summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorShaohua Li <shli@kernel.org>2012-08-02 08:33:00 +1000
committerNeilBrown <neilb@suse.de>2012-08-02 08:33:00 +1000
commit8811b5968f6216e97ccb9fe7b9883af39e339921 (patch)
tree9693073f6cb1bcd8743d429d7d3330018bc0ce6d
parent74018dc3063a2c729fc73041c0a9f03aac995920 (diff)
raid5: make_request use batch stripe release
make_request() does stripe release for every stripe and the stripe usually has count 1, which makes previous release_stripe() optimization not work. In my test, this release_stripe() becomes the heaviest pleace to take conf->device_lock after previous patches applied. Below patch makes stripe release batch. All the stripes will be released in unplug. The STRIPE_ON_UNPLUG_LIST bit is to protect concurrent access stripe lru. Signed-off-by: Shaohua Li <shli@fusionio.com> Signed-off-by: NeilBrown <neilb@suse.de>
-rw-r--r--drivers/md/raid5.c62
-rw-r--r--drivers/md/raid5.h1
2 files changed, 60 insertions, 3 deletions
diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c
index bde9da2baa39..978ba9b7a3c4 100644
--- a/drivers/md/raid5.c
+++ b/drivers/md/raid5.c
@@ -471,7 +471,8 @@ get_active_stripe(struct r5conf *conf, sector_t sector,
} else {
if (atomic_read(&sh->count)) {
BUG_ON(!list_empty(&sh->lru)
- && !test_bit(STRIPE_EXPANDING, &sh->state));
+ && !test_bit(STRIPE_EXPANDING, &sh->state)
+ && !test_bit(STRIPE_ON_UNPLUG_LIST, &sh->state));
} else {
if (!test_bit(STRIPE_HANDLE, &sh->state))
atomic_inc(&conf->active_stripes);
@@ -3988,6 +3989,62 @@ static struct stripe_head *__get_priority_stripe(struct r5conf *conf)
return sh;
}
+struct raid5_plug_cb {
+ struct blk_plug_cb cb;
+ struct list_head list;
+};
+
+static void raid5_unplug(struct blk_plug_cb *blk_cb, bool from_schedule)
+{
+ struct raid5_plug_cb *cb = container_of(
+ blk_cb, struct raid5_plug_cb, cb);
+ struct stripe_head *sh;
+ struct mddev *mddev = cb->cb.data;
+ struct r5conf *conf = mddev->private;
+
+ if (cb->list.next && !list_empty(&cb->list)) {
+ spin_lock_irq(&conf->device_lock);
+ while (!list_empty(&cb->list)) {
+ sh = list_first_entry(&cb->list, struct stripe_head, lru);
+ list_del_init(&sh->lru);
+ /*
+ * avoid race release_stripe_plug() sees
+ * STRIPE_ON_UNPLUG_LIST clear but the stripe
+ * is still in our list
+ */
+ smp_mb__before_clear_bit();
+ clear_bit(STRIPE_ON_UNPLUG_LIST, &sh->state);
+ __release_stripe(conf, sh);
+ }
+ spin_unlock_irq(&conf->device_lock);
+ }
+ kfree(cb);
+}
+
+static void release_stripe_plug(struct mddev *mddev,
+ struct stripe_head *sh)
+{
+ struct blk_plug_cb *blk_cb = blk_check_plugged(
+ raid5_unplug, mddev,
+ sizeof(struct raid5_plug_cb));
+ struct raid5_plug_cb *cb;
+
+ if (!blk_cb) {
+ release_stripe(sh);
+ return;
+ }
+
+ cb = container_of(blk_cb, struct raid5_plug_cb, cb);
+
+ if (cb->list.next == NULL)
+ INIT_LIST_HEAD(&cb->list);
+
+ if (!test_and_set_bit(STRIPE_ON_UNPLUG_LIST, &sh->state))
+ list_add_tail(&sh->lru, &cb->list);
+ else
+ release_stripe(sh);
+}
+
static void make_request(struct mddev *mddev, struct bio * bi)
{
struct r5conf *conf = mddev->private;
@@ -4116,8 +4173,7 @@ static void make_request(struct mddev *mddev, struct bio * bi)
if ((bi->bi_rw & REQ_SYNC) &&
!test_and_set_bit(STRIPE_PREREAD_ACTIVE, &sh->state))
atomic_inc(&conf->preread_active_stripes);
- mddev_check_plugged(mddev);
- release_stripe(sh);
+ release_stripe_plug(mddev, sh);
} else {
/* cannot get stripe for read-ahead, just give-up */
clear_bit(BIO_UPTODATE, &bi->bi_flags);
diff --git a/drivers/md/raid5.h b/drivers/md/raid5.h
index 2164021f3b5f..9a7b36f0a425 100644
--- a/drivers/md/raid5.h
+++ b/drivers/md/raid5.h
@@ -319,6 +319,7 @@ enum {
STRIPE_BIOFILL_RUN,
STRIPE_COMPUTE_RUN,
STRIPE_OPS_REQ_PENDING,
+ STRIPE_ON_UNPLUG_LIST,
};
/*