Whamcloud - gitweb
LU-10805 libcfs: timer_setup() API changes
[fs/lustre-release.git] / libcfs / libcfs / watchdog.c
index 6ce41c3..4c437ab 100644 (file)
  *
  * You should have received a copy of the GNU General Public License
  * version 2 along with this program; If not, see
- * http://www.sun.com/software/products/lustre/docs/GPLv2.pdf
- *
- * Please contact Sun Microsystems, Inc., 4150 Network Circle, Santa Clara,
- * CA 95054 USA or visit www.sun.com if you need additional information or
- * have any questions.
+ * http://www.gnu.org/licenses/gpl-2.0.html
  *
  * GPL HEADER END
  */
 /*
  * Copyright (c) 2008, 2010, Oracle and/or its affiliates. All rights reserved.
  * Use is subject to license terms.
+ *
+ * Copyright (c) 2012, 2017, Intel Corporation.
  */
 /*
  * This file is part of Lustre, http://www.lustre.org/
 
 #define DEBUG_SUBSYSTEM S_LNET
 
+#include <linux/kthread.h>
 #include <libcfs/libcfs.h>
 #include "tracefile.h"
 
+#ifndef WITH_WATCHDOG
+#define WITH_WATCHDOG
+#endif
+
 struct lc_watchdog {
-        cfs_spinlock_t  lcw_lock;     /* check or change lcw_list */
-        int             lcw_refcount; /* must hold lcw_pending_timers_lock */
-        cfs_timer_t     lcw_timer;    /* kernel timer */
-        cfs_list_t      lcw_list;     /* chain on pending list */
-        cfs_time_t      lcw_last_touched; /* last touched stamp */
-        cfs_task_t     *lcw_task;     /* owner task */
-        void          (*lcw_callback)(pid_t, void *);
-        void           *lcw_data;
-
-        pid_t           lcw_pid;
-
-        enum {
-                LC_WATCHDOG_DISABLED,
-                LC_WATCHDOG_ENABLED,
-                LC_WATCHDOG_EXPIRED
-        } lcw_state;
+       spinlock_t              lcw_lock;       /* check or change lcw_list */
+       int                     lcw_refcount;   /* must hold lcw_pending_timers_lock */
+       struct timer_list       lcw_timer;      /* kernel timer */
+       struct list_head        lcw_list;       /* chain on pending list */
+       ktime_t                 lcw_last_touched;/* last touched stamp */
+       struct task_struct     *lcw_task;       /* owner task */
+       void                    (*lcw_callback)(pid_t, void *);
+       void                    *lcw_data;
+
+       pid_t                   lcw_pid;
+
+       enum {
+               LC_WATCHDOG_DISABLED,
+               LC_WATCHDOG_ENABLED,
+               LC_WATCHDOG_EXPIRED
+       } lcw_state;
 };
 
 #ifdef WITH_WATCHDOG
@@ -66,9 +69,9 @@ struct lc_watchdog {
  * and lcw_stop_completion when it exits.
  * Wake lcw_event_waitq to signal timer callback dispatches.
  */
-static cfs_completion_t lcw_start_completion;
-static cfs_completion_t  lcw_stop_completion;
-static cfs_waitq_t lcw_event_waitq;
+static struct completion lcw_start_completion;
+static struct completion  lcw_stop_completion;
+static wait_queue_head_t lcw_event_waitq;
 
 /*
  * Set this and wake lcw_event_waitq to stop the dispatcher.
@@ -84,49 +87,40 @@ static unsigned long lcw_flags = 0;
  * When it hits 0, we stop the dispatcher.
  */
 static __u32         lcw_refcount = 0;
-static CFS_DEFINE_MUTEX(lcw_refcount_mutex);
+static DEFINE_MUTEX(lcw_refcount_mutex);
 
 /*
  * List of timers that have fired that need their callbacks run by the
  * dispatcher.
  */
 /* BH lock! */
-static cfs_spinlock_t lcw_pending_timers_lock = CFS_SPIN_LOCK_UNLOCKED;
-static cfs_list_t lcw_pending_timers = \
-        CFS_LIST_HEAD_INIT(lcw_pending_timers);
+static DEFINE_SPINLOCK(lcw_pending_timers_lock);
+static struct list_head lcw_pending_timers = LIST_HEAD_INIT(lcw_pending_timers);
 
