drbd: introduce stop-sector to online verify

We now can schedule only a specific range of sectors for online verify,
or interrupt a running verify without interrupting the connection.

Had to bump the protocol version differently, we are now 101.
Added verify_can_do_stop_sector() { protocol >= 97 && protocol != 100; }

Also, the return value convention for worker callbacks has changed,
we returned "true/false" for "keep the connection up" in 8.3,
we return 0 for success and <= for failure in 8.4.
Affected: receive_state()

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

6 files changed, 74 insertions, 19 deletions

diff --git a/drivers/block/drbd/drbd_int.h b/drivers/block/drbd/drbd_int.h
index 22adfc7189de..eddc4388a1b1 100644
--- a/drivers/block/drbd/drbd_int.h
+++ b/drivers/block/drbd/drbd_int.h
@@ -971,6 +971,7 @@ struct drbd_conf {
 
 	/* where does the admin want us to start? (sector) */
 	sector_t ov_start_sector;
+	sector_t ov_stop_sector;
 	/* where are we now? (sector) */
 	sector_t ov_position;
 	/* Start sector of out of sync range (to merge printk reporting). */
@@ -2264,6 +2265,12 @@ static inline void dec_ap_bio(struct drbd_conf *mdev)
 		wake_up(&mdev->misc_wait);
 }
 
+static inline bool verify_can_do_stop_sector(struct drbd_conf *mdev)
+{
+	return mdev->tconn->agreed_pro_version >= 97 &&
+		mdev->tconn->agreed_pro_version != 100;
+}
+
 static inline int drbd_set_ed_uuid(struct drbd_conf *mdev, u64 val)
 {
 	int changed = mdev->ed_uuid != val;
diff --git a/drivers/block/drbd/drbd_nl.c b/drivers/block/drbd/drbd_nl.c
index 4afd626ca3dc..eefb56308aea 100644
--- a/drivers/block/drbd/drbd_nl.c
+++ b/drivers/block/drbd/drbd_nl.c
@@ -2939,6 +2939,7 @@ int drbd_adm_start_ov(struct sk_buff *skb, struct genl_info *info)
 {
 	struct drbd_conf *mdev;
 	enum drbd_ret_code retcode;
+	struct start_ov_parms parms;
 
 	retcode = drbd_adm_prepare(skb, info, DRBD_ADM_NEED_MINOR);
 	if (!adm_ctx.reply_skb)
@@ -2947,19 +2948,22 @@ int drbd_adm_start_ov(struct sk_buff *skb, struct genl_info *info)
 		goto out;
 
 	mdev = adm_ctx.mdev;
+
+	/* resume from last known position, if possible */
+	parms.ov_start_sector = mdev->ov_start_sector;
+	parms.ov_stop_sector = ULLONG_MAX;
 	if (info->attrs[DRBD_NLA_START_OV_PARMS]) {
-		/* resume from last known position, if possible */
-		struct start_ov_parms parms =
-			{ .ov_start_sector = mdev->ov_start_sector };
 		int err = start_ov_parms_from_attrs(&parms, info);
 		if (err) {
 			retcode = ERR_MANDATORY_TAG;
 			drbd_msg_put_info(from_attrs_err_to_txt(err));
 			goto out;
 		}
-		/* w_make_ov_request expects position to be aligned */
-		mdev->ov_start_sector = parms.ov_start_sector & ~BM_SECT_PER_BIT;
 	}
+	/* w_make_ov_request expects position to be aligned */
+	mdev->ov_start_sector = parms.ov_start_sector & ~(BM_SECT_PER_BIT-1);
+	mdev->ov_stop_sector = parms.ov_stop_sector;
+
 	/* If there is still bitmap IO pending, e.g. previous resync or verify
 	 * just being finished, wait for it before requesting a new resync. */
 	drbd_suspend_io(mdev);
diff --git a/drivers/block/drbd/drbd_proc.c b/drivers/block/drbd/drbd_proc.c
index e0f0d2a6d538..56672a61eb94 100644
--- a/drivers/block/drbd/drbd_proc.c
+++ b/drivers/block/drbd/drbd_proc.c
@@ -167,18 +167,24 @@ static void drbd_syncer_progress(struct drbd_conf *mdev, struct seq_file *seq)
 		 * we convert to sectors in the display below. */
 		unsigned long bm_bits = drbd_bm_bits(mdev);
 		unsigned long bit_pos;
+		unsigned long long stop_sector = 0;
 		if (mdev->state.conn == C_VERIFY_S ||
-		    mdev->state.conn == C_VERIFY_T)
+		    mdev->state.conn == C_VERIFY_T) {
 			bit_pos = bm_bits - mdev->ov_left;
-		else
+			if (verify_can_do_stop_sector(mdev))
+				stop_sector = mdev->ov_stop_sector;
+		} else
 			bit_pos = mdev->bm_resync_fo;
 		/* Total sectors may be slightly off for oddly
 		 * sized devices. So what. */
 		seq_printf(seq,
-			"\t%3d%% sector pos: %llu/%llu\n",
+			"\t%3d%% sector pos: %llu/%llu",
 			(int)(bit_pos / (bm_bits/100+1)),
 			(unsigned long long)bit_pos * BM_SECT_PER_BIT,
 			(unsigned long long)bm_bits * BM_SECT_PER_BIT);
+		if (stop_sector != 0 && stop_sector != ULLONG_MAX)
+			seq_printf(seq, " stop sector: %llu", stop_sector);
+		seq_printf(seq, "\n");
 	}
 }
 
