, req->r_request_started);
- dput(dn);
+ if (dn)
+ update_dentry_lease(dn, rinfo->dir_dlease[i],
+ req->r_session,
+ req->r_request_started);
+next_item:
+ if (dn)
+ dput(dn);
}
req->r_did_prepopulate = true;
diff --git a/fs/ceph/mds_client.c b/fs/ceph/mds_client.c
index 1766947fc07a..416c08d315db 100644
--- a/fs/ceph/mds_client.c
+++ b/fs/ceph/mds_client.c
@@ -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)
@@ -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;
}
@@ -2783,6 +2808,12 @@ void ceph_mdsc_pre_umount(struct ceph_mds_client *mdsc)
drop_leases(mdsc);
ceph_flush_dirty_caps(mdsc);
wait_requests(mdsc);
+
+ /*
+ * wait for reply handlers to drop their request refs and
+ * their inode/dcache refs
+ */
+ ceph_msgr_flush();
}
/*
diff --git a/fs/ceph/mds_client.h b/fs/ceph/mds_client.h
index b292fa42a66d..952410c60d09 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 64b8b1f7863d..15167b2daa55 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),
@@ -657,7 +647,7 @@ static void prepare_write_connect(struct ceph_messenger *msgr,
dout("prepare_write_connect %p cseq=%d gseq=%d proto=%d\n", con,
con->connect_seq, global_seq, proto);
- con->out_connect.features = CEPH_FEATURE_SUPPORTED_CLIENT;
+ con->out_connect.features = cpu_to_le64(CEPH_FEATURE_SUPPORTED_CLIENT);
con->out_connect.host_type = cpu_to_le32(CEPH_ENTITY_TYPE_CLIENT);
con->out_connect.connect_seq = cpu_to_le32(con->connect_seq);
con->out_connect.global_seq = cpu_to_le32(global_seq);
@@ -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;
}
@@ -1396,10 +1399,12 @@ static int read_partial_message(struct ceph_connection *con)
if (!con->in_msg) {
dout("got hdr type %d front %d data %d\n", con->in_hdr.type,
con->in_hdr.front_len, con->in_hdr.data_len);
+ skip = 0;
con->in_msg = ceph_alloc_msg(con, &con->in_hdr, &skip);
if (skip) {
/* skip this message */
dout("alloc_msg said skip message\n");
+ BUG_ON(con->in_msg);
con->in_base_pos = -front_len - middle_len - data_len -
sizeof(m->footer);
con->in_tag = CEPH_MSGR_TAG_READY;
@@ -2013,20 +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 07a539906e67..cc115eafae11 100644
--- a/fs/ceph/mon_client.c
+++ b/fs/ceph/mon_client.c
@@ -725,7 +725,8 @@ static void handle_auth_reply(struct ceph_mon_client *monc,
dout("authenticated, starting session\n");
monc->client->msgr->inst.name.type = CEPH_ENTITY_TYPE_CLIENT;
- monc->client->msgr->inst.name.num = monc->auth->global_id;
+ monc->client->msgr->inst.name.num =
+ cpu_to_le64(monc->auth->global_id);
__send_subscribe(monc);
__resend_generic_request(monc);
diff --git a/fs/ceph/osd_client.c b/fs/ceph/osd_client.c
index d25b4add85b4..92b7251a53f1 100644
--- a/fs/ceph/osd_client.c
+++ b/fs/ceph/osd_client.c
@@ -1344,7 +1344,7 @@ static void dispatch(struct ceph_connection *con, struct ceph_msg *msg)
int type = le16_to_cpu(msg->hdr.type);
if (!osd)
- return;
+ goto out;
osdc = osd->o_osdc;
switch (type) {
@@ -1359,6 +1359,7 @@ static void dispatch(struct ceph_connection *con, struct ceph_msg *msg)
pr_err("received unknown message type %d %s\n", type,
ceph_msg_type_name(type));
}
+out:
ceph_msg_put(msg);
}
diff --git a/fs/ceph/osdmap.c b/fs/ceph/osdmap.c
index ddc656fb5c05..277f8b339577 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);
@@ -707,6 +708,7 @@ struct ceph_osdmap *osdmap_apply_incremental(void **p, void *end,
newcrush = crush_decode(*p, min(*p+len, end));
if (IS_ERR(newcrush))
return ERR_CAST(newcrush);
+ *p += len;
}
/* new flags? */