 /* Last time a watchdog expired */
-static cfs_time_t lcw_last_watchdog_time;
+static time64_t lcw_last_watchdog_time;
 static int lcw_recent_watchdog_count;
 
 static void
 lcw_dump(struct lc_watchdog *lcw)
 {
         ENTRY;
-#if defined(HAVE_TASKLIST_LOCK)
-        cfs_read_lock(&tasklist_lock);
-#else
         rcu_read_lock();
-#endif
        if (lcw->lcw_task == NULL) {
-                LCONSOLE_WARN("Process " LPPID " was not found in the task "
+               LCONSOLE_WARN("Process %d was not found in the task "
                               "list; watchdog callback may be incomplete\n",
                               (int)lcw->lcw_pid);
         } else {
                 libcfs_debug_dumpstack(lcw->lcw_task);
         }
 
-#if defined(HAVE_TASKLIST_LOCK)
-        cfs_read_unlock(&tasklist_lock);
-#else
         rcu_read_unlock();
-#endif
         EXIT;
 }
 
-static void lcw_cb(ulong_ptr_t data)
+static void lcw_cb(cfs_timer_cb_arg_t data)
 {
-        struct lc_watchdog *lcw = (struct lc_watchdog *)data;
+       struct lc_watchdog *lcw = cfs_from_timer(lcw, data, lcw_timer);
         ENTRY;
 
         if (lcw->lcw_state != LC_WATCHDOG_ENABLED) {
@@ -136,148 +130,144 @@ static void lcw_cb(ulong_ptr_t data)
 
         lcw->lcw_state = LC_WATCHDOG_EXPIRED;
 
-        cfs_spin_lock_bh(&lcw->lcw_lock);
-        LASSERT(cfs_list_empty(&lcw->lcw_list));
+       spin_lock_bh(&lcw->lcw_lock);
+       LASSERT(list_empty(&lcw->lcw_list));
 
-        cfs_spin_lock_bh(&lcw_pending_timers_lock);
-        lcw->lcw_refcount++; /* +1 for pending list */
-        cfs_list_add(&lcw->lcw_list, &lcw_pending_timers);
-        cfs_waitq_signal(&lcw_event_waitq);
+       spin_lock_bh(&lcw_pending_timers_lock);
+       lcw->lcw_refcount++; /* +1 for pending list */
+       list_add(&lcw->lcw_list, &lcw_pending_timers);
+       wake_up(&lcw_event_waitq);
 
-        cfs_spin_unlock_bh(&lcw_pending_timers_lock);
-        cfs_spin_unlock_bh(&lcw->lcw_lock);
-        EXIT;
+       spin_unlock_bh(&lcw_pending_timers_lock);
+       spin_unlock_bh(&lcw->lcw_lock);
+       EXIT;
 }
 
 static int is_watchdog_fired(void)
 {
-        int rc;
+       int rc;
 
-        if (cfs_test_bit(LCW_FLAG_STOP, &lcw_flags))
-                return 1;
+       if (test_bit(LCW_FLAG_STOP, &lcw_flags))
+               return 1;
 
-        cfs_spin_lock_bh(&lcw_pending_timers_lock);
-        rc = !cfs_list_empty(&lcw_pending_timers);
-        cfs_spin_unlock_bh(&lcw_pending_timers_lock);
-        return rc;
+       spin_lock_bh(&lcw_pending_timers_lock);
+       rc = !list_empty(&lcw_pending_timers);
+       spin_unlock_bh(&lcw_pending_timers_lock);
+       return rc;
 }
 
