blockstorage: Add prune locks to BlockManager

This change also introduces an aditional buffer of 10 blocks (PRUNE_LOCK_BUFFER) that will not be pruned before the best block.

Co-authored-by: Luke Dashjr <luke-jr+git@utopios.org>

src/node/blockstorage.cpp

@@ -21,6 +21,8 @@
 #include <util/system.h>
 #include <validation.h>
 
+#include <unordered_map>
+
 namespace node {
 std::atomic_bool fImporting(false);
 std::atomic_bool fReindex(false);
@@ -230,6 +232,11 @@ void BlockManager::FindFilesToPrune(std::set<int>& setFilesToPrune, uint64_t nPr
            nLastBlockWeCanPrune, count);
 }
 
+void BlockManager::UpdatePruneLock(const std::string& name, const PruneLockInfo& lock_info) {
+    AssertLockHeld(::cs_main);
+    m_prune_locks[name] = lock_info;
+}
+
 CBlockIndex* BlockManager::InsertBlockIndex(const uint256& hash)
 {
     AssertLockHeld(cs_main);

src/node/blockstorage.h

@@ -13,6 +13,7 @@
 
 #include <atomic>
 #include <cstdint>
+#include <unordered_map>
 #include <vector>
 
 extern RecursiveMutex cs_main;
@@ -65,6 +66,10 @@ struct CBlockIndexHeightOnlyComparator {
     bool operator()(const CBlockIndex* pa, const CBlockIndex* pb) const;
 };
 
+struct PruneLockInfo {
+    int height_first{std::numeric_limits<int>::max()}; //! Height of earliest block that should be kept and not pruned
+};
+
 /**
  * Maintains a tree of blocks (stored in `m_block_index`) which is consulted
  * to determine where the most-work tip is.
@@ -118,6 +123,14 @@ private:
     /** Dirty block file entries. */
     std::set<int> m_dirty_fileinfo;
 
+    /**
+     * Map from external index name to oldest block that must not be pruned.
+     *
+     * @note Internally, only blocks at height (height_first - PRUNE_LOCK_BUFFER - 1) and
+     * below will be pruned, but callers should avoid assuming any particular buffer size.
+     */
+    std::unordered_map<std::string, PruneLockInfo> m_prune_locks GUARDED_BY(::cs_main);
+
 public:
     BlockMap m_block_index GUARDED_BY(cs_main);
 
@@ -175,6 +188,9 @@ public:
     //! Check whether the block associated with this index entry is pruned or not.
     bool IsBlockPruned(const CBlockIndex* pblockindex) EXCLUSIVE_LOCKS_REQUIRED(::cs_main);
 
+    //! Create or update a prune lock identified by its name
+    void UpdatePruneLock(const std::string& name, const PruneLockInfo& lock_info) EXCLUSIVE_LOCKS_REQUIRED(::cs_main);
+
     ~BlockManager()
     {
         Unload();

src/validation.cpp

@@ -106,6 +106,12 @@ const std::vector<std::string> CHECKLEVEL_DOC {
     "level 4 tries to reconnect the blocks",
     "each level includes the checks of the previous levels",
 };
+/** The number of blocks to keep below the deepest prune lock.
+ *  There is nothing special about this number. It is higher than what we
+ *  expect to see in regular mainnet reorgs, but not so high that it would
+ *  noticeably interfere with the pruning mechanism.
+ * */
+static constexpr int PRUNE_LOCK_BUFFER{10};
 
 /**
  * Mutex to guard access to validation specific variables, such as reading
@@ -2338,13 +2344,29 @@ bool CChainState::FlushStateToDisk(
         CoinsCacheSizeState cache_state = GetCoinsCacheSizeState();
         LOCK(m_blockman.cs_LastBlockFile);
         if (fPruneMode && (m_blockman.m_check_for_pruning || nManualPruneHeight > 0) && !fReindex) {
-            // make sure we don't prune above the blockfilterindexes bestblocks
+            // make sure we don't prune above any of the prune locks bestblocks
             // pruning is height-based
-            int last_prune = m_chain.Height(); // last height we can prune
+            int last_prune{m_chain.Height()}; // last height we can prune
+            std::optional<std::string> limiting_lock; // prune lock that actually was the limiting factor, only used for logging
+
             ForEachBlockFilterIndex([&](BlockFilterIndex& index) {
                 last_prune = std::max(1, std::min(last_prune, index.GetSummary().best_block_height));
             });
 
+            for (const auto& prune_lock : m_blockman.m_prune_locks) {
+                if (prune_lock.second.height_first == std::numeric_limits<int>::max()) continue;
+                // Remove the buffer and one additional block here to get actual height that is outside of the buffer
+                const int lock_height{prune_lock.second.height_first - PRUNE_LOCK_BUFFER - 1};
+                last_prune = std::max(1, std::min(last_prune, lock_height));
+                if (last_prune == lock_height) {
+                    limiting_lock = prune_lock.first;
+                }
+            }
+
+            if (limiting_lock) {
+                LogPrint(BCLog::PRUNE, "%s limited pruning to height %d\n", limiting_lock.value(), last_prune);
+            }
+
             if (nManualPruneHeight > 0) {
                 LOG_TIME_MILLIS_WITH_CATEGORY("find files to prune (manual)", BCLog::BENCH);
 
@@ -2581,6 +2603,18 @@ bool CChainState::DisconnectTip(BlockValidationState& state, DisconnectedBlockTr
         assert(flushed);
     }
     LogPrint(BCLog::BENCH, "- Disconnect block: %.2fms\n", (GetTimeMicros() - nStart) * MILLI);
+
+    {
+        // Prune locks that began at or after the tip should be moved backward so they get a chance to reorg
+        const int max_height_first{pindexDelete->nHeight - 1};
+        for (auto& prune_lock : m_blockman.m_prune_locks) {
+            if (prune_lock.second.height_first <= max_height_first) continue;
+
+            prune_lock.second.height_first = max_height_first;
+            LogPrint(BCLog::PRUNE, "%s prune lock moved back to %d\n", prune_lock.first, max_height_first);
+        }
+    }
+
     // Write the chain state to disk, if necessary.
     if (!FlushStateToDisk(state, FlushStateMode::IF_NEEDED)) {
         return false;