ceph: flush the mdlog before waiting on unsafe reqs

[platform/kernel/linux-rpi.git] / fs / ceph / caps.c

diff --git a/fs/ceph/caps.c b/fs/ceph/caps.c
index c2d6541..d5a4655 100644 (file)
--- a/fs/ceph/caps.c
+++ b/fs/ceph/caps.c
@@ -2219,6 +2219,7 @@ static int caps_are_flushed(struct inode *inode, u64 flush_tid)
  */
 static int unsafe_request_wait(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;
        int ret, err = 0;
@@ -2238,6 +2239,81 @@ static int unsafe_request_wait(struct inode *inode)
        }
        spin_unlock(&ci->i_unsafe_lock);
 
+       /*
+        * 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.
+        */
+       if (req1 || req2) {
+               struct ceph_mds_session **sessions = NULL;
+               struct ceph_mds_session *s;
+               struct ceph_mds_request *req;
+               unsigned int max;
+               int i;
+
+               /*
+                * The mdsc->max_sessions is unlikely to be changed
+                * mostly, here we will retry it by reallocating the
+                * sessions arrary memory to get rid of the mdsc->mutex
+                * lock.
+                */
+retry:
+               max = mdsc->max_sessions;
+               sessions = krealloc(sessions, max * sizeof(s), __GFP_ZERO);
+               if (!sessions)
+                       return -ENOMEM;
+
+               spin_lock(&ci->i_unsafe_lock);
+               if (req1) {
+                       list_for_each_entry(req, &ci->i_unsafe_dirops,
+                                           r_unsafe_dir_item) {
+                               s = req->r_session;
+                               if (unlikely(s->s_mds > max)) {
+                                       spin_unlock(&ci->i_unsafe_lock);
+                                       goto retry;
+                               }
+                               if (!sessions[s->s_mds]) {
+                                       s = ceph_get_mds_session(s);
+                                       sessions[s->s_mds] = s;
+                               }
+                       }
+               }
+               if (req2) {
+                       list_for_each_entry(req, &ci->i_unsafe_iops,
+                                           r_unsafe_target_item) {
+                               s = req->r_session;
+                               if (unlikely(s->s_mds > max)) {
+                                       spin_unlock(&ci->i_unsafe_lock);
+                                       goto retry;
+                               }
+                               if (!sessions[s->s_mds]) {
+                                       s = ceph_get_mds_session(s);
+                                       sessions[s->s_mds] = s;
+                               }
+                       }
+               }
+               spin_unlock(&ci->i_unsafe_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])
+                             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; i++) {
+                       s = sessions[i];
+                       if (s) {
+                               send_flush_mdlog(s);
+                               ceph_put_mds_session(s);
+                       }
+               }
+               kfree(sessions);
+       }
+
        dout("unsafe_request_wait %p wait on tid %llu %llu\n",
             inode, req1 ? req1->r_tid : 0ULL, req2 ? req2->r_tid : 0ULL);
        if (req1) {
@@ -4226,33 +4302,9 @@ static void flush_dirty_session_caps(struct ceph_mds_session *s)
        dout("flush_dirty_caps done\n");
 }
 
-static void iterate_sessions(struct ceph_mds_client *mdsc,
-                            void (*cb)(struct ceph_mds_session *))
-{
-       int mds;
-
-       mutex_lock(&mdsc->mutex);
-       for (mds = 0; mds < mdsc->max_sessions; ++mds) {
-               struct ceph_mds_session *s;
-
-               if (!mdsc->sessions[mds])
-                       continue;
-
-               s = ceph_get_mds_session(mdsc->sessions[mds]);
-               if (!s)
-                       continue;
-
-               mutex_unlock(&mdsc->mutex);
-               cb(s);
-               ceph_put_mds_session(s);
-               mutex_lock(&mdsc->mutex);
-       }
-       mutex_unlock(&mdsc->mutex);
-}
-
 void ceph_flush_dirty_caps(struct ceph_mds_client *mdsc)
 {
-       iterate_sessions(mdsc, flush_dirty_session_caps);
+       ceph_mdsc_iterate_sessions(mdsc, flush_dirty_session_caps, true);
 }
 
 void __ceph_touch_fmode(struct ceph_inode_info *ci,