Qortal/altcoinj
3
0
Fork 0
mirror of https://github.com/Qortal/altcoinj.git synced 2025-02-14 19:25:51 +00:00
Code Issues Projects Releases Wiki Activity

Add top 50 blocks to the block locator. Should make chain split handling more robust.

Browse Source
This commit is contained in:
Mike Hearn 2012-02-13 23:39:23 +01:00
parent 2f72eb64d7
commit 0e198e68e8
2 changed files with 35 additions and 19 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
src/com/google/bitcoin/core/BlockChain.java
View File

@ -118,6 +118,13 @@ public class BlockChain {
wallets.add(wallet);
}
/**
* Returns the {@link BlockStore} the chain was constructed with. You can use this to iterate over the chain.
*/
public BlockStore getBlockStore() {
return blockStore;
}
/**
* Processes a received block and tries to add it to the chain. If there's something wrong with the block an
* exception is thrown. If the block is OK but cannot be connected to the chain at this time, returns false.

47
src/com/google/bitcoin/core/Peer.java
View File

@ -16,12 +16,17 @@
package com.google.bitcoin.core;
import com.google.bitcoin.store.BlockStore;
import com.google.bitcoin.store.BlockStoreException;
import com.google.bitcoin.utils.EventListenerInvoker;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import java.io.IOException;
import java.util.*;
import java.util.ArrayList;
import java.util.LinkedHashMap;
import java.util.List;
import java.util.Map;
import java.util.concurrent.*;
/**
@ -600,27 +605,31 @@ public class Peer {
log.info("blockChainDownload({})", toHash.toString());
// TODO: Block locators should be abstracted out rather than special cased here.
List<Sha256Hash> blockLocator = new LinkedList<Sha256Hash>();
// For now we don't do the exponential thinning as suggested here:
List<Sha256Hash> blockLocator = new ArrayList<Sha256Hash>(51);
// For now we don't do the exponential thinning as suggested here:
//
// https://en.bitcoin.it/wiki/Protocol_specification#getblocks
// However, this should be taken seriously going forward. The old implementation only added the hash of the
// genesis block and the current chain head, which randomly led us to halt block fetching when ending on a
// chain that turned out not to be the longest. This happened roughly once a week.
// Now we add three hashes to the locator:
// 1. Hash of genesis block
// 2. Hash of the block previous to the chain head
// 3. Hash of the chain head
// This allows our peer to see that we are on the wrong track if we ended up on the wrong side of a chain fork
// if the fork is only one block deep.
blockLocator.add(params.genesisBlock.getHash());
Block topBlock = blockChain.getChainHead().getHeader();
if (!topBlock.equals(params.genesisBlock)) {
if (!topBlock.getPrevBlockHash().equals(params.genesisBlock.getHash())){
blockLocator.add(0, topBlock.getPrevBlockHash());
//
// This is because it requires scanning all the block chain headers, which is very slow. Instead we add the top
// 50 block headers. If there is a re-org deeper than that, we'll end up downloading the entire chain. We
// must always put the genesis block as the first entry.
BlockStore store = blockChain.getBlockStore();
StoredBlock cursor = blockChain.getChainHead();
for (int i = 50; cursor != null && i > 0; i--) {
blockLocator.add(cursor.getHeader().getHash());
try {
cursor = cursor.getPrev(store);
} catch (BlockStoreException e) {
log.error("Failed to walk the block chain whilst constructing a locator");
throw new RuntimeException(e);
}
blockLocator.add(0, topBlock.getHash());
}
// The toHash field is set to zero already by the constructor.
// Only add the locator if we didn't already do so. If the chain is < 50 blocks we already reached it.
if (cursor != null) {
blockLocator.add(params.genesisBlock.getHash());
}
// The toHash field is set to zero already by the constructor. This is how we indicate "never stop".
if (downloadBlockBodies) {
GetBlocksMessage message = new GetBlocksMessage(params, blockLocator, toHash);