diff --git a/drivers/block/drbd/drbd_receiver.c b/drivers/block/drbd/drbd_receiver.c
index 7fe6b01618d4..8fddec96dfbe 100644
--- a/drivers/block/drbd/drbd_receiver.c
+++ b/drivers/block/drbd/drbd_receiver.c
@@ -3843,7 +3843,7 @@ static int receive_state(struct drbd_tconn *tconn, struct packet_info *pi)
 	 * already decided to close the connection again,
 	 * we must not "re-establish" it here. */
 	if (os.conn <= C_TEAR_DOWN)
-		return false;
+		return -ECONNRESET;
 
 	/* If this is the "end of sync" confirmation, usually the peer disk
 	 * transitions from D_INCONSISTENT to D_UP_TO_DATE. For empty (0 bits
@@ -3875,6 +3875,14 @@ static int receive_state(struct drbd_tconn *tconn, struct packet_info *pi)
 		}
 	}
 
+	/* explicit verify finished notification, stop sector reached. */
+	if (os.conn == C_VERIFY_T && os.disk == D_UP_TO_DATE &&
+	    peer_state.conn == C_CONNECTED && real_peer_disk == D_UP_TO_DATE) {
+		ov_out_of_sync_print(mdev);
+		drbd_resync_finished(mdev);
+		return 0;
+	}
+
 	/* peer says his disk is inconsistent, while we think it is uptodate,
 	 * and this happens while the peer still thinks we have a sync going on,
 	 * but we think we are already done with the sync.
diff --git a/drivers/block/drbd/drbd_state.c b/drivers/block/drbd/drbd_state.c
index 444581828d70..12f2b4fbe559 100644
--- a/drivers/block/drbd/drbd_state.c
+++ b/drivers/block/drbd/drbd_state.c
@@ -975,13 +975,15 @@ __drbd_set_state(struct drbd_conf *mdev, union drbd_state ns,
 	wake_up(&mdev->state_wait);
 	wake_up(&mdev->tconn->ping_wait);
 
-	/* aborted verify run. log the last position */
+	/* Aborted verify run, or we reached the stop sector.
+	 * Log the last position, unless end-of-device. */
 	if ((os.conn == C_VERIFY_S || os.conn == C_VERIFY_T) &&
-	    ns.conn < C_CONNECTED) {
+	    ns.conn <= C_CONNECTED) {
 		mdev->ov_start_sector =
 			BM_BIT_TO_SECT(drbd_bm_bits(mdev) - mdev->ov_left);
-		dev_info(DEV, "Online Verify reached sector %llu\n",
-			(unsigned long long)mdev->ov_start_sector);
+		if (mdev->ov_left)
+			dev_info(DEV, "Online Verify reached sector %llu\n",
+				(unsigned long long)mdev->ov_start_sector);
 	}
 
 	if ((os.conn == C_PAUSED_SYNC_T || os.conn == C_PAUSED_SYNC_S) &&
@@ -1422,6 +1424,13 @@ static void after_state_ch(struct drbd_conf *mdev, union drbd_state os,
 	if (os.disk < D_UP_TO_DATE && os.conn >= C_SYNC_SOURCE && ns.conn == C_CONNECTED)
 		drbd_send_state(mdev, ns);
 
