Merge bitcoin/bitcoin#25667: assumeutxo: snapshot initialization

bf9597606166323158bbf631137b82d41f39334f doc: add note about snapshot chainstate init (James O'Beirne)
e4d799528696c5ede38c257afaffd367917e0de8 test: add testcases for snapshot initialization (James O'Beirne)
cced4e7336d93a2dc88e4a61c49941887766bd72 test: move-only-ish: factor out LoadVerifyActivateChainstate() (James O'Beirne)
51fc9241c08a00f1f407f1534853a5cddbbc0a23 test: allow on-disk coins and block tree dbs in tests (James O'Beirne)
3c361391b8f5971eb3c7b620aa7ad9b437cc515e test: add reset_chainstate parameter for snapshot unittests (James O'Beirne)
00b357c215ed900145bd770525a341ba0ed9c027 validation: add ResetChainstates() (James O'Beirne)
3a29dfbfb2c16a50d854f6f81428a68aa9180509 move-only: test: make snapshot chainstate setup reusable (James O'Beirne)
8153bd9247dad3982d54488bcdb3960470315290 blockmanager: avoid undefined behavior during FlushBlockFile (James O'Beirne)
ad67ff377c2b271cb4683da2fb25fd295557f731 validation: remove snapshot datadirs upon validation failure (James O'Beirne)
34d159033106cc595cfa852695610bfe419c989c add utilities for deleting on-disk leveldb data (James O'Beirne)
252abd1e8bc5cdf4368ad55e827a873240535b28 init: add utxo snapshot detection (James O'Beirne)
f9f1735f139b6a1f1c7fea50717ff90dc4ba2bce validation: rename snapshot chainstate dir (James O'Beirne)
d14bebf100aaaa25c7558eeed8b5c536da99885f db: add StoragePath to CDBWrapper/CCoinsViewDB (James O'Beirne)

Pull request description:

  This is part of the [assumeutxo project](https://github.com/bitcoin/bitcoin/projects/11) (parent PR: https://github.com/bitcoin/bitcoin/pull/15606)

  ---

  Half of the replacement for #24232. The original PR grew larger than expected throughout the review process.

  This change adds the ability to initialize a snapshot-based chainstate during init if one is detected on disk. This is of course unused as of now (aside from in unittests) given that we haven't yet enabled actually loading snapshots.

  Don't be scared! There are some big move-only commits in here.

  Accompanying changes include:

  - moving the snapshot coinsdb directory from being called `chainstate_[base blockhash]` to `chainstate_snapshot`, since we only support one snapshot in use at a time. This simplifies some logic, but it necessitates writing that base blockhash out to a file within the coinsdb dir. See [discussion here](https://github.com/bitcoin/bitcoin/pull/24232#discussion_r832762880).
  - adding a simple fix in `FlushBlockFile()` that avoids a crash when attemping to flush to disk before `LoadBlockIndexDB()` is called, which happens when calling `MaybeRebalanceCaches()` during multiple chainstate init.
  - improving the unittest to allow testing with on-disk chainstates - necessary to test a simulated restart and re-initialization.

ACKs for top commit:
  naumenkogs:
    utACK bf9597606166323158bbf631137b82d41f39334f
  ariard:
    Code Review ACK bf9597606
  ryanofsky:
    Code review ACK bf9597606166323158bbf631137b82d41f39334f. Changes since last review: rebasing, switching from CAutoFile to AutoFile, adding comments, switching from BOOST_CHECK to Assert in test util, using chainman.GetMutex() in tests, destroying one ChainstateManager before creating a new one in tests
  fjahr:
    utACK bf9597606166323158bbf631137b82d41f39334f
  aureleoules:
    ACK bf9597606166323158bbf631137b82d41f39334f

Tree-SHA512: 15ae75caf19f8d12a12d2647c52897904d27b265a7af6b4ae7b858592eeadb8f9da6c2394b6baebec90adc28742c053e3eb506119577dae7c1e722ebb3b7bcc0