 static void lcw_dump_stack(struct lc_watchdog *lcw)
 {
-        cfs_time_t      current_time;
-        cfs_duration_t  delta_time;
-        struct timeval  timediff;
-
-        current_time = cfs_time_current();
-        delta_time = cfs_time_sub(current_time, lcw->lcw_last_touched);
-        cfs_duration_usec(delta_time, &timediff);
-
-        /*
-         * Check to see if we should throttle the watchdog timer to avoid
-         * too many dumps going to the console thus triggering an NMI.
-         */
-        delta_time = cfs_duration_sec(cfs_time_sub(current_time,
-                                                   lcw_last_watchdog_time));
-
-        if (delta_time < libcfs_watchdog_ratelimit &&
-            lcw_recent_watchdog_count > 3) {
-                LCONSOLE_WARN("Service thread pid %u was inactive for "
-                              "%lu.%.02lus. Watchdog stack traces are limited "
-                              "to 3 per %d seconds, skipping this one.\n",
-                              (int)lcw->lcw_pid,
-                              timediff.tv_sec,
-                              timediff.tv_usec / 10000,
-                              libcfs_watchdog_ratelimit);
-        } else {
-                if (delta_time < libcfs_watchdog_ratelimit) {
-                        lcw_recent_watchdog_count++;
-                } else {
-                        memcpy(&lcw_last_watchdog_time, &current_time,
-                               sizeof(current_time));
-                        lcw_recent_watchdog_count = 0;
-                }
-
-                LCONSOLE_WARN("Service thread pid %u was inactive for "
-                              "%lu.%.02lus. The thread might be hung, or it "
-                              "might only be slow and will resume later. "
-                              "Dumping the stack trace for debugging purposes:"
-                              "\n",
-                              (int)lcw->lcw_pid,
-                              timediff.tv_sec,
-                              timediff.tv_usec / 10000);
-                lcw_dump(lcw);
-        }
+       time64_t current_time = ktime_get_seconds();
+       struct timespec64 timediff;
+       time64_t delta_time;
+
+       timediff = ktime_to_timespec64(ktime_sub(ktime_get(),
+                                      lcw->lcw_last_touched));
+
+       /* LU-9235: Don't dump stack if the thread is just touched. */
+       if (timediff.tv_sec == 0)
+               return;
+
+       /*
+        * Check to see if we should throttle the watchdog timer to avoid
+        * too many dumps going to the console thus triggering an NMI.
+        */
+       delta_time = current_time - lcw_last_watchdog_time;
+       if (delta_time < libcfs_watchdog_ratelimit &&
+           lcw_recent_watchdog_count > 3) {
+               LCONSOLE_WARN("Service thread pid %u was inactive for %lu.%.02lus. Watchdog stack traces are limited to 3 per %d seconds, skipping this one.\n",
+                             (int)lcw->lcw_pid,
+                             timediff.tv_sec,
+                             timediff.tv_nsec / (NSEC_PER_SEC / 100),
+                             libcfs_watchdog_ratelimit);
+       } else {
+               if (delta_time < libcfs_watchdog_ratelimit) {
+                       lcw_recent_watchdog_count++;
+               } else {
+                       memcpy(&lcw_last_watchdog_time, &current_time,
+                              sizeof(current_time));
+                       lcw_recent_watchdog_count = 0;
+               }
+
+               LCONSOLE_WARN("Service thread pid %u was inactive for %lu.%.02lus. The thread might be hung, or it might only be slow and will resume later. Dumping the stack trace for debugging purposes:\n",
+                             (int)lcw->lcw_pid,
+                             timediff.tv_sec,
+                             timediff.tv_nsec / (NSEC_PER_SEC / 100));
+               lcw_dump(lcw);
+       }
+}
+
+/*
+ * Provided watchdog handlers
+ */
+
+static void lc_watchdog_dumplog(pid_t pid, void *data)
+{
+       libcfs_debug_dumplog_internal((void *)((uintptr_t)pid));
 }
 
 static int lcw_dispatch_main(void *data)
 {
         int                 rc = 0;
-        unsigned long       flags;
         struct lc_watchdog *lcw;
-        CFS_LIST_HEAD      (zombies);
+       struct list_head zombies = LIST_HEAD_INIT(zombies);
 
         ENTRY;
 
-        cfs_daemonize("lc_watchdogd");
-
-        SIGNAL_MASK_LOCK(current, flags);
-        sigfillset(&current->blocked);
-        RECALC_SIGPENDING;
-        SIGNAL_MASK_UNLOCK(current, flags);
-
-        cfs_complete(&lcw_start_completion);
+       complete(&lcw_start_completion);
 
         while (1) {
                 int dumplog = 1;
 
