make sync faster because of massive block processing is executed in a sperated thread
From now during massive sync blocks are taken from HAF in the main thread, then processed in another thread. The main loop do not wait to finish processing a batch of blocks, but backs to getting new blocks while processing is in progress.