Limit full_stack fuzz runtime by limiting block connection ops

The `full_stack_target` fuzzer is generally pretty slow, but on
some inputs the fuzzer is able to connect hundreds of thousands to
millions of blocks, which is by far the slowest operation. In some
cases, individual inputs can take tens of minutes to complete,
which is not a realistic issue in LDK (we don't expect there to be
million-block bursts in Bitcoin) so we simply disable it.

Author: TheBlueMatt

fuzz/src/full_stack.rs

@@ -299,6 +299,14 @@ impl<'a> MoneyLossDetector<'a> {
 	}
 
 	fn connect_block(&mut self, all_txn: &[Transaction]) {
+		if self.blocks_connected > 50_000 {
+			// Connecting blocks is relatively slow, and some commands can connect many blocks.
+			// This can inflate the total runtime substantially, leading to spurious timeouts.
+			// Instead, because block connection rate is expected to be limited by PoW, simply
+			// start ignoring blocks after the first 50k.
+			return;
+		}
+
 		let mut txdata = Vec::with_capacity(all_txn.len());
 		for (idx, tx) in all_txn.iter().enumerate() {
 			let txid = tx.compute_txid();