writeback: move bdi threads exiting logic to the forker thread

Currently, bdi threads can decide to exit if there were no useful activities
for 5 minutes. However, this causes nasty races: we can easily oops in the
'bdi_queue_work()' if the bdi thread decides to exit while we are waking it up.

And even if we do not oops, but the bdi tread exits immediately after we wake
it up, we'd lose the wake-up event and have an unnecessary delay (up to 5 secs)
in the bdi work processing.

This patch makes the forker thread to be the central place which not only
creates bdi threads, but also kills them if they were inactive long enough.
This better design-wise.

Another reason why this change was done is to prepare for the further changes
which will prevent the bdi threads from waking up every 5 sec and wasting
power. Indeed, when the task does not wake up periodically anymore, it won't be
able to exit either.

This patch also moves the the 'wake_up_bit()' call from the bdi thread to the
forker thread as well. So now the forker thread sets the BDI_pending bit, then
forks the task or kills it, then clears the bit and wakes up the waiting
process.

The only process which may wain on the bit is 'bdi_wb_shutdown()'. This
function was changed as well - now it first removes the bdi from the
'bdi_list', then waits on the 'BDI_pending' bit. Once it wakes up, it is
guaranteed that the forker thread won't race with it, because the bdi is not
visible. Note, the forker thread sets the 'BDI_pending' bit under the
'bdi->wb_lock' which is essential for proper serialization.

And additionally, when we change 'bdi->wb.task', we now take the
'bdi->work_lock', to make sure that we do not lose wake-ups which we otherwise
would when raced with, say, 'bdi_queue_work()'.

Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>

1 files changed, 58 insertions, 11 deletions

diff --git a/mm/backing-dev.c b/mm/backing-dev.c
index e104e32c2ee8..9c1c199f88ce 100644
--- a/mm/backing-dev.c
+++ b/mm/backing-dev.c
@@ -316,6 +316,18 @@ static void sync_supers_timer_fn(unsigned long unused)
         bdi_arm_supers_timer();
 }
 
+/*
+ * Calculate the longest interval (jiffies) bdi threads are allowed to be
+ * inactive.
+ */
+static unsigned long bdi_longest_inactive(void)
+{
+        unsigned long interval;
+
+        interval = msecs_to_jiffies(dirty_writeback_interval * 10);
+        return max(5UL * 60 * HZ, interval);
+}
+
 static int bdi_forker_thread(void *ptr)
 {
         struct bdi_writeback *me = ptr;
@@ -329,11 +341,12 @@ static int bdi_forker_thread(void *ptr)
         set_user_nice(current, 0);
 
         for (;;) {
-                struct task_struct *task;
+                struct task_struct *task = NULL;
                 struct backing_dev_info *bdi;
                 enum {
                         NO_ACTION,   /* Nothing to do */
                         FORK_THREAD, /* Fork bdi thread */
+                        KILL_THREAD, /* Kill inactive bdi thread */
                 } action = NO_ACTION;
 
                 /*
@@ -346,10 +359,6 @@ static int bdi_forker_thread(void *ptr)
                 spin_lock_bh(&bdi_lock);
                 set_current_state(TASK_INTERRUPTIBLE);
 
-                /*
-                 * Check if any existing bdi's have dirty data without
-                 * a thread registered. If so, set that up.
-                 */
                 list_for_each_entry(bdi, &bdi_list, bdi_list) {
                         bool have_dirty_io;
 
@@ -376,6 +385,25 @@ static int bdi_forker_thread(void *ptr)
                                 action = FORK_THREAD;
                                 break;
                         }
+
+                        spin_lock(&bdi->wb_lock);
+                        /*
+                         * If there is no work to do and the bdi thread was
+                         * inactive long enough - kill it. The wb_lock is taken
+                         * to make sure no-one adds more work to this bdi and
+                         * wakes the bdi thread up.
+                         */
+                        if (bdi->wb.task && !have_dirty_io &&
+                            time_after(jiffies, bdi->wb.last_active +
+                                                bdi_longest_inactive())) {
+                                task = bdi->wb.task;
+                                bdi->wb.task = NULL;
+                                spin_unlock(&bdi->wb_lock);
+                                set_bit(BDI_pending, &bdi->state);
+                                action = KILL_THREAD;
+                                break;
+                        }
+                        spin_unlock(&bdi->wb_lock);
                 }
                 spin_unlock_bh(&bdi_lock);
 
@@ -394,8 +422,20 @@ static int bdi_forker_thread(void *ptr)
                                  * the bdi from the thread.
                                  */
                                 bdi_flush_io(bdi);
-                        } else
+                        } else {
+                                /*
+                                 * The spinlock makes sure we do not lose
+                                 * wake-ups when racing with 'bdi_queue_work()'.
+                                 */
+                                spin_lock(&bdi->wb_lock);
                                 bdi->wb.task = task;
+                                spin_unlock(&bdi->wb_lock);
+                        }
+                        break;
+
+                case KILL_THREAD:
+                        __set_current_state(TASK_RUNNING);
+                        kthread_stop(task);
                         break;
 
                 case NO_ACTION:
@@ -407,6 +447,13 @@ static int bdi_forker_thread(void *ptr)
                         /* Back to the main loop */
                         continue;
                 }
+
+                /*
+                 * Clear pending bit and wakeup anybody waiting to tear us down.
+                 */
+                clear_bit(BDI_pending, &bdi->state);
+                smp_mb__after_clear_bit();
+                wake_up_bit(&bdi->state, BDI_pending);
         }
 
         return 0;
@@ -490,15 +537,15 @@ static void bdi_wb_shutdown(struct backing_dev_info *bdi)
                 return;
 
         /*
-         * If setup is pending, wait for that to complete first
+         * Make sure nobody finds us on the bdi_list anymore
          */
-        wait_on_bit(&bdi->state, BDI_pending, bdi_sched_wait,
-                        TASK_UNINTERRUPTIBLE);
+        bdi_remove_from_list(bdi);
 
         /*
-         * Make sure nobody finds us on the bdi_list anymore
+         * If setup is pending, wait for that to complete first
          */
-        bdi_remove_from_list(bdi);
+        wait_on_bit(&bdi->state, BDI_pending, bdi_sched_wait,
+                        TASK_UNINTERRUPTIBLE);
 
         /*
          * Finally, kill the kernel thread. We don't need to be RCU