-                cfs_wait_event_interruptible(lcw_event_waitq,
-                                             is_watchdog_fired(), rc);
+               rc = wait_event_interruptible(lcw_event_waitq,
+                                             is_watchdog_fired());
                 CDEBUG(D_INFO, "Watchdog got woken up...\n");
-                if (cfs_test_bit(LCW_FLAG_STOP, &lcw_flags)) {
-                        CDEBUG(D_INFO, "LCW_FLAG_STOP was set, shutting down...\n");
-
-                        cfs_spin_lock_bh(&lcw_pending_timers_lock);
-                        rc = !cfs_list_empty(&lcw_pending_timers);
-                        cfs_spin_unlock_bh(&lcw_pending_timers_lock);
-                        if (rc) {
-                                CERROR("pending timers list was not empty at "
-                                       "time of watchdog dispatch shutdown\n");
-                        }
-                        break;
-                }
-
-                cfs_spin_lock_bh(&lcw_pending_timers_lock);
-                while (!cfs_list_empty(&lcw_pending_timers)) {
-                        int is_dumplog;
-
-                        lcw = cfs_list_entry(lcw_pending_timers.next,
-                                             struct lc_watchdog, lcw_list);
-                        /* +1 ref for callback to make sure lwc wouldn't be
-                         * deleted after releasing lcw_pending_timers_lock */
-                        lcw->lcw_refcount++;
-                        cfs_spin_unlock_bh(&lcw_pending_timers_lock);
-
-                        /* lock ordering */
-                        cfs_spin_lock_bh(&lcw->lcw_lock);
-                        cfs_spin_lock_bh(&lcw_pending_timers_lock);
-
-                        if (cfs_list_empty(&lcw->lcw_list)) {
-                                /* already removed from pending list */
-                                lcw->lcw_refcount--; /* -1 ref for callback */
-                                if (lcw->lcw_refcount == 0)
-                                        cfs_list_add(&lcw->lcw_list, &zombies);
-                                cfs_spin_unlock_bh(&lcw->lcw_lock);
-                                /* still hold lcw_pending_timers_lock */
-                                continue;
-                        }
-
-                        cfs_list_del_init(&lcw->lcw_list);
-                        lcw->lcw_refcount--; /* -1 ref for pending list */
-
-                        cfs_spin_unlock_bh(&lcw_pending_timers_lock);
-                        cfs_spin_unlock_bh(&lcw->lcw_lock);
-
-                        CDEBUG(D_INFO, "found lcw for pid " LPPID "\n",
+               if (test_bit(LCW_FLAG_STOP, &lcw_flags)) {
+                       CDEBUG(D_INFO, "LCW_FLAG_STOP set, shutting down...\n");
+
+                       spin_lock_bh(&lcw_pending_timers_lock);
+                       rc = !list_empty(&lcw_pending_timers);
+                       spin_unlock_bh(&lcw_pending_timers_lock);
+                       if (rc) {
+                               CERROR("pending timers list was not empty at "
+                                      "time of watchdog dispatch shutdown\n");
+                       }
+                       break;
+               }
+
+               spin_lock_bh(&lcw_pending_timers_lock);
+               while (!list_empty(&lcw_pending_timers)) {
+                       int is_dumplog;
+
+                       lcw = list_entry(lcw_pending_timers.next,
+                                        struct lc_watchdog, lcw_list);
+                       /* +1 ref for callback to make sure lwc wouldn't be
+                        * deleted after releasing lcw_pending_timers_lock */
+                       lcw->lcw_refcount++;
+                       spin_unlock_bh(&lcw_pending_timers_lock);
+
+                       /* lock ordering */
+                       spin_lock_bh(&lcw->lcw_lock);
+                       spin_lock_bh(&lcw_pending_timers_lock);
+
+                       if (list_empty(&lcw->lcw_list)) {
+                               /* already removed from pending list */
+                               lcw->lcw_refcount--; /* -1 ref for callback */
+                               if (lcw->lcw_refcount == 0)
+                                       list_add(&lcw->lcw_list, &zombies);
+                               spin_unlock_bh(&lcw->lcw_lock);
+                               /* still hold lcw_pending_timers_lock */
+                               continue;
+                       }
+
+                       list_del_init(&lcw->lcw_list);
+                       lcw->lcw_refcount--; /* -1 ref for pending list */
+
+                       spin_unlock_bh(&lcw_pending_timers_lock);
+                       spin_unlock_bh(&lcw->lcw_lock);
+
+                       CDEBUG(D_INFO, "found lcw for pid %d\n",
                                lcw->lcw_pid);
                         lcw_dump_stack(lcw);
 
@@ -289,65 +279,67 @@ static int lcw_dispatch_main(void *data)
                                         dumplog = 0;
                         }
 