1 files changed, 107 insertions, 22 deletions

diff --git a/src/validation.cpp b/src/validation.cpp
index be2836e603..4941d2bcb6 100644
--- a/src/validation.cpp
+++ b/src/validation.cpp
@@ -1513,7 +1513,7 @@ void Chainstate::InitCoinsDB(
     fs::path leveldb_name)
 {
     if (m_from_snapshot_blockhash) {
-        leveldb_name += "_" + m_from_snapshot_blockhash->ToString();
+        leveldb_name += node::SNAPSHOT_CHAINSTATE_SUFFIX;
     }
 
     m_coins_views = std::make_unique<CoinsViews>(
@@ -4804,28 +4804,15 @@ std::vector<Chainstate*> ChainstateManager::GetAll()
     return out;
 }
 
-Chainstate& ChainstateManager::InitializeChainstate(
-    CTxMemPool* mempool, const std::optional<uint256>& snapshot_blockhash)
+Chainstate& ChainstateManager::InitializeChainstate(CTxMemPool* mempool)
 {
     AssertLockHeld(::cs_main);
-    bool is_snapshot = snapshot_blockhash.has_value();
-    std::unique_ptr<Chainstate>& to_modify =
-        is_snapshot ? m_snapshot_chainstate : m_ibd_chainstate;
+    assert(!m_ibd_chainstate);
+    assert(!m_active_chainstate);
 
-    if (to_modify) {
-        throw std::logic_error("should not be overwriting a chainstate");
-    }
-    to_modify.reset(new Chainstate(mempool, m_blockman, *this, snapshot_blockhash));
-
-    // Snapshot chainstates and initial IBD chaintates always become active.
-    if (is_snapshot || (!is_snapshot && !m_active_chainstate)) {
-        LogPrintf("Switching active chainstate to %s\n", to_modify->ToString());
-        m_active_chainstate = to_modify.get();
-    } else {
-        throw std::logic_error("unexpected chainstate activation");
-    }
-
-    return *to_modify;
+    m_ibd_chainstate = std::make_unique<Chainstate>(mempool, m_blockman, *this);
+    m_active_chainstate = m_ibd_chainstate.get();
+    return *m_active_chainstate;
 }
 
 const AssumeutxoData* ExpectedAssumeutxo(
@@ -4840,6 +4827,46 @@ const AssumeutxoData* ExpectedAssumeutxo(
     return nullptr;
 }
 
+static bool DeleteCoinsDBFromDisk(const fs::path db_path, bool is_snapshot)
+    EXCLUSIVE_LOCKS_REQUIRED(::cs_main)
+{
+    AssertLockHeld(::cs_main);
+
+    if (is_snapshot) {
+        fs::path base_blockhash_path = db_path / node::SNAPSHOT_BLOCKHASH_FILENAME;
+
+        if (fs::exists(base_blockhash_path)) {
+            bool removed = fs::remove(base_blockhash_path);
+            if (!removed) {
+                LogPrintf("[snapshot] failed to remove file %s\n",
+                          fs::PathToString(base_blockhash_path));
+            }
+        } else {
+            LogPrintf("[snapshot] snapshot chainstate dir being removed lacks %s file\n",
+                    fs::PathToString(node::SNAPSHOT_BLOCKHASH_FILENAME));
+        }
+    }
+
+    std::string path_str = fs::PathToString(db_path);
+    LogPrintf("Removing leveldb dir at %s\n", path_str);
+
+    // We have to destruct before this call leveldb::DB in order to release the db
+    // lock, otherwise `DestroyDB` will fail. See `leveldb::~DBImpl()`.
+    const bool destroyed = dbwrapper::DestroyDB(path_str, {}).ok();
+
+    if (!destroyed) {
+        LogPrintf("error: leveldb DestroyDB call failed on %s\n", path_str);
+    }
+
+    // Datadir should be removed from filesystem; otherwise initialization may detect
+    // it on subsequent statups and get confused.
+    //
+    // If the base_blockhash_path removal above fails in the case of snapshot
+    // chainstates, this will return false since leveldb won't remove a non-empty
+    // directory.
+    return destroyed && !fs::exists(db_path);
+}
+
 bool ChainstateManager::ActivateSnapshot(
         AutoFile& coins_file,
         const SnapshotMetadata& metadata,
@@ -4897,11 +4924,34 @@ bool ChainstateManager::ActivateSnapshot(
             static_cast<size_t>(current_coinstip_cache_size * SNAPSHOT_CACHE_PERC));
     }
 
