diff options
| author | Andy Grover <andy.grover@oracle.com> | 2010-03-03 19:25:21 -0800 |
|---|---|---|
| committer | Andy Grover <andy.grover@oracle.com> | 2010-09-08 18:12:09 -0700 |
| commit | e32b4a70495aac6a612e13f4c21db09dd756ff2c (patch) | |
| tree | 2f0a7187e5ddb8a79d5bec0cefade198b99ea138 /net/rds | |
| parent | ff3d7d36134ef7138803734fdbf91cc986ea7976 (diff) | |
| download | kernel_samsung_aries-e32b4a70495aac6a612e13f4c21db09dd756ff2c.zip kernel_samsung_aries-e32b4a70495aac6a612e13f4c21db09dd756ff2c.tar.gz kernel_samsung_aries-e32b4a70495aac6a612e13f4c21db09dd756ff2c.tar.bz2 | |
RDS/IB: Do not wait for send ring to be empty on conn shutdown
Now that we are signaling send completions much less, we are likely
to have dirty entries in the send queue when the connection is
shut down (on rmmod, for example.) These are cleaned up a little
further down in conn_shutdown, but if we wait on the ring_empty_wait
for them, it'll never happen, and we hand on unload.
Signed-off-by: Andy Grover <andy.grover@oracle.com>
Diffstat (limited to 'net/rds')
| -rw-r--r-- | net/rds/ib_cm.c | 5 |
1 files changed, 4 insertions, 1 deletions
diff --git a/net/rds/ib_cm.c b/net/rds/ib_cm.c index b5b5ebb..1a91af7 100644 --- a/net/rds/ib_cm.c +++ b/net/rds/ib_cm.c @@ -620,8 +620,11 @@ void rds_ib_conn_shutdown(struct rds_connection *conn) ic->i_cm_id, err); } + /* + * Don't wait for the send ring to be empty -- there may be completed + * non-signaled entries sitting on there. We unmap these below. + */ wait_event(rds_ib_ring_empty_wait, - rds_ib_ring_empty(&ic->i_send_ring) && rds_ib_ring_empty(&ic->i_recv_ring)); if (ic->i_send_hdrs) |