drbd: improve throttling decisions of background resynchronisation

Background resynchronisation does some "side-stepping", or throttles
itself, if it detects application IO activity, and the current resync
rate estimate is above the configured "cmin-rate".

What was not detected: if there is no application IO,
because it blocks on activity log transactions.

Introduce a new atomic_t ap_actlog_cnt, tracking such blocked requests,
and count non-zero as application IO activity.
This counter is exposed at proc_details level 2 and above.

Also make sure to release the currently locked resync extent
if we side-step due to such voluntary throttling.

Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>

1 files changed, 26 insertions, 3 deletions

diff --git a/drivers/block/drbd/drbd_actlog.c b/drivers/block/drbd/drbd_actlog.c
index d7e80663187c..6ce5c76d642b 100644
--- a/drivers/block/drbd/drbd_actlog.c
+++ b/drivers/block/drbd/drbd_actlog.c
@@ -991,6 +991,15 @@ int drbd_try_rs_begin_io(struct drbd_device *device, sector_t sector)
 	struct lc_element *e;
 	struct bm_extent *bm_ext;
 	int i;
+	bool throttle = drbd_rs_should_slow_down(device, sector, true);
+
+	/* If we need to throttle, a half-locked (only marked BME_NO_WRITES,
+	 * not yet BME_LOCKED) extent needs to be kicked out explicitly if we
+	 * need to throttle. There is at most one such half-locked extent,
+	 * which is remembered in resync_wenr. */
+
+	if (throttle && device->resync_wenr != enr)
+		return -EAGAIN;
 
 	spin_lock_irq(&device->al_lock);
 	if (device->resync_wenr != LC_FREE && device->resync_wenr != enr) {
@@ -1014,8 +1023,10 @@ int drbd_try_rs_begin_io(struct drbd_device *device, sector_t sector)
 			D_ASSERT(device, test_bit(BME_NO_WRITES, &bm_ext->flags));
 			clear_bit(BME_NO_WRITES, &bm_ext->flags);
 			device->resync_wenr = LC_FREE;
-			if (lc_put(device->resync, &bm_ext->lce) == 0)
+			if (lc_put(device->resync, &bm_ext->lce) == 0) {
+				bm_ext->flags = 0;
 				device->resync_locked--;
+			}
 			wake_up(&device->al_wait);
 		} else {
 			drbd_alert(device, "LOGIC BUG\n");
@@ -1077,8 +1088,20 @@ proceed:
 	return 0;
 
 try_again:
-	if (bm_ext)
-		device->resync_wenr = enr;
+	if (bm_ext) {
+		if (throttle) {
+			D_ASSERT(device, !test_bit(BME_LOCKED, &bm_ext->flags));
+			D_ASSERT(device, test_bit(BME_NO_WRITES, &bm_ext->flags));
+			clear_bit(BME_NO_WRITES, &bm_ext->flags);
+			device->resync_wenr = LC_FREE;
+			if (lc_put(device->resync, &bm_ext->lce) == 0) {
+				bm_ext->flags = 0;
+				device->resync_locked--;
+			}
+			wake_up(&device->al_wait);
+		} else
+			device->resync_wenr = enr;
+	}
 	spin_unlock_irq(&device->al_lock);
 	return -EAGAIN;
 }