+	/* Verify finished, or reached stop sector.  Peer did not know about
+	 * the stop sector, and we may even have changed the stop sector during
+	 * verify to interrupt/stop early.  Send the new state. */
+	if (os.conn == C_VERIFY_S && ns.conn == C_CONNECTED
+	&& verify_can_do_stop_sector(mdev))
+		drbd_send_state(mdev, ns);
+
 	/* Wake up role changes, that were delayed because of connection establishing */
 	if (os.conn == C_WF_REPORT_PARAMS && ns.conn != C_WF_REPORT_PARAMS) {
 		if (test_and_clear_bit(STATE_SENT, &mdev->tconn->flags))
diff --git a/drivers/block/drbd/drbd_worker.c b/drivers/block/drbd/drbd_worker.c
index 9d7e1fb0f431..1c9c6fd332c3 100644
--- a/drivers/block/drbd/drbd_worker.c
+++ b/drivers/block/drbd/drbd_worker.c
@@ -692,6 +692,7 @@ static int w_make_ov_request(struct drbd_work *w, int cancel)
 	int number, i, size;
 	sector_t sector;
 	const sector_t capacity = drbd_get_capacity(mdev->this_bdev);
+	bool stop_sector_reached = false;
 
 	if (unlikely(cancel))
 		return 1;
@@ -700,9 +701,17 @@ static int w_make_ov_request(struct drbd_work *w, int cancel)
 
 	sector = mdev->ov_position;
 	for (i = 0; i < number; i++) {
-		if (sector >= capacity) {
+		if (sector >= capacity)
 			return 1;
-		}
+
+		/* We check for "finished" only in the reply path:
+		 * w_e_end_ov_reply().
+		 * We need to send at least one request out. */
+		stop_sector_reached = i > 0
+			&& verify_can_do_stop_sector(mdev)
+			&& sector >= mdev->ov_stop_sector;
+		if (stop_sector_reached)
+			break;
 
 		size = BM_BLOCK_SIZE;
 
@@ -726,7 +735,8 @@ static int w_make_ov_request(struct drbd_work *w, int cancel)
 
  requeue:
 	mdev->rs_in_flight += (i << (BM_BLOCK_SHIFT - 9));
-	mod_timer(&mdev->resync_timer, jiffies + SLEEP_TIME);
+	if (i == 0 || !stop_sector_reached)
+		mod_timer(&mdev->resync_timer, jiffies + SLEEP_TIME);
 	return 1;
 }
 
@@ -792,7 +802,12 @@ int drbd_resync_finished(struct drbd_conf *mdev)
 	dt = (jiffies - mdev->rs_start - mdev->rs_paused) / HZ;
 	if (dt <= 0)
 		dt = 1;
+	
 	db = mdev->rs_total;
+	/* adjust for verify start and stop sectors, respective reached position */
+	if (mdev->state.conn == C_VERIFY_S || mdev->state.conn == C_VERIFY_T)
+		db -= mdev->ov_left;
+
 	dbdt = Bit2KB(db/dt);
 	mdev->rs_paused /= HZ;
 
@@ -815,7 +830,7 @@ int drbd_resync_finished(struct drbd_conf *mdev)
 	ns.conn = C_CONNECTED;
 
 	dev_info(DEV, "%s done (total %lu sec; paused %lu sec; %lu K/sec)\n",
-	     verify_done ? "Online verify " : "Resync",
+	     verify_done ? "Online verify" : "Resync",
 	     dt + mdev->rs_paused, mdev->rs_paused, dbdt);
 
 	n_oos = drbd_bm_total_weight(mdev);
@@ -896,7 +911,9 @@ out:
 	mdev->rs_total  = 0;
 	mdev->rs_failed = 0;
 	mdev->rs_paused = 0;
-	if (verify_done)
+
+	/* reset start sector, if we reached end of device */
+	if (verify_done && mdev->ov_left == 0)
 		mdev->ov_start_sector = 0;
 
 	drbd_md_sync(mdev);
@@ -1144,6 +1161,7 @@ int w_e_end_ov_reply(struct drbd_work *w, int cancel)
 	unsigned int size = peer_req->i.size;
 	int digest_size;
 	int err, eq = 0;
+	bool stop_sector_reached = false;
 
 	if (unlikely(cancel)) {
 		drbd_free_peer_req(mdev, peer_req);
@@ -1194,7 +1212,10 @@ int w_e_end_ov_reply(struct drbd_work *w, int cancel)
 	if ((mdev->ov_left & 0x200) == 0x200)
 		drbd_advance_rs_marks(mdev, mdev->ov_left);
 
-	if (mdev->ov_left == 0) {
+	stop_sector_reached = verify_can_do_stop_sector(mdev) &&
+		(sector + (size>>9)) >= mdev->ov_stop_sector;
+
+	if (mdev->ov_left == 0 || stop_sector_reached) {
 		ov_out_of_sync_print(mdev);
 		drbd_resync_finished(mdev);
 	}