blkio: Add root group to td->tg_list
authorVivek Goyal <vgoyal@redhat.com>
Fri, 1 Oct 2010 12:49:48 +0000 (14:49 +0200)
committerJens Axboe <jaxboe@fusionio.com>
Fri, 1 Oct 2010 12:49:48 +0000 (14:49 +0200)
o Currently all the dynamically allocated groups, except root grp is added
  to td->tg_list. This was not a problem so far but in next patch I will
  travel through td->tg_list to process any updates of limits on the group.
  If root group is not in tg_list, then root group's updates are not
  processed.

o It is better to root group also to tg_list instead of doing special
  processing for it during limit updates.

Signed-off-by: Vivek Goyal <vgoyal@redhat.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
block/blk-throttle.c

index af53f37c1b13188bcc599b3c02b786a99d25e008..bc2936b80addb0d7bf89c8bb39f8e1d3bedc08c7 100644 (file)
@@ -87,7 +87,7 @@ struct throtl_data
        unsigned int nr_queued[2];
 
        /*
-        * number of total undestroyed groups (excluding root group)
+        * number of total undestroyed groups
         */
        unsigned int nr_undestroyed_grps;
 
@@ -940,7 +940,17 @@ int blk_throtl_init(struct request_queue *q)
        /* Practically unlimited BW */
        tg->bps[0] = tg->bps[1] = -1;
        tg->iops[0] = tg->iops[1] = -1;
-       atomic_set(&tg->ref, 1);
+
+       /*
+        * Set root group reference to 2. One reference will be dropped when
+        * all groups on tg_list are being deleted during queue exit. Other
+        * reference will remain there as we don't want to delete this group
+        * as it is statically allocated and gets destroyed when throtl_data
+        * goes away.
+        */
+       atomic_set(&tg->ref, 2);
+       hlist_add_head(&tg->tg_node, &td->tg_list);
+       td->nr_undestroyed_grps++;
 
        INIT_DELAYED_WORK(&td->throtl_work, blk_throtl_work);
 
@@ -966,10 +976,9 @@ void blk_throtl_exit(struct request_queue *q)
 
        spin_lock_irq(q->queue_lock);
        throtl_release_tgs(td);
-       blkiocg_del_blkio_group(&td->root_tg.blkg);
 
        /* If there are other groups */
-       if (td->nr_undestroyed_grps >= 1)
+       if (td->nr_undestroyed_grps > 0)
                wait = true;
 
        spin_unlock_irq(q->queue_lock);