Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/sage/ceph...
[pandora-kernel.git] / drivers / md / raid5.c
index e867ee4..49bf5f8 100644 (file)
  *
  * We group bitmap updates into batches.  Each batch has a number.
  * We may write out several batches at once, but that isn't very important.
- * conf->bm_write is the number of the last batch successfully written.
- * conf->bm_flush is the number of the last batch that was closed to
+ * conf->seq_write is the number of the last batch successfully written.
+ * conf->seq_flush is the number of the last batch that was closed to
  *    new additions.
  * When we discover that we will need to write to any block in a stripe
  * (in add_stripe_bio) we update the in-memory bitmap and record in sh->bm_seq
- * the number of the batch it will be in. This is bm_flush+1.
+ * the number of the batch it will be in. This is seq_flush+1.
  * When we are ready to do a write, if that batch hasn't been written yet,
  *   we plug the array and queue the stripe for later.
  * When an unplug happens, we increment bm_flush, thus closing the current
@@ -199,14 +199,12 @@ static void __release_stripe(raid5_conf_t *conf, struct stripe_head *sh)
                BUG_ON(!list_empty(&sh->lru));
                BUG_ON(atomic_read(&conf->active_stripes)==0);
                if (test_bit(STRIPE_HANDLE, &sh->state)) {
-                       if (test_bit(STRIPE_DELAYED, &sh->state)) {
+                       if (test_bit(STRIPE_DELAYED, &sh->state))
                                list_add_tail(&sh->lru, &conf->delayed_list);
-                               plugger_set_plug(&conf->plug);
-                       } else if (test_bit(STRIPE_BIT_DELAY, &sh->state) &&
-                                  sh->bm_seq - conf->seq_write > 0) {
+                       else if (test_bit(STRIPE_BIT_DELAY, &sh->state) &&
+                                  sh->bm_seq - conf->seq_write > 0)
                                list_add_tail(&sh->lru, &conf->bitmap_list);
-                               plugger_set_plug(&conf->plug);
-                       } else {
+                       else {
                                clear_bit(STRIPE_BIT_DELAY, &sh->state);
                                list_add_tail(&sh->lru, &conf->handle_list);
                        }
@@ -461,7 +459,7 @@ get_active_stripe(raid5_conf_t *conf, sector_t sector,
                                                     < (conf->max_nr_stripes *3/4)
                                                     || !conf->inactive_blocked),
                                                    conf->device_lock,
-                                                   md_raid5_kick_device(conf));
+                                                   );
                                conf->inactive_blocked = 0;
                        } else
                                init_stripe(sh, sector, previous);
@@ -1470,7 +1468,7 @@ static int resize_stripes(raid5_conf_t *conf, int newsize)
                wait_event_lock_irq(conf->wait_for_stripe,
                                    !list_empty(&conf->inactive_list),
                                    conf->device_lock,
-                                   blk_flush_plug(current));
+                                   );
                osh = get_free_stripe(conf);
                spin_unlock_irq(&conf->device_lock);
                atomic_set(&nsh->count, 1);
@@ -3623,8 +3621,7 @@ static void raid5_activate_delayed(raid5_conf_t *conf)
                                atomic_inc(&conf->preread_active_stripes);
                        list_add_tail(&sh->lru, &conf->hold_list);
                }
-       } else
-               plugger_set_plug(&conf->plug);
+       }
 }
 
 static void activate_bit_delay(raid5_conf_t *conf)
@@ -3641,21 +3638,6 @@ static void activate_bit_delay(raid5_conf_t *conf)
        }
 }
 
-void md_raid5_kick_device(raid5_conf_t *conf)
-{
-       blk_flush_plug(current);
-       raid5_activate_delayed(conf);
-       md_wakeup_thread(conf->mddev->thread);
-}
-EXPORT_SYMBOL_GPL(md_raid5_kick_device);
-
-static void raid5_unplug(struct plug_handle *plug)
-{
-       raid5_conf_t *conf = container_of(plug, raid5_conf_t, plug);
-
-       md_raid5_kick_device(conf);
-}
-
 int md_raid5_congested(mddev_t *mddev, int bits)
 {
        raid5_conf_t *conf = mddev->private;
@@ -3945,6 +3927,7 @@ static int make_request(mddev_t *mddev, struct bio * bi)
        struct stripe_head *sh;
        const int rw = bio_data_dir(bi);
        int remaining;
+       int plugged;
 
        if (unlikely(bi->bi_rw & REQ_FLUSH)) {
                md_flush_request(mddev, bi);
@@ -3963,6 +3946,7 @@ static int make_request(mddev_t *mddev, struct bio * bi)
        bi->bi_next = NULL;
        bi->bi_phys_segments = 1;       /* over-loaded to count active stripes */
 
