summaryrefslogtreecommitdiffstats
diff options
context:
space:
mode:
authorMarcelo Ricardo Leitner <marcelo.leitner@gmail.com>2016-10-08 10:14:37 -0300
committerDavid Teigland <teigland@redhat.com>2016-10-10 09:54:00 -0500
commit3a8db79889ce16930aff19b818f5b09651bb7644 (patch)
treee4395e493ec475bbcb9402e647e9429b56be3db3
parentc8d2bc9bc39ebea8437fd974fdbc21847bb897a3 (diff)
downloadlinux-3a8db79889ce16930aff19b818f5b09651bb7644.tar.bz2
dlm: free workqueues after the connections
After backporting commit ee44b4bc054a ("dlm: use sctp 1-to-1 API") series to a kernel with an older workqueue which didn't use RCU yet, it was noticed that we are freeing the workqueues in dlm_lowcomms_stop() too early as free_conn() will try to access that memory for canceling the queued works if any. This issue was introduced by commit 0d737a8cfd83 as before it such attempt to cancel the queued works wasn't performed, so the issue was not present. This patch fixes it by simply inverting the free order. Cc: stable@vger.kernel.org Fixes: 0d737a8cfd83 ("dlm: fix race while closing connections") Signed-off-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com> Signed-off-by: David Teigland <teigland@redhat.com>
-rw-r--r--fs/dlm/lowcomms.c8
1 files changed, 2 insertions, 6 deletions
diff --git a/fs/dlm/lowcomms.c b/fs/dlm/lowcomms.c
index 963016c8f3d1..609998de533e 100644
--- a/fs/dlm/lowcomms.c
+++ b/fs/dlm/lowcomms.c
@@ -1656,16 +1656,12 @@ void dlm_lowcomms_stop(void)
mutex_lock(&connections_lock);
dlm_allow_conn = 0;
foreach_conn(stop_conn);
+ clean_writequeues();
+ foreach_conn(free_conn);
mutex_unlock(&connections_lock);
work_stop();
- mutex_lock(&connections_lock);
- clean_writequeues();
-
- foreach_conn(free_conn);
-
- mutex_unlock(&connections_lock);
kmem_cache_destroy(con_cache);
}