[Devel] [PATCH RHEL7 COMMIT] fence-watchdog: panic after 30 sec of reboot/halt

[Konstantin Khorenko]

The commit is pushed to "branch-rh7-3.10.0-514.26.1.vz7.33.x-ovz" and will appear at https://src.openvz.org/scm/ovz/vzkernel.git
after rh7-3.10.0-514.26.1.vz7.33.12
------>
commit ad1311fb4f6f439aa6c4596e99a36caa4fcfe489
Author: Pavel Tikhomirov <ptikhomirov at virtuozzo.com>
Date:   Fri Jul 21 12:20:56 2017 +0400

    fence-watchdog: panic after 30 sec of reboot/halt
    
    As we do reboot and halt actions in scope of scheduled worker
    it can never happen if scheduling does not work properly, so
    panic in case that previous action was not successful.
    
    https://jira.sw.ru/browse/PSBM-54747
    
    Signed-off-by: Pavel Tikhomirov <ptikhomirov at virtuozzo.com>
    Reviewed-by: Kirill Tkhai <ktkhai at virtuozzo.com>
---
 kernel/fence-watchdog.c | 20 +++++++++++++++++---
 1 file changed, 17 insertions(+), 3 deletions(-)

diff --git a/kernel/fence-watchdog.c b/kernel/fence-watchdog.c
index c35b3ad..40d258a 100644
--- a/kernel/fence-watchdog.c
+++ b/kernel/fence-watchdog.c
@@ -42,7 +42,14 @@ const char *action_names[] = {"crash", "reboot", "halt", "netfilter", NULL};
 
 DEFINE_VVAR(volatile unsigned long, fence_wdog_jiffies64) = MAX_U64;
 static int fence_wdog_action = FENCE_WDOG_CRASH;
-static atomic_t not_fenced = ATOMIC_INIT(-1);
+
+enum {
+	NOT_FENCED = 0,
+	FENCED = 1,
+	FENCED_TIMEOUT = 2,
+};
+
+static atomic_t fence_stage = ATOMIC_INIT(NOT_FENCED);
 static char fence_wdog_log_path[PATH_MAX] = "/fence_wdog.log";
 
 #define SECS_PER_MIN	60
@@ -132,19 +139,26 @@ static DECLARE_WORK(halt_or_reboot_work, do_halt_or_reboot);
 
 void fence_wdog_do_fence(void)
 {
-	if (fence_wdog_action == FENCE_WDOG_CRASH)
+	if (fence_wdog_action == FENCE_WDOG_CRASH ||
+			atomic_read(&fence_stage) == FENCED_TIMEOUT)
 		panic("fence-watchdog: %s\n",
 		      action_names[fence_wdog_action]);
 	else
 		schedule_work(&halt_or_reboot_work);
 }
 
+#define FENCE_WDOG_TIMEOUT 30
+
 inline int fence_wdog_check_timer(void)
 {
 	if (unlikely(get_jiffies_64() > fence_wdog_jiffies64 &&
 			fence_wdog_action != FENCE_WDOG_NETFILTER)) {
-		if (atomic_inc_not_zero(&not_fenced))
+		if (atomic_cmpxchg(&fence_stage, NOT_FENCED, FENCED) == NOT_FENCED
+		    || (get_jiffies_64() > fence_wdog_jiffies64
+		    + FENCE_WDOG_TIMEOUT * HZ
+		    && atomic_cmpxchg(&fence_stage, FENCED, FENCED_TIMEOUT) == FENCED))
 			fence_wdog_do_fence();
+
 		return 1;
 	}