-                        cfs_spin_lock_bh(&lcw_pending_timers_lock);
-                        lcw->lcw_refcount--; /* -1 ref for callback */
-                        if (lcw->lcw_refcount == 0)
-                                cfs_list_add(&lcw->lcw_list, &zombies);
-                }
-                cfs_spin_unlock_bh(&lcw_pending_timers_lock);
-
-                while (!cfs_list_empty(&zombies)) {
-                        lcw = cfs_list_entry(lcw_pending_timers.next,
-                                         struct lc_watchdog, lcw_list);
-                        cfs_list_del(&lcw->lcw_list);
-                        LIBCFS_FREE(lcw, sizeof(*lcw));
-                }
-        }
+                       spin_lock_bh(&lcw_pending_timers_lock);
+                       lcw->lcw_refcount--; /* -1 ref for callback */
+                       if (lcw->lcw_refcount == 0)
+                               list_add(&lcw->lcw_list, &zombies);
+               }
+               spin_unlock_bh(&lcw_pending_timers_lock);
 
-        cfs_complete(&lcw_stop_completion);
+               while (!list_empty(&zombies)) {
+                       lcw = list_entry(zombies.next,
+                                            struct lc_watchdog, lcw_list);
+                       list_del_init(&lcw->lcw_list);
+                       LIBCFS_FREE(lcw, sizeof(*lcw));
+               }
+       }
 
-        RETURN(rc);
+       complete(&lcw_stop_completion);
+
+       RETURN(rc);
 }
 
 static void lcw_dispatch_start(void)
 {
-        int rc;
-
-        ENTRY;
-        LASSERT(lcw_refcount == 1);
-
-        cfs_init_completion(&lcw_stop_completion);
-        cfs_init_completion(&lcw_start_completion);
-        cfs_waitq_init(&lcw_event_waitq);
-
-        CDEBUG(D_INFO, "starting dispatch thread\n");
-        rc = cfs_create_thread(lcw_dispatch_main, NULL, 0);
-        if (rc < 0) {
-                CERROR("error spawning watchdog dispatch thread: %d\n", rc);
-                EXIT;
-                return;
-        }
-        cfs_wait_for_completion(&lcw_start_completion);
-        CDEBUG(D_INFO, "watchdog dispatcher initialization complete.\n");
-
-        EXIT;
+       struct task_struct *task;
+
+       ENTRY;
+       LASSERT(lcw_refcount == 1);
+
+       init_completion(&lcw_stop_completion);
+       init_completion(&lcw_start_completion);
+       init_waitqueue_head(&lcw_event_waitq);
+
+       CDEBUG(D_INFO, "starting dispatch thread\n");
+       task = kthread_run(lcw_dispatch_main, NULL, "lc_watchdogd");
+       if (IS_ERR(task)) {
+               CERROR("error spawning watchdog dispatch thread: %ld\n",
+                       PTR_ERR(task));
+               EXIT;
+               return;
+       }
+       wait_for_completion(&lcw_start_completion);
+       CDEBUG(D_INFO, "watchdog dispatcher initialization complete.\n");
+
+       EXIT;
 }
 
 static void lcw_dispatch_stop(void)
 {
-        ENTRY;
-        LASSERT(lcw_refcount == 0);
+       ENTRY;
+       LASSERT(lcw_refcount == 0);
 
-        CDEBUG(D_INFO, "trying to stop watchdog dispatcher.\n");
+       CDEBUG(D_INFO, "trying to stop watchdog dispatcher.\n");
 
-        cfs_set_bit(LCW_FLAG_STOP, &lcw_flags);
-        cfs_waitq_signal(&lcw_event_waitq);
+       set_bit(LCW_FLAG_STOP, &lcw_flags);
+       wake_up(&lcw_event_waitq);
 
-        cfs_wait_for_completion(&lcw_stop_completion);
+       wait_for_completion(&lcw_stop_completion);
+       clear_bit(LCW_FLAG_STOP, &lcw_flags);
 
-        CDEBUG(D_INFO, "watchdog dispatcher has shut down.\n");
+       CDEBUG(D_INFO, "watchdog dispatcher has shut down.\n");
 
-        EXIT;
+       EXIT;
 }
 
 struct lc_watchdog *lc_watchdog_add(int timeout,
@@ -363,28 +355,28 @@ struct lc_watchdog *lc_watchdog_add(int timeout,
                 RETURN(ERR_PTR(-ENOMEM));
         }
 
