ceph: handle the new nfiles/nsubdirs fields in cap message

Without these new fields, stale st_size is returned in following
case.

1. MDS modifies a directory
2. MDS issues CEPH_CAP_ANY_SHARED to client
3. The client satifies stat(2) by its cached metadata. set st_size
   to "i_files + i_subdirs".

Link: http://tracker.ceph.com/issues/23855
Signed-off-by: "Yan, Zheng" <zyan@redhat.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>

diff --git a/fs/ceph/caps.c b/fs/ceph/caps.c
index de7b7a3..477b822 100644 (file)
--- a/fs/ceph/caps.c
+++ b/fs/ceph/caps.c
@@ -3030,6 +3030,10 @@ struct cap_extra_info {
        u64 inline_version;
        void *inline_data;
        u32 inline_len;
+       /* dirstat */
+       bool dirstat_valid;
+       u64 nfiles;
+       u64 nsubdirs;
        /* currently issued */
        int issued;
 };
@@ -3154,6 +3158,11 @@ static void handle_cap_grant(struct inode *inode,
                                    &ctime, &mtime, &atime);
        }
 
+       if ((newcaps & CEPH_CAP_FILE_SHARED) && extra_info->dirstat_valid) {
+               ci->i_files = extra_info->nfiles;
+               ci->i_subdirs = extra_info->nsubdirs;
+       }
+
        if (newcaps & (CEPH_CAP_ANY_FILE_RD | CEPH_CAP_ANY_FILE_WR)) {
                /* file layout may have changed */
                s64 old_pool = ci->i_layout.pool_id;
@@ -3741,6 +3750,7 @@ void ceph_handle_caps(struct ceph_mds_session *session,
        struct ceph_mds_cap_peer *peer = NULL;
        struct ceph_snap_realm *realm = NULL;
        int op;
+       int msg_version = le16_to_cpu(msg->hdr.version);
        u32 seq, mseq;
        struct ceph_vino vino;
        void *snaptrace;
@@ -3765,7 +3775,7 @@ void ceph_handle_caps(struct ceph_mds_session *session,
        snaptrace_len = le32_to_cpu(h->snap_trace_len);
        p = snaptrace + snaptrace_len;
 
-       if (le16_to_cpu(msg->hdr.version) >= 2) {
+       if (msg_version >= 2) {
                u32 flock_len;
                ceph_decode_32_safe(&p, end, flock_len, bad);
                if (p + flock_len > end)
@@ -3773,7 +3783,7 @@ void ceph_handle_caps(struct ceph_mds_session *session,
                p += flock_len;
        }
 
-       if (le16_to_cpu(msg->hdr.version) >= 3) {
+       if (msg_version >= 3) {
                if (op == CEPH_CAP_OP_IMPORT) {
                        if (p + sizeof(*peer) > end)
                                goto bad;
@@ -3785,7 +3795,7 @@ void ceph_handle_caps(struct ceph_mds_session *session,
                }
        }
 
-       if (le16_to_cpu(msg->hdr.version) >= 4) {
+       if (msg_version >= 4) {
                ceph_decode_64_safe(&p, end, extra_info.inline_version, bad);
                ceph_decode_32_safe(&p, end, extra_info.inline_len, bad);
                if (p + extra_info.inline_len > end)
@@ -3794,7 +3804,7 @@ void ceph_handle_caps(struct ceph_mds_session *session,
                p += extra_info.inline_len;
        }
 
-       if (le16_to_cpu(msg->hdr.version) >= 5) {
+       if (msg_version >= 5) {
                struct ceph_osd_client  *osdc = &mdsc->fsc->client->osdc;
                u32                     epoch_barrier;
 
@@ -3802,7 +3812,7 @@ void ceph_handle_caps(struct ceph_mds_session *session,
                ceph_osdc_update_epoch_barrier(osdc, epoch_barrier);
        }
 
-       if (le16_to_cpu(msg->hdr.version) >= 8) {
+       if (msg_version >= 8) {
                u64 flush_tid;
                u32 caller_uid, caller_gid;
                u32 pool_ns_len;
@@ -3822,6 +3832,25 @@ void ceph_handle_caps(struct ceph_mds_session *session,
                }
        }
 
+       if (msg_version >= 11) {
+               struct ceph_timespec *btime;
+               u64 change_attr;
+               u32 flags;
+
+               /* version >= 9 */
+               if (p + sizeof(*btime) > end)
+                       goto bad;
+               btime = p;
+               p += sizeof(*btime);
+               ceph_decode_64_safe(&p, end, change_attr, bad);
+               /* version >= 10 */
+               ceph_decode_32_safe(&p, end, flags, bad);
+               /* version >= 11 */
+               extra_info.dirstat_valid = true;
+               ceph_decode_64_safe(&p, end, extra_info.nfiles, bad);
+               ceph_decode_64_safe(&p, end, extra_info.nsubdirs, bad);
+       }
+
        /* lookup ino */
        inode = ceph_find_inode(mdsc->fsc->sb, vino);
        ci = ceph_inode(inode);