drbd: differentiate early and later "postponing" of requests

We use the RQ_POSTPONED flag to mark a request for several reasons.

It may be a conflicting request in a dual-primary setup,
where conflict detection and resolution on the peer decided that
this request needs to be re-submitted, it needs to re-enter
drbd_make_request() to fix the data divergence caused by these
conflicting, partially overlapping, quasi-simultaneous requests.

In this case we need to mark the corresponding area as out-of-sync,
before we call drbd_al_complete_io().

We also use the RQ_POSTPONED flag to just "push back" a request,
before even processing it, if IO is suspended for some reason.
In this case, as this request was neither submitted nor sent yet,
we must not touch the bitmap.

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

drivers/block/drbd/drbd_req.c

@@ -123,7 +123,14 @@ void drbd_req_destroy(struct kref *kref)
 		 * (local only or remote failed).
 		 * Other places where we set out-of-sync:
 		 * READ with local io-error */
-		if (!(s & RQ_POSTPONED)) {
+
+		/* There is a special case:
+		 * we may notice late that IO was suspended,
+		 * and postpone, or schedule for retry, a write,
+		 * before it even was submitted or sent.
+		 * In that case we do not want to touch the bitmap at all.
+		 */
+		if ((s & (RQ_POSTPONED|RQ_LOCAL_MASK|RQ_NET_MASK)) != RQ_POSTPONED) {
 			if (!(s & RQ_NET_OK) || !(s & RQ_LOCAL_OK))
 				drbd_set_out_of_sync(mdev, req->i.sector, req->i.size);