+       plugged = mddev_check_plugged(mddev);
        for (;logical_sector < last_sector; logical_sector += STRIPE_SECTORS) {
                DEFINE_WAIT(w);
                int disks, data_disks;
@@ -4057,7 +4041,7 @@ static int make_request(mddev_t *mddev, struct bio * bi)
                                 * add failed due to overlap.  Flush everything
                                 * and wait a while
                                 */
-                               md_raid5_kick_device(conf);
+                               md_wakeup_thread(mddev->thread);
                                release_stripe(sh);
                                schedule();
                                goto retry;
@@ -4077,6 +4061,9 @@ static int make_request(mddev_t *mddev, struct bio * bi)
                }
                        
        }
+       if (!plugged)
+               md_wakeup_thread(mddev->thread);
+
        spin_lock_irq(&conf->device_lock);
        remaining = raid5_dec_bi_phys_segments(bi);
        spin_unlock_irq(&conf->device_lock);
@@ -4478,24 +4465,30 @@ static void raid5d(mddev_t *mddev)
        struct stripe_head *sh;
        raid5_conf_t *conf = mddev->private;
        int handled;
+       struct blk_plug plug;
 
        pr_debug("+++ raid5d active\n");
 
        md_check_recovery(mddev);
 
+       blk_start_plug(&plug);
        handled = 0;
        spin_lock_irq(&conf->device_lock);
        while (1) {
                struct bio *bio;
 
-               if (conf->seq_flush != conf->seq_write) {
-                       int seq = conf->seq_flush;
+               if (atomic_read(&mddev->plug_cnt) == 0 &&
+                   !list_empty(&conf->bitmap_list)) {
+                       /* Now is a good time to flush some bitmap updates */
+                       conf->seq_flush++;
                        spin_unlock_irq(&conf->device_lock);
                        bitmap_unplug(mddev->bitmap);
                        spin_lock_irq(&conf->device_lock);
-                       conf->seq_write = seq;
+                       conf->seq_write = conf->seq_flush;
                        activate_bit_delay(conf);
                }
+               if (atomic_read(&mddev->plug_cnt) == 0)
+                       raid5_activate_delayed(conf);
 
                while ((bio = remove_bio_from_retry(conf))) {
                        int ok;
@@ -4525,6 +4518,7 @@ static void raid5d(mddev_t *mddev)
        spin_unlock_irq(&conf->device_lock);
 
        async_tx_issue_pending_all();
+       blk_finish_plug(&plug);
 
        pr_debug("--- raid5d inactive\n");
 }
@@ -5141,8 +5135,6 @@ static int run(mddev_t *mddev)
                       mdname(mddev));
        md_set_array_sectors(mddev, raid5_size(mddev, 0, 0));
 
-       plugger_init(&conf->plug, raid5_unplug);
-       mddev->plug = &conf->plug;
        if (mddev->queue) {
                int chunk_size;
                /* read-ahead size must cover two whole stripes, which
@@ -5159,7 +5151,6 @@ static int run(mddev_t *mddev)
 
                mddev->queue->backing_dev_info.congested_data = mddev;
                mddev->queue->backing_dev_info.congested_fn = raid5_congested;
-               mddev->queue->queue_lock = &conf->device_lock;
 
                chunk_size = mddev->chunk_sectors << 9;
                blk_queue_io_min(mddev->queue, chunk_size);
@@ -5192,7 +5183,6 @@ static int stop(mddev_t *mddev)
        mddev->thread = NULL;
        if (mddev->queue)
                mddev->queue->backing_dev_info.congested_fn = NULL;
-       plugger_flush(&conf->plug); /* the unplug fn references 'conf'*/
        free_conf(conf);
        mddev->private = NULL;
        mddev->to_remove = &raid5_attrs_group;
@@ -5688,6 +5678,7 @@ static void raid5_quiesce(mddev_t *mddev, int state)
 static void *raid45_takeover_raid0(mddev_t *mddev, int level)
 {
        struct raid0_private_data *raid0_priv = mddev->private;
+       sector_t sectors;
 
        /* for raid0 takeover only one zone is supported */
        if (raid0_priv->nr_strip_zones > 1) {
@@ -5696,6 +5687,9 @@ static void *raid45_takeover_raid0(mddev_t *mddev, int level)
                return ERR_PTR(-EINVAL);
        }
 
+       sectors = raid0_priv->strip_zone[0].zone_end;
+       sector_div(sectors, raid0_priv->strip_zone[0].nb_dev);
+       mddev->dev_sectors = sectors;
        mddev->new_level = level;
        mddev->new_layout = ALGORITHM_PARITY_N;
        mddev->new_chunk_sectors = mddev->chunk_sectors;