diff options
author | Yan, Zheng <zheng.z.yan@intel.com> | 2013-09-26 08:25:36 +0200 |
---|---|---|
committer | Sage Weil <sage@inktank.com> | 2013-11-23 20:01:04 +0100 |
commit | eb1b8af33c2e42a9a57fc0a7588f4a7b255d2e79 (patch) | |
tree | bfcabbe9697759b719e6b4b49b99e02e8e8de030 /fs | |
parent | ceph: handle race between cap reconnect and cap release (diff) | |
download | linux-eb1b8af33c2e42a9a57fc0a7588f4a7b255d2e79.tar.xz linux-eb1b8af33c2e42a9a57fc0a7588f4a7b255d2e79.zip |
ceph: cleanup aborted requests when re-sending requests.
Aborted requests usually get cleared when the reply is received.
If MDS crashes, no reply will be received. So we need to cleanup
aborted requests when re-sending requests.
Signed-off-by: Yan, Zheng <zheng.z.yan@intel.com>
Reviewed-by: Greg Farnum <greg@inktank.com>
Signed-off-by: Sage Weil <sage@inktank.com>
Diffstat (limited to 'fs')
-rw-r--r-- | fs/ceph/mds_client.c | 5 |
1 files changed, 4 insertions, 1 deletions
diff --git a/fs/ceph/mds_client.c b/fs/ceph/mds_client.c index 6d953ab0ac06..8ef79266d064 100644 --- a/fs/ceph/mds_client.c +++ b/fs/ceph/mds_client.c @@ -1872,8 +1872,11 @@ static int __do_request(struct ceph_mds_client *mdsc, int mds = -1; int err = -EAGAIN; - if (req->r_err || req->r_got_result) + if (req->r_err || req->r_got_result) { + if (req->r_aborted) + __unregister_request(mdsc, req); goto out; + } if (req->r_timeout && time_after_eq(jiffies, req->r_started + req->r_timeout)) { |