drbd: Do not full sync if a P_SYNC_UUID packet gets lost

See also commit from 2009-08-15
"drbd_uuid_compare(): Do not full sync in case a P_SYNC_UUID packet gets lost."

We saw cases where the History UUIDs where not as expected. So the
detection of the special case did not trigger. With the sync UUID
no longer being a random number, but deducible from the previous
bitmap UUID, the detection of this special case becomes more
reliable.

The SyncUUID now is the previous bitmap UUID + 0x1000000000000.

Rule 5a:
Cs = H1p & H1p + Offset = Bp
  Connection was lost before SyncUUID Packet came through.
  Corrent (peer) UUIDs:
   Bp = H1p
   H1p = H2p
   H2p = 0
  Become Sync target.

Rule 7a:
Cp = H1s & H1s + Offset = Bs
  Connection was lost before SyncUUID Packet came through.
  Correct (own) UUIDs:
   Bs = H1s
   H1s = H2s
   H2s = 0
  Become Sync source.

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

3 files changed, 23 insertions, 15 deletions

diff --git a/drivers/block/drbd/drbd_int.h b/drivers/block/drbd/drbd_int.h
index 84b4575fdf4..3c8eecd9666 100644
--- a/drivers/block/drbd/drbd_int.h
+++ b/drivers/block/drbd/drbd_int.h
@@ -97,6 +97,7 @@ extern char usermode_helper[];
 #define ID_SYNCER (-1ULL)
 #define ID_VACANT 0
 #define is_syncer_block_id(id) ((id) == ID_SYNCER)
+#define UUID_NEW_BM_OFFSET ((u64)0x0001000000000000ULL)
 
 struct drbd_conf;
 
diff --git a/drivers/block/drbd/drbd_main.c b/drivers/block/drbd/drbd_main.c
index da98bff7c33..b3b6d3190f6 100644
--- a/drivers/block/drbd/drbd_main.c
+++ b/drivers/block/drbd/drbd_main.c
@@ -2022,7 +2022,7 @@ int drbd_gen_and_send_sync_uuid(struct drbd_conf *mdev)
 
 	D_ASSERT(mdev->state.disk == D_UP_TO_DATE);
 
-	get_random_bytes(&uuid, sizeof(u64));
+	uuid = mdev->ldev->md.uuid[UI_BITMAP] + UUID_NEW_BM_OFFSET;
 	drbd_uuid_set(mdev, UI_BITMAP, uuid);
 	drbd_md_sync(mdev);
 	p.uuid = cpu_to_be64(uuid);
diff --git a/drivers/block/drbd/drbd_receiver.c b/drivers/block/drbd/drbd_receiver.c
index fd0957f9c23..7991183749e 100644
--- a/drivers/block/drbd/drbd_receiver.c
+++ b/drivers/block/drbd/drbd_receiver.c
@@ -2293,6 +2293,8 @@ static void drbd_uuid_dump(struct drbd_conf *mdev, char *text, u64 *uuid,
    -2	C_SYNC_TARGET set BitMap
  -100	after split brain, disconnect
 -1000	unrelated data
+-1091   requires proto 91
+-1096   requires proto 96
  */
 static int drbd_uuid_compare(struct drbd_conf *mdev, int *rule_nr) __must_hold(local)
 {
@@ -2322,7 +2324,7 @@ static int drbd_uuid_compare(struct drbd_conf *mdev, int *rule_nr) __must_hold(l
 		if (mdev->p_uuid[UI_BITMAP] == (u64)0 && mdev->ldev->md.uuid[UI_BITMAP] != (u64)0) {
 
 			if (mdev->agreed_pro_version < 91)
-				return -1001;
+				return -1091;
 
 			if ((mdev->ldev->md.uuid[UI_BITMAP] & ~((u64)1)) == (mdev->p_uuid[UI_HISTORY_START] & ~((u64)1)) &&
 			    (mdev->ldev->md.uuid[UI_HISTORY_START] & ~((u64)1)) == (mdev->p_uuid[UI_HISTORY_START + 1] & ~((u64)1))) {
@@ -2343,7 +2345,7 @@ static int drbd_uuid_compare(struct drbd_conf *mdev, int *rule_nr) __must_hold(l
 		if (mdev->ldev->md.uuid[UI_BITMAP] == (u64)0 && mdev->p_uuid[UI_BITMAP] != (u64)0) {
 
 			if (mdev->agreed_pro_version < 91)
-				return -1001;
+				return -1091;
 
