From: Joseph Qi Date: Thu, 9 Oct 2014 22:25:13 +0000 (-0700) Subject: ocfs2: fix deadlock between o2hb thread and o2net_wq X-Git-Tag: v5.15~17103^2~143 X-Git-Url: https://git.proxmox.com/?a=commitdiff_plain;h=70e82a12dbfa3acbff41be08a36e8be4578878c9;p=mirror_ubuntu-kernels.git ocfs2: fix deadlock between o2hb thread and o2net_wq The following case may lead to o2net_wq and o2hb thread deadlock on o2hb_callback_sem. Currently there are 2 nodes say N1, N2 in the cluster. And N2 down, at the same time, N3 tries to join the cluster. So N1 will handle node down (N2) and join (N3) simultaneously. o2hb o2net_wq ->o2hb_do_disk_heartbeat ->o2hb_check_slot ->o2hb_run_event_list ->o2hb_fire_callbacks ->down_write(&o2hb_callback_sem) ->o2net_hb_node_down_cb ->flush_workqueue(o2net_wq) ->o2net_process_message ->dlm_query_join_handler ->o2hb_check_node_heartbeating ->o2hb_fill_node_map ->down_read(&o2hb_callback_sem) No need to take o2hb_callback_sem in dlm_query_join_handler, o2hb_live_lock is enough to protect live node map. Signed-off-by: Joseph Qi Cc: xMark Fasheh Cc: Joel Becker Cc: jiangyiwen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- diff --git a/fs/ocfs2/cluster/heartbeat.c b/fs/ocfs2/cluster/heartbeat.c index 73039295d0d1..d13385448168 100644 --- a/fs/ocfs2/cluster/heartbeat.c +++ b/fs/ocfs2/cluster/heartbeat.c @@ -2572,6 +2572,25 @@ int o2hb_check_node_heartbeating(u8 node_num) } EXPORT_SYMBOL_GPL(o2hb_check_node_heartbeating); +int o2hb_check_node_heartbeating_no_sem(u8 node_num) +{ + unsigned long testing_map[BITS_TO_LONGS(O2NM_MAX_NODES)]; + unsigned long flags; + + spin_lock_irqsave(&o2hb_live_lock, flags); + o2hb_fill_node_map_from_callback(testing_map, sizeof(testing_map)); + spin_unlock_irqrestore(&o2hb_live_lock, flags); + if (!test_bit(node_num, testing_map)) { + mlog(ML_HEARTBEAT, + "node (%u) does not have heartbeating enabled.\n", + node_num); + return 0; + } + + return 1; +} +EXPORT_SYMBOL_GPL(o2hb_check_node_heartbeating_no_sem); + int o2hb_check_node_heartbeating_from_callback(u8 node_num) { unsigned long testing_map[BITS_TO_LONGS(O2NM_MAX_NODES)]; diff --git a/fs/ocfs2/cluster/heartbeat.h b/fs/ocfs2/cluster/heartbeat.h index 00ad8e8fea51..3ef5137dc362 100644 --- a/fs/ocfs2/cluster/heartbeat.h +++ b/fs/ocfs2/cluster/heartbeat.h @@ -80,6 +80,7 @@ void o2hb_fill_node_map(unsigned long *map, void o2hb_exit(void); int o2hb_init(void); int o2hb_check_node_heartbeating(u8 node_num); +int o2hb_check_node_heartbeating_no_sem(u8 node_num); int o2hb_check_node_heartbeating_from_callback(u8 node_num); int o2hb_check_local_node_heartbeating(void); void o2hb_stop_all_regions(void); diff --git a/fs/ocfs2/dlm/dlmdomain.c b/fs/ocfs2/dlm/dlmdomain.c index 257a6dfe3f13..02d315fef432 100644 --- a/fs/ocfs2/dlm/dlmdomain.c +++ b/fs/ocfs2/dlm/dlmdomain.c @@ -839,7 +839,7 @@ static int dlm_query_join_handler(struct o2net_msg *msg, u32 len, void *data, * to back off and try again. This gives heartbeat a chance * to catch up. */ - if (!o2hb_check_node_heartbeating(query->node_idx)) { + if (!o2hb_check_node_heartbeating_no_sem(query->node_idx)) { mlog(0, "node %u is not in our live map yet\n", query->node_idx);