From: Chris Horn <hornc@cray.com>
Date: Fri, 21 Feb 2020 00:01:45 +0000 (-0600)
Subject: LU-13277 lnet: Discovery thread can deadlock on shutdown
X-Git-Tag: 2.13.53~117
X-Git-Url: https://git.whamcloud.com/?p=fs%2Flustre-release.git;a=commitdiff_plain;h=82bb93410fc6f74e32ad74339ece5b4f62dc9967

LU-13277 lnet: Discovery thread can deadlock on shutdown

Drop the net_lock/EX before breaking out of the loop to avoid
deadlock.

Cray-bug-id: LUS-8525
Signed-off-by: Chris Horn <hornc@cray.com>
Change-Id: Ie62b55fa45b5795937eb1480a1fcabe295fed0ee
Reviewed-on: https://review.whamcloud.com/37675
Reviewed-by: Amir Shehata <ashehata@whamcloud.com>
Tested-by: jenkins <devops@whamcloud.com>
Reviewed-by: Serguei Smirnov <ssmirnov@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Shaun Tancheff <shaun.tancheff@hpe.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>
---

diff --git a/lnet/lnet/peer.c b/lnet/lnet/peer.c
index 5ea9206..fe4f028 100644
--- a/lnet/lnet/peer.c
+++ b/lnet/lnet/peer.c
@@ -3291,8 +3291,10 @@ static int lnet_peer_discovery(void *arg)
 			lnet_push_target_resize();
 
 		lnet_net_lock(LNET_LOCK_EX);
-		if (the_lnet.ln_dc_state == LNET_DC_STATE_STOPPING)
+		if (the_lnet.ln_dc_state == LNET_DC_STATE_STOPPING) {
+			lnet_net_unlock(LNET_LOCK_EX);
 			break;
+		}
 
 		/*
 		 * Process all incoming discovery work requests.  When