 			if ((mdev->ldev->md.uuid[UI_HISTORY_START] & ~((u64)1)) == (mdev->p_uuid[UI_BITMAP] & ~((u64)1)) &&
 			    (mdev->ldev->md.uuid[UI_HISTORY_START + 1] & ~((u64)1)) == (mdev->p_uuid[UI_HISTORY_START] & ~((u64)1))) {
@@ -2388,17 +2390,22 @@ static int drbd_uuid_compare(struct drbd_conf *mdev, int *rule_nr) __must_hold(l
 	*rule_nr = 51;
 	peer = mdev->p_uuid[UI_HISTORY_START] & ~((u64)1);
 	if (self == peer) {
-		self = mdev->ldev->md.uuid[UI_HISTORY_START] & ~((u64)1);
-		peer = mdev->p_uuid[UI_HISTORY_START + 1] & ~((u64)1);
-		if (self == peer) {
+		if (mdev->agreed_pro_version < 96 ?
+		    (mdev->ldev->md.uuid[UI_HISTORY_START] & ~((u64)1)) ==
+		    (mdev->p_uuid[UI_HISTORY_START + 1] & ~((u64)1)) :
+		    peer + UUID_NEW_BM_OFFSET == (mdev->p_uuid[UI_BITMAP] & ~((u64)1))) {
 			/* The last P_SYNC_UUID did not get though. Undo the last start of
 			   resync as sync source modifications of the peer's UUIDs. */
 
 			if (mdev->agreed_pro_version < 91)
-				return -1001;
+				return -1091;
 
 			mdev->p_uuid[UI_BITMAP] = mdev->p_uuid[UI_HISTORY_START];
 			mdev->p_uuid[UI_HISTORY_START] = mdev->p_uuid[UI_HISTORY_START + 1];
+
+			dev_info(DEV, "Did not got last syncUUID packet, corrected:\n");
+			drbd_uuid_dump(mdev, "peer", mdev->p_uuid, mdev->p_uuid[UI_SIZE], mdev->p_uuid[UI_FLAGS]);
+
 			return -1;
 		}
 	}
@@ -2420,20 +2427,20 @@ static int drbd_uuid_compare(struct drbd_conf *mdev, int *rule_nr) __must_hold(l
 	*rule_nr = 71;
 	self = mdev->ldev->md.uuid[UI_HISTORY_START] & ~((u64)1);
 	if (self == peer) {
-		self = mdev->ldev->md.uuid[UI_HISTORY_START + 1] & ~((u64)1);
-		peer = mdev->p_uuid[UI_HISTORY_START] & ~((u64)1);
-		if (self == peer) {
+		if (mdev->agreed_pro_version < 96 ?
+		    (mdev->ldev->md.uuid[UI_HISTORY_START + 1] & ~((u64)1)) ==
+		    (mdev->p_uuid[UI_HISTORY_START] & ~((u64)1)) :
+		    self + UUID_NEW_BM_OFFSET == (mdev->ldev->md.uuid[UI_BITMAP] & ~((u64)1))) {
 			/* The last P_SYNC_UUID did not get though. Undo the last start of
 			   resync as sync source modifications of our UUIDs. */
 
 			if (mdev->agreed_pro_version < 91)
-				return -1001;
+				return -1091;
 
 			_drbd_uuid_set(mdev, UI_BITMAP, mdev->ldev->md.uuid[UI_HISTORY_START]);
 			_drbd_uuid_set(mdev, UI_HISTORY_START, mdev->ldev->md.uuid[UI_HISTORY_START + 1]);
 
-			dev_info(DEV, "Undid last start of resync:\n");
-
+			dev_info(DEV, "Last syncUUID did not get through, corrected:\n");
 			drbd_uuid_dump(mdev, "self", mdev->ldev->md.uuid,
 				       mdev->state.disk >= D_NEGOTIATING ? drbd_bm_total_weight(mdev) : 0, 0);
 
@@ -2496,8 +2503,8 @@ static enum drbd_conns drbd_sync_handshake(struct drbd_conf *mdev, enum drbd_rol
 		dev_alert(DEV, "Unrelated data, aborting!\n");
 		return C_MASK;
 	}
-	if (hg == -1001) {
-		dev_alert(DEV, "To resolve this both sides have to support at least protocol 91\n");
+	if (hg < -1000) {
+		dev_alert(DEV, "To resolve this both sides have to support at least protocol %d\n", -hg - 1000);
 		return C_MASK;
 	}