-    const bool snapshot_ok = this->PopulateAndValidateSnapshot(
+    bool snapshot_ok = this->PopulateAndValidateSnapshot(
         *snapshot_chainstate, coins_file, metadata);
 
+    // If not in-memory, persist the base blockhash for use during subsequent
+    // initialization.
+    if (!in_memory) {
+        LOCK(::cs_main);
+        if (!node::WriteSnapshotBaseBlockhash(*snapshot_chainstate)) {
+            snapshot_ok = false;
+        }
+    }
     if (!snapshot_ok) {
-        WITH_LOCK(::cs_main, this->MaybeRebalanceCaches());
+        LOCK(::cs_main);
+        this->MaybeRebalanceCaches();
+
+        // PopulateAndValidateSnapshot can return (in error) before the leveldb datadir
+        // has been created, so only attempt removal if we got that far.
+        if (auto snapshot_datadir = node::FindSnapshotChainstateDir()) {
+            // We have to destruct leveldb::DB in order to release the db lock, otherwise
+            // DestroyDB() (in DeleteCoinsDBFromDisk()) will fail. See `leveldb::~DBImpl()`.
+            // Destructing the chainstate (and so resetting the coinsviews object) does this.
+            snapshot_chainstate.reset();
+            bool removed = DeleteCoinsDBFromDisk(*snapshot_datadir, /*is_snapshot=*/true);
+            if (!removed) {
+                AbortNode(strprintf("Failed to remove snapshot chainstate dir (%s). "
+                    "Manually remove it before restarting.\n", fs::PathToString(*snapshot_datadir)));
+            }
+        }
         return false;
     }
 
@@ -5175,6 +5225,13 @@ void ChainstateManager::MaybeRebalanceCaches()
     }
 }
 
+void ChainstateManager::ResetChainstates()
+{
+    m_ibd_chainstate.reset();
+    m_snapshot_chainstate.reset();
+    m_active_chainstate = nullptr;
+}
+
 ChainstateManager::~ChainstateManager()
 {
     LOCK(::cs_main);
@@ -5186,3 +5243,31 @@ ChainstateManager::~ChainstateManager()
         i.clear();
     }
 }
+
+bool ChainstateManager::DetectSnapshotChainstate(CTxMemPool* mempool)
+{
+    assert(!m_snapshot_chainstate);
+    std::optional<fs::path> path = node::FindSnapshotChainstateDir();
+    if (!path) {
+        return false;
+    }
+    std::optional<uint256> base_blockhash = node::ReadSnapshotBaseBlockhash(*path);
+    if (!base_blockhash) {
+        return false;
+    }
+    LogPrintf("[snapshot] detected active snapshot chainstate (%s) - loading\n",
+        fs::PathToString(*path));
+
+    this->ActivateExistingSnapshot(mempool, *base_blockhash);
+    return true;
+}
+
+Chainstate& ChainstateManager::ActivateExistingSnapshot(CTxMemPool* mempool, uint256 base_blockhash)
+{
+    assert(!m_snapshot_chainstate);
+    m_snapshot_chainstate =
+        std::make_unique<Chainstate>(mempool, m_blockman, *this, base_blockhash);
+    LogPrintf("[snapshot] switching active chainstate to %s\n", m_snapshot_chainstate->ToString());
+    m_active_chainstate = m_snapshot_chainstate.get();
+    return *m_snapshot_chainstate;
+}