-        cfs_spin_lock_init(&lcw->lcw_lock);
-        lcw->lcw_refcount = 1; /* refcount for owner */
-        lcw->lcw_task     = cfs_current();
-        lcw->lcw_pid      = cfs_curproc_pid();
-        lcw->lcw_callback = (callback != NULL) ? callback : lc_watchdog_dumplog;
-        lcw->lcw_data     = data;
-        lcw->lcw_state    = LC_WATCHDOG_DISABLED;
-
-        CFS_INIT_LIST_HEAD(&lcw->lcw_list);
-        cfs_timer_init(&lcw->lcw_timer, lcw_cb, lcw);
-
-        cfs_mutex_lock(&lcw_refcount_mutex);
-        if (++lcw_refcount == 1)
-                lcw_dispatch_start();
-        cfs_mutex_unlock(&lcw_refcount_mutex);
-
-        /* Keep this working in case we enable them by default */
-        if (lcw->lcw_state == LC_WATCHDOG_ENABLED) {
-                lcw->lcw_last_touched = cfs_time_current();
-                cfs_timer_arm(&lcw->lcw_timer, cfs_time_seconds(timeout) +
-                              cfs_time_current());
-        }
+       spin_lock_init(&lcw->lcw_lock);
+       lcw->lcw_refcount = 1; /* refcount for owner */
+       lcw->lcw_task     = current;
+       lcw->lcw_pid      = current_pid();
+       lcw->lcw_callback = (callback != NULL) ? callback : lc_watchdog_dumplog;
+       lcw->lcw_data     = data;
+       lcw->lcw_state    = LC_WATCHDOG_DISABLED;
+
+       INIT_LIST_HEAD(&lcw->lcw_list);
+       cfs_timer_setup(&lcw->lcw_timer, lcw_cb, (unsigned long)lcw, 0);
+
+       mutex_lock(&lcw_refcount_mutex);
+       if (++lcw_refcount == 1)
+               lcw_dispatch_start();
+       mutex_unlock(&lcw_refcount_mutex);
+
+       /* Keep this working in case we enable them by default */
+       if (lcw->lcw_state == LC_WATCHDOG_ENABLED) {
+               lcw->lcw_last_touched = ktime_get();
+               mod_timer(&lcw->lcw_timer, cfs_time_seconds(timeout) +
+                         jiffies);
+       }
 
         RETURN(lcw);
 }
@@ -392,53 +384,47 @@ EXPORT_SYMBOL(lc_watchdog_add);
 
 static void lcw_update_time(struct lc_watchdog *lcw, const char *message)
 {
-        cfs_time_t newtime = cfs_time_current();;
-
-        if (lcw->lcw_state == LC_WATCHDOG_EXPIRED) {
-                struct timeval timediff;
-                cfs_time_t delta_time = cfs_time_sub(newtime,
-                                                     lcw->lcw_last_touched);
-                cfs_duration_usec(delta_time, &timediff);
-
-                LCONSOLE_WARN("Service thread pid %u %s after %lu.%.02lus. "
-                              "This indicates the system was overloaded (too "
-                              "many service threads, or there were not enough "
-                              "hardware resources).\n",
-                              lcw->lcw_pid,
-                              message,
-                              timediff.tv_sec,
-                              timediff.tv_usec / 10000);
-        }
-        lcw->lcw_last_touched = newtime;
+       ktime_t newtime = ktime_get();
+
+       if (lcw->lcw_state == LC_WATCHDOG_EXPIRED) {
+               ktime_t lapse = ktime_sub(newtime, lcw->lcw_last_touched);
+               struct timespec64 timediff;
+
+               timediff = ktime_to_timespec64(lapse);
+               LCONSOLE_WARN("Service thread pid %u %s after %lu.%.02lus. This indicates the system was overloaded (too many service threads, or there were not enough hardware resources).\n",
+                             lcw->lcw_pid, message,
+                             timediff.tv_sec,
+                             timediff.tv_nsec / (NSEC_PER_SEC / 100));
+       }
+       lcw->lcw_last_touched = newtime;
 }
 
