@@ -2297,8 +2297,9 @@ static int flush_mdlog_and_wait_inode_unsafe_requests(struct inode *inode)
struct ceph_mds_client *mdsc = ceph_sb_to_client(inode->i_sb)->mdsc;
struct ceph_inode_info *ci = ceph_inode(inode);
struct ceph_mds_request *req1 = NULL, *req2 = NULL;
+ struct ceph_mds_session *s, **sessions = NULL;
unsigned int max_sessions;
- int ret, err = 0;
+ int i, ret, err = 0;
spin_lock(&ci->i_unsafe_lock);
if (S_ISDIR(inode->i_mode) && !list_empty(&ci->i_unsafe_dirops)) {
@@ -2315,31 +2316,22 @@ static int flush_mdlog_and_wait_inode_unsafe_requests(struct inode *inode)
}
spin_unlock(&ci->i_unsafe_lock);
- /*
- * The mdsc->max_sessions is unlikely to be changed
- * mostly, here we will retry it by reallocating the
- * sessions array memory to get rid of the mdsc->mutex
- * lock.
- */
-retry:
- max_sessions = mdsc->max_sessions;
-
/*
* Trigger to flush the journal logs in all the relevant MDSes
* manually, or in the worst case we must wait at most 5 seconds
* to wait the journal logs to be flushed by the MDSes periodically.
*/
+ mutex_lock(&mdsc->mutex);
+ max_sessions = mdsc->max_sessions;
+ sessions = kcalloc(max_sessions, sizeof(s), GFP_KERNEL);
+ if (!sessions) {
+ mutex_unlock(&mdsc->mutex);
+ err = -ENOMEM;
+ goto out;
+ }
+
if ((req1 || req2) && likely(max_sessions)) {
- struct ceph_mds_session **sessions = NULL;
- struct ceph_mds_session *s;
struct ceph_mds_request *req;
- int i;
-
- sessions = kcalloc(max_sessions, sizeof(s), GFP_KERNEL);
- if (!sessions) {
- err = -ENOMEM;
- goto out;
- }
spin_lock(&ci->i_unsafe_lock);
if (req1) {
@@ -2348,16 +2340,8 @@ static int flush_mdlog_and_wait_inode_unsafe_requests(struct inode *inode)
s = req->r_session;
if (!s)
continue;
- if (unlikely(s->s_mds >= max_sessions)) {
- spin_unlock(&ci->i_unsafe_lock);
- for (i = 0; i < max_sessions; i++) {
- s = sessions[i];
- if (s)
- ceph_put_mds_session(s);
- }
- kfree(sessions);
- goto retry;
- }
+ if (unlikely(s->s_mds >= max_sessions))
+ continue;
if (!sessions[s->s_mds]) {
s = ceph_get_mds_session(s);
sessions[s->s_mds] = s;
@@ -2370,16 +2354,8 @@ static int flush_mdlog_and_wait_inode_unsafe_requests(struct inode *inode)
s = req->r_session;
if (!s)
continue;
- if (unlikely(s->s_mds >= max_sessions)) {
- spin_unlock(&ci->i_unsafe_lock);
- for (i = 0; i < max_sessions; i++) {
- s = sessions[i];
- if (s)
- ceph_put_mds_session(s);
- }
- kfree(sessions);
- goto retry;
- }
+ if (unlikely(s->s_mds >= max_sessions))
+ continue;
if (!sessions[s->s_mds]) {
s = ceph_get_mds_session(s);
sessions[s->s_mds] = s;
@@ -2387,25 +2363,26 @@ static int flush_mdlog_and_wait_inode_unsafe_requests(struct inode *inode)
}
}
spin_unlock(&ci->i_unsafe_lock);
+ }
+ mutex_unlock(&mdsc->mutex);
- /* the auth MDS */
- spin_lock(&ci->i_ceph_lock);
- if (ci->i_auth_cap) {
- s = ci->i_auth_cap->session;
- if (!sessions[s->s_mds])
- sessions[s->s_mds] = ceph_get_mds_session(s);
- }
- spin_unlock(&ci->i_ceph_lock);
+ /* the auth MDS */
+ spin_lock(&ci->i_ceph_lock);
+ if (ci->i_auth_cap) {
+ s = ci->i_auth_cap->session;
+ if (!sessions[s->s_mds] &&
+ likely(s->s_mds < max_sessions))
+ sessions[s->s_mds] = ceph_get_mds_session(s);
+ }
+ spin_unlock(&ci->i_ceph_lock);
- /* send flush mdlog request to MDSes */
- for (i = 0; i < max_sessions; i++) {
- s = sessions[i];
- if (s) {
- send_flush_mdlog(s);
- ceph_put_mds_session(s);
- }
+ /* send flush mdlog request to MDSes */
+ for (i = 0; i < max_sessions; i++) {
+ s = sessions[i];
+ if (s) {
+ send_flush_mdlog(s);
+ ceph_put_mds_session(s);
}
- kfree(sessions);
}
dout("%s %p wait on tid %llu %llu\n", __func__,
@@ -2428,6 +2405,7 @@ static int flush_mdlog_and_wait_inode_unsafe_requests(struct inode *inode)
ceph_mdsc_put_request(req1);
if (req2)
ceph_mdsc_put_request(req2);
+ kfree(sessions);
return err;
}