pytest: speed up test_restart_many_payments.

I thought this was timing out because I made it slow with the
change to txprepare as a plugin.  In fact, it was timing out
because sometimes gossip comes so fast it gets suppressed
and we never get the log messags.

Still, before this it took 98 seconds under valgrind and
24 under non-valgrind, so it's an improvement to time as
well as robustness.

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>

tests/test_connection.py

@@ -2000,40 +2000,52 @@ def test_fulfill_incoming_first(node_factory, bitcoind):
 def test_restart_many_payments(node_factory, bitcoind):
     l1 = node_factory.get_node(may_reconnect=True)
 
-    # On my laptop, these take 74 seconds and 44 seconds (with restart commented out)
+    # On my laptop, these take 89 seconds and 12 seconds
     if VALGRIND:
         num = 2
     else:
         num = 5
 
+    nodes = node_factory.get_nodes(num * 2, opts={'may_reconnect': True})
+    innodes = nodes[:num]
+    outnodes = nodes[num:]
+
+    # Fund up-front to save some time.
+    dests = {l1.rpc.newaddr()['bech32']: (10**6 + 1000000) / 10**8 * num}
+    for n in innodes:
+        dests[n.rpc.newaddr()['bech32']] = (10**6 + 1000000) / 10**8
+    bitcoind.rpc.sendmany("", dests)
+    bitcoind.generate_block(1)
+    sync_blockheight(bitcoind, [l1] + innodes)
+
     # Nodes with channels into the main node
-    innodes = node_factory.get_nodes(num, opts={'may_reconnect': True})
-    inchans = []
     for n in innodes:
         n.rpc.connect(l1.info['id'], 'localhost', l1.port)
-        inchans.append(n.fund_channel(l1, 10**6, False))
+        n.rpc.fundchannel(l1.info['id'], 10**6)
 
     # Nodes with channels out of the main node
-    outnodes = node_factory.get_nodes(len(innodes), opts={'may_reconnect': True})
-    outchans = []
     for n in outnodes:
-        n.rpc.connect(l1.info['id'], 'localhost', l1.port)
-        outchans.append(l1.fund_channel(n, 10**6, False))
+        l1.rpc.connect(n.info['id'], 'localhost', n.port)
+        # OK to use change from previous fundings
+        l1.rpc.fundchannel(n.info['id'], 10**6, minconf=0)
 
-    # Make sure they're all announced.
-    bitcoind.generate_block(5)
+    # Now mine them, get scids.
+    bitcoind.generate_block(6, wait_for_mempool=num * 2)
+    sync_blockheight(bitcoind, [l1] + nodes)
+
+    wait_for(lambda: [only_one(n.rpc.listpeers()['peers'])['channels'][0]['state'] for n in nodes] == ['CHANNELD_NORMAL'] * len(nodes))
+
+    inchans = []
+    for n in innodes:
+        inchans.append(only_one(n.rpc.listpeers()['peers'])['channels'][0]['short_channel_id'])
+
+    outchans = []
+    for n in outnodes:
+        outchans.append(only_one(n.rpc.listpeers()['peers'])['channels'][0]['short_channel_id'])
 
-    # We wait for each node to see each dir active, and its own
-    # channel CHANNELD_NORMAL
-    logs = ([r'update for channel {}/0 now ACTIVE'.format(scid)
-             for scid in inchans + outchans]
-            + [r'update for channel {}/1 now ACTIVE'.format(scid)
-               for scid in inchans + outchans]
-            + ['to CHANNELD_NORMAL'])
-
-    # Now do all the waiting at once: if !DEVELOPER, this can be *very* slow!
-    for n in innodes + outnodes:
-        n.daemon.wait_for_logs(logs)
+    # Now make sure every node sees every channel.
+    for n in nodes + [l1]:
+        wait_for(lambda: [c['public'] for c in n.rpc.listchannels()['channels']] == [True] * len(nodes) * 2)
 
     # Manually create routes, get invoices
     Payment = namedtuple('Payment', ['innode', 'route', 'payment_hash'])