 static void lc_watchdog_del_pending(struct lc_watchdog *lcw)
 {
-        cfs_spin_lock_bh(&lcw->lcw_lock);
-        if (unlikely(!cfs_list_empty(&lcw->lcw_list))) {
-                cfs_spin_lock_bh(&lcw_pending_timers_lock);
-                cfs_list_del_init(&lcw->lcw_list);
-                lcw->lcw_refcount--; /* -1 ref for pending list */
-                cfs_spin_unlock_bh(&lcw_pending_timers_lock);
-        }
-
-        cfs_spin_unlock_bh(&lcw->lcw_lock);
+       spin_lock_bh(&lcw->lcw_lock);
+       if (unlikely(!list_empty(&lcw->lcw_list))) {
+               spin_lock_bh(&lcw_pending_timers_lock);
+               list_del_init(&lcw->lcw_list);
+               lcw->lcw_refcount--; /* -1 ref for pending list */
+               spin_unlock_bh(&lcw_pending_timers_lock);
+       }
+
+       spin_unlock_bh(&lcw->lcw_lock);
 }
 
 void lc_watchdog_touch(struct lc_watchdog *lcw, int timeout)
 {
-        ENTRY;
-        LASSERT(lcw != NULL);
+       ENTRY;
+       LASSERT(lcw != NULL);
 
-        lc_watchdog_del_pending(lcw);
+       lc_watchdog_del_pending(lcw);
 
-        lcw_update_time(lcw, "resumed");
-        lcw->lcw_state = LC_WATCHDOG_ENABLED;
+       lcw_update_time(lcw, "resumed");
 
-        cfs_timer_arm(&lcw->lcw_timer, cfs_time_current() +
-                      cfs_time_seconds(timeout));
+       mod_timer(&lcw->lcw_timer, jiffies + cfs_time_seconds(timeout));
+       lcw->lcw_state = LC_WATCHDOG_ENABLED;
 
-        EXIT;
+       EXIT;
 }
 EXPORT_SYMBOL(lc_watchdog_touch);
 
@@ -463,44 +449,34 @@ void lc_watchdog_delete(struct lc_watchdog *lcw)
         ENTRY;
         LASSERT(lcw != NULL);
 
-        cfs_timer_disarm(&lcw->lcw_timer);
+       del_timer(&lcw->lcw_timer);
 
         lcw_update_time(lcw, "stopped");
 
-        cfs_spin_lock_bh(&lcw->lcw_lock);
-        cfs_spin_lock_bh(&lcw_pending_timers_lock);
-        if (unlikely(!cfs_list_empty(&lcw->lcw_list))) {
-                cfs_list_del_init(&lcw->lcw_list);
-                lcw->lcw_refcount--; /* -1 ref for pending list */
-        }
+       spin_lock_bh(&lcw->lcw_lock);
+       spin_lock_bh(&lcw_pending_timers_lock);
+       if (unlikely(!list_empty(&lcw->lcw_list))) {
+               list_del_init(&lcw->lcw_list);
+               lcw->lcw_refcount--; /* -1 ref for pending list */
+       }
 
-        lcw->lcw_refcount--; /* -1 ref for owner */
-        dead = lcw->lcw_refcount == 0;
-        cfs_spin_unlock_bh(&lcw_pending_timers_lock);
-        cfs_spin_unlock_bh(&lcw->lcw_lock);
+       lcw->lcw_refcount--; /* -1 ref for owner */
+       dead = lcw->lcw_refcount == 0;
+       spin_unlock_bh(&lcw_pending_timers_lock);
+       spin_unlock_bh(&lcw->lcw_lock);
 
-        if (dead)
-                LIBCFS_FREE(lcw, sizeof(*lcw));
+       if (dead)
+               LIBCFS_FREE(lcw, sizeof(*lcw));
 
-        cfs_mutex_lock(&lcw_refcount_mutex);
-        if (--lcw_refcount == 0)
-                lcw_dispatch_stop();
-        cfs_mutex_unlock(&lcw_refcount_mutex);
+       mutex_lock(&lcw_refcount_mutex);
+       if (--lcw_refcount == 0)
+               lcw_dispatch_stop();
+       mutex_unlock(&lcw_refcount_mutex);
 
-        EXIT;
+       EXIT;
 }
 EXPORT_SYMBOL(lc_watchdog_delete);
 
-/*
- * Provided watchdog handlers
- */
-
-void lc_watchdog_dumplog(pid_t pid, void *data)
-{
-        libcfs_debug_dumplog_internal((void *)((long_ptr_t)pid));
-}
-EXPORT_SYMBOL(lc_watchdog_dumplog);
-
 #else   /* !defined(WITH_WATCHDOG) */
 
 struct lc_watchdog *lc_watchdog_add(int timeout,