cidstore: implemented validation of data during index initialization

1 files changed, 24 insertions, 9 deletions

diff --git a/zenstore/cidstore.cpp b/zenstore/cidstore.cpp
index 7a5d7bcf4..eaad0f5a8 100644
--- a/zenstore/cidstore.cpp
+++ b/zenstore/cidstore.cpp
@@ -4,7 +4,9 @@
 
 #include <zencore/compress.h>
 #include <zencore/filesystem.h>
+#include <zencore/fmtutils.h>
 #include <zencore/logging.h>
+#include <zencore/string.h>
 #include <zenstore/CAS.h>
 #include <zenstore/caslog.h>
 
@@ -65,9 +67,10 @@ struct CidStore::Impl
 		LogMapping(DecompressedId, Compressed);
 	}
 
-	void LogMapping(const IoHash& DecompressedId, const IoHash& Compressed)
+	void LogMapping(const IoHash& DecompressedId, const IoHash& CompressedHash)
 	{
-		m_LogFile.Append({.Uncompressed = DecompressedId, .Compressed = Compressed});
+		ZEN_ASSERT(DecompressedId != CompressedHash);
+		m_LogFile.Append({.Uncompressed = DecompressedId, .Compressed = CompressedHash});
 	}
 
 	IoBuffer FindChunkByCid(const IoHash& DecompressedId)
@@ -118,23 +121,35 @@ struct CidStore::Impl
 
 		m_LogFile.Open(SlogPath, IsNew);
 
+		ZEN_DEBUG("Initializing index from '{}' ({})", SlogPath, NiceBytes(m_LogFile.GetLogSize()));
+
 		uint64_t TombstoneCount = 0;
+		uint64_t InvalidCount	= 0;
 
-		m_LogFile.Replay([&](const IndexEntry& Ie) {
-			if (Ie.Compressed != IoHash::Zero)
+		m_LogFile.Replay([&](const IndexEntry& Entry) {
+			if (Entry.Compressed != IoHash::Zero)
 			{
 				// Update
-				m_CidMap.insert_or_assign(Ie.Uncompressed, Ie.Compressed);
+				m_CidMap.insert_or_assign(Entry.Uncompressed, Entry.Compressed);
 			}
 			else
 			{
-				// Tombstone
-				m_CidMap.erase(Ie.Uncompressed);
-				++TombstoneCount;
+				if (Entry.Uncompressed != IoHash::Zero)
+				{
+					// Tombstone
+					m_CidMap.erase(Entry.Uncompressed);
+					++TombstoneCount;
+				}
+				else
+				{
+					// Completely uninitialized entry with both hashes set to zero indicates a
+					// problem. Might be an unwritten page due to BSOD or some other problem
+					++InvalidCount;
+				}
 			}
 		});
 
-		ZEN_INFO("CID index initialized: {} entries found ({} tombstones)", m_CidMap.size(), TombstoneCount);
+		ZEN_INFO("CID index initialized: {} entries found ({} tombstones, {} invalid)", m_CidMap.size(), TombstoneCount, InvalidCount);
 	}
 
 	void Flush() { m_LogFile.Flush(); }