Headers-first synchronization

Many changes:
* Do not use 'getblocks', but 'getheaders', and use it to build a headers tree.
* Blocks are fetched in parallel from all available outbound peers, using a
  limited moving window. When one peer stalls the movement of the window, it is
  disconnected.
* No more orphan blocks. At all. We only ever request a block for which we have
  verified the headers, and store it to disk immediately. This means that a
  disk-fill attack would require PoW.
* Require protocol version 31800 for every peer (released in december 2010).
* No more syncnode (we sync from everyone we can, though limited to 1 during
  initial *headers* sync).
* Introduce some extra named constants, comments and asserts.

src/main.h

@@ -72,9 +72,17 @@ static const int MAX_SCRIPTCHECK_THREADS = 16;
 /** -par default (number of script-checking threads, 0 = auto) */
 static const int DEFAULT_SCRIPTCHECK_THREADS = 0;
 /** Number of blocks that can be requested at any given time from a single peer. */
-static const int MAX_BLOCKS_IN_TRANSIT_PER_PEER = 128;
-/** Timeout in seconds before considering a block download peer unresponsive. */
-static const unsigned int BLOCK_DOWNLOAD_TIMEOUT = 60;
+static const int MAX_BLOCKS_IN_TRANSIT_PER_PEER = 16;
+/** Timeout in seconds during which a peer must stall block download progress before being disconnected. */
+static const unsigned int BLOCK_STALLING_TIMEOUT = 2;
+/** Number of headers sent in one getheaders result. We rely on the assumption that if a peer sends
+ *  less than this number, we reached their tip. Changing this value is a protocol upgrade. */
+static const unsigned int MAX_HEADERS_RESULTS = 2000;
+/** Size of the "block download window": how far ahead of our current height do we fetch?
+ *  Larger windows tolerate larger download speed differences between peer, but increase the potential
+ *  degree of disordering of blocks on disk (which make reindexing and in the future perhaps pruning
+ *  harder). We'll probably want to make this a per-peer adaptive value at some point. */
+static const unsigned int BLOCK_DOWNLOAD_WINDOW = 1024;
 
 /** "reject" message codes **/
 static const unsigned char REJECT_MALFORMED = 0x01;
@@ -137,8 +145,6 @@ void RegisterNodeSignals(CNodeSignals& nodeSignals);
 /** Unregister a network node */
 void UnregisterNodeSignals(CNodeSignals& nodeSignals);
 
-void PushGetBlocks(CNode* pnode, CBlockIndex* pindexBegin, uint256 hashEnd);
-
 /** Process an incoming block */
 bool ProcessBlock(CValidationState &state, CNode* pfrom, CBlock* pblock, CDiskBlockPos *dbp = NULL);
 /** Check whether enough disk space is available for an incoming block */
@@ -439,9 +445,6 @@ bool DisconnectBlock(CBlock& block, CValidationState& state, CBlockIndex* pindex
 // Apply the effects of this block (with given index) on the UTXO set represented by coins
 bool ConnectBlock(CBlock& block, CValidationState& state, CBlockIndex* pindex, CCoinsViewCache& coins, bool fJustCheck = false);
 
-// Add this block to the block index, and if necessary, switch the active block chain to this
-bool AddToBlockIndex(CBlock& block, CValidationState& state, const CDiskBlockPos& pos);
-
 // Context-independent validity checks
 bool CheckBlockHeader(const CBlockHeader& block, CValidationState& state, bool fCheckPOW = true);
 bool CheckBlock(const CBlock& block, CValidationState& state, bool fCheckPOW = true, bool fCheckMerkleRoot = true);
@@ -449,7 +452,7 @@ bool CheckBlock(const CBlock& block, CValidationState& state, bool fCheckPOW = t
 // Store block on disk
 // if dbp is provided, the file is known to already reside on disk
 bool AcceptBlock(CBlock& block, CValidationState& state, CBlockIndex **pindex, CDiskBlockPos* dbp = NULL);
-bool AcceptBlockHeader(CBlockHeader& block, CValidationState& state, CBlockIndex **ppindex= NULL);
+bool AcceptBlockHeader(const CBlockHeader& block, CValidationState& state, CBlockIndex **ppindex= NULL);