fuse: improve BATCH_FORGET debug log output

Node numbers are now written as "i1234" everywhere except in the
BATCH_FORGET output, fix that. Also show the transaction number
so we can follow where each forget comes from.

As a normal FORGET looks like this:

	2019/01/27 14:38:07 rx 1488: FORGET i822 {Nlookup=1}

with the new output, "grep FORGET i1234 " will catch both single forgets
and batch forgets.

Before:

	2019/01/20 18:15:31 rx 1372: BATCH_FORGET i0 {Count=3}  48b
	2019/01/20 18:15:31 doBatchForget: forgetting 1 of 3: NodeId: 40, Nlookup: 1
	2019/01/20 18:15:31 doBatchForget: forgetting 2 of 3: NodeId: 41, Nlookup: 1
	2019/01/20 18:15:31 doBatchForget: forgetting 3 of 3: NodeId: 42, Nlookup: 1

After:

	2019/01/27 14:48:35 rx 2982: BATCH_FORGET i0 {Count=2}  32b
	2019/01/27 14:48:35 doBatchForget: rx 2982 1/2: FORGET i804 {Nlookup=1}
	2019/01/27 14:48:35 doBatchForget: rx 2982 2/2: FORGET i803 {Nlookup=1}

fuse/opcode.go

@@ -313,7 +313,8 @@ func doBatchForget(server *Server, req *request) {
 	forgets := *(*[]_ForgetOne)(unsafe.Pointer(h))
 	for i, f := range forgets {
 		if server.opts.Debug {
-			log.Printf("doBatchForget: forgetting %d of %d: NodeId: %d, Nlookup: %d", i+1, len(forgets), f.NodeId, f.Nlookup)
+			log.Printf("doBatchForget: rx %d %d/%d: FORGET i%d {Nlookup=%d}",
+				req.inHeader.Unique, i+1, len(forgets), f.NodeId, f.Nlookup)
 		}
 		if f.NodeId == pollHackInode {
 			continue