make GetFirstStoredBlock assert that 'start_block' always has data

And transfer the responsibility of verifying whether 'start_block'
has data or not to the caller.

This is because the 'GetFirstStoredBlock' function responsibility
is to return the first block containing data. And the current
implementation can return 'start_block' when it has no data!. Which
is misleading at least.

Edge case behavior change:
Previously, if the block tip lacked data but all preceding blocks
contained data, there was no prune violation. And now, such
scenario will result in a prune violation.

Author: furszy

src/index/base.cpp

@@ -115,7 +115,8 @@ bool BaseIndex::Init()
         if (!start_block) {
             // index is not built yet
             // make sure we have all block data back to the genesis
-            prune_violation = m_chainstate->m_blockman.GetFirstStoredBlock(*active_chain.Tip()) != active_chain.Genesis();
+            bool has_tip_data = active_chain.Tip()->nStatus & BLOCK_HAVE_DATA;
+            prune_violation = !has_tip_data || m_chainstate->m_blockman.GetFirstStoredBlock(*active_chain.Tip()) != active_chain.Genesis();
         }
         // in case the index has a best block set and is not fully synced
         // check if we have the required blocks to continue building the index

src/node/blockstorage.cpp

@@ -407,6 +407,7 @@ const CBlockIndex* BlockManager::GetFirstStoredBlock(const CBlockIndex& start_bl
 {
     AssertLockHeld(::cs_main);
     const CBlockIndex* last_block = &start_block;
+    assert(last_block->nStatus & BLOCK_HAVE_DATA);
     while (last_block->pprev && (last_block->pprev->nStatus & BLOCK_HAVE_DATA)) {
         last_block = last_block->pprev;
     }

src/node/blockstorage.h

@@ -222,7 +222,9 @@ class BlockManager
     //! Returns last CBlockIndex* that is a checkpoint
     const CBlockIndex* GetLastCheckpoint(const CCheckpointData& data) EXCLUSIVE_LOCKS_REQUIRED(cs_main);
 
-    //! Find the first block that is not pruned
+    //! Find the first stored ancestor of start_block immediately after the last
+    //! pruned ancestor. Return value will never be null. Caller is responsible
+    //! for ensuring that start_block has data is not pruned.
     const CBlockIndex* GetFirstStoredBlock(const CBlockIndex& start_block LIFETIMEBOUND) EXCLUSIVE_LOCKS_REQUIRED(::cs_main);
 
     /** True if any block files have ever been pruned. */

src/rpc/blockchain.cpp

@@ -812,9 +812,7 @@ static RPCHelpMan pruneblockchain()
 
     PruneBlockFilesManual(active_chainstate, height);
     const CBlockIndex& block{*CHECK_NONFATAL(active_chain.Tip())};
-    const CBlockIndex* last_block{active_chainstate.m_blockman.GetFirstStoredBlock(block)};
-
-    return static_cast<int64_t>(last_block->nHeight - 1);
+    return block.nStatus & BLOCK_HAVE_DATA ? active_chainstate.m_blockman.GetFirstStoredBlock(block)->nHeight - 1 : block.nHeight;
 },
     };
 }
@@ -1267,7 +1265,8 @@ RPCHelpMan getblockchaininfo()
     obj.pushKV("size_on_disk", chainman.m_blockman.CalculateCurrentUsage());
     obj.pushKV("pruned", chainman.m_blockman.IsPruneMode());
     if (chainman.m_blockman.IsPruneMode()) {
-        obj.pushKV("pruneheight", chainman.m_blockman.GetFirstStoredBlock(tip)->nHeight);
+        bool has_tip_data = tip.nStatus & BLOCK_HAVE_DATA;
+        obj.pushKV("pruneheight", has_tip_data ? chainman.m_blockman.GetFirstStoredBlock(tip)->nHeight : tip.nHeight + 1);
 
         const bool automatic_pruning{chainman.m_blockman.GetPruneTarget() != BlockManager::PRUNE_TARGET_MANUAL};
         obj.pushKV("automatic_pruning",  automatic_pruning);

src/test/blockmanager_tests.cpp

@@ -90,4 +90,35 @@ BOOST_FIXTURE_TEST_CASE(blockmanager_scan_unlink_already_pruned_files, TestChain
     BOOST_CHECK(!AutoFile(blockman.OpenBlockFile(new_pos, true)).IsNull());
 }
 
+BOOST_FIXTURE_TEST_CASE(blockmanager_block_data_availability, TestChain100Setup)
+{
+    LOCK(::cs_main);
+    auto& chainman = m_node.chainman;
+    auto& blockman = chainman->m_blockman;
+    const CBlockIndex& tip = *chainman->ActiveTip();
+
+    // Function to prune all blocks from 'last_pruned_block' down to the genesis block
+    const auto& func_prune_blocks = [&](CBlockIndex* last_pruned_block)
+    {
+        LOCK(::cs_main);
+        CBlockIndex* it = last_pruned_block;
+        while (it != nullptr && it->nStatus & BLOCK_HAVE_DATA) {
+            it->nStatus &= ~BLOCK_HAVE_DATA;
+            it = it->pprev;
+        }
+    };
+
+    // 1) Return genesis block when all blocks are available
+    BOOST_CHECK_EQUAL(blockman.GetFirstStoredBlock(tip), chainman->ActiveChain()[0]);
+
+    // Prune half of the blocks
+    int height_to_prune = tip.nHeight / 2;
+    CBlockIndex* first_available_block = chainman->ActiveChain()[height_to_prune + 1];
+    CBlockIndex* last_pruned_block = first_available_block->pprev;
+    func_prune_blocks(last_pruned_block);
+
+    // 2) The last block not pruned is in-between upper-block and the genesis block
+    BOOST_CHECK_EQUAL(blockman.GetFirstStoredBlock(tip), first_available_block);
+}
+
 BOOST_AUTO_TEST_SUITE_END()