Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/sage/ceph-client

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/sage/ceph-client:
  ceph: fix crush device 'out' threshold to 1.0, not 0.1
  ceph: fix caps usage accounting for import (non-reserved) case
  ceph: only release clean, unused caps with mds requests
  ceph: fix crush CHOOSE_LEAF when type is already a leaf
  ceph: fix crush recursion
  ceph: fix caps debugfs entry
  ceph: delay umount until all mds requests drop inode+dentry refs
  ceph: handle splice_dentry/d_materialize_unique error in readdir_prepopulate
  ceph: fix crush map update decoding
  ceph: fix message memory leak, uninitialized variable
  ceph: fix map handler error path
  ceph: some endianity fixes

10 files changed, 69 insertions, 33 deletions

diff --git a/fs/ceph/auth_x.c b/fs/ceph/auth_x.c
index 83d4d2785ffe..3fe49042d8ad 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) {
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..3ab79f6c4ce8 100644
--- a/fs/ceph/mds_client.c
+++ b/fs/ceph/mds_client.c
@@ -2783,6 +2783,12 @@ 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/messenger.c b/fs/ceph/messenger.c
index 64b8b1f7863d..9ad43a310a41 100644
--- a/fs/ceph/messenger.c
+++ b/fs/ceph/messenger.c
@@ -657,7 +657,7 @@ 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);
@@ -1396,10 +1396,12 @@ 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;
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..50ce64ebd330 100644
--- a/fs/ceph/osdmap.c
+++ b/fs/ceph/osdmap.c
@@ -707,6 +707,7 @@ 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? */