Use cluster linearization for transaction relay sort order

Previously, transaction batches were first sorted by ancestor count and then
feerate, to ensure transactions are announced in a topologically valid order,
while prioritizing higher feerate transactions. Ancestor count is a crude
topological sort criteria, so replace this with linearization order so that the
highest feerate transactions (as would be observed by the mining algorithm) are
relayed before lower feerate ones, in a topologically valid way.

This also fixes a test that only worked due to the ancestor-count-based sort
order.

test/functional/mempool_packages.py

@@ -245,21 +245,17 @@ class MempoolPackagesTest(BitcoinTestFramework):
         mempool1 = self.nodes[1].getrawmempool(False)
         assert set(mempool1).issubset(set(mempool0))
         assert parent_transaction in mempool1
-        # Note: this test is brittle, because it relies on the relay sort order
-        # of node0 to be based on ancestor count (so that the first 10
-        # descendants of parent_transaction relay before the later ones).
-        for tx in chain[:CUSTOM_DESCENDANT_LIMIT-1]:
-            assert tx in mempool1
-        for tx in chain[CUSTOM_DESCENDANT_LIMIT:]:
-            assert tx not in mempool1
-        for tx in mempool1:
-            entry0 = self.nodes[0].getmempoolentry(tx)
-            entry1 = self.nodes[1].getmempoolentry(tx)
-            assert not entry0['unbroadcast']
-            assert not entry1['unbroadcast']
-            assert_equal(entry1['fees']['base'], entry0['fees']['base'])
-            assert_equal(entry1['vsize'], entry0['vsize'])
-            assert_equal(entry1['depends'], entry0['depends'])
+        for tx in chain:
+            if tx in mempool1:
+                entry0 = self.nodes[0].getmempoolentry(tx)
+                entry1 = self.nodes[1].getmempoolentry(tx)
+                assert not entry0['unbroadcast']
+                assert not entry1['unbroadcast']
+                assert entry1["descendantcount"] <= CUSTOM_DESCENDANT_LIMIT
+                assert_equal(entry1['fees']['base'], entry0['fees']['base'])
+                assert_equal(entry1['vsize'], entry0['vsize'])
+                assert_equal(entry1['depends'], entry0['depends'])
+
         # Test reorg handling
         # First, the basics:
         self.generate(self.nodes[0], 1)