From bc2c969fe333bfe4551f24439f82f8abfabd243e Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?=E5=B0=8F=E4=BA=BA=E9=B1=BC?= <1823748191@qq.com>
Date: Tue, 10 Dec 2024 20:46:04 +0800
Subject: [PATCH] fix fast remove bug for GC

---
 db/db_impl.cc     | 19 +++++++++++--------
 db/version_set.cc |  8 ++++++++
 db/version_set.h  |  5 +++++
 3 files changed, 24 insertions(+), 8 deletions(-)

diff --git a/db/db_impl.cc b/db/db_impl.cc
index c8e5ea2..27a4019 100644
--- a/db/db_impl.cc
+++ b/db/db_impl.cc
@@ -1742,14 +1742,14 @@ void DBImpl::GarbageCollect() {
       valuelog_set.emplace(filename);
     }
   }
-  bool tmp_judge=false;//only clean one file
+  //bool tmp_judge=false;//only clean one file
   for (std::string valuelog_name : valuelog_set) {
-    if(tmp_judge){
-      break;
-    }
-    else{
-      tmp_judge=true;
-    }
+    // if(tmp_judge){
+    //   break;
+    // }
+    // else{
+    //   tmp_judge=true;
+    // }
     uint64_t cur_log_number = GetValueLogID(valuelog_name);
     valuelog_name = ValueLogFileName(dbname_, cur_log_number);
     if (cur_log_number == valuelogfile_number_) {
@@ -1936,7 +1936,10 @@ void DBImpl::GarbageCollect() {
     // 清理旧文件（如果需要）
     cur_valuelog.close();
 
-    env_->RemoveFile(valuelog_name);
+    mutex_.Lock();
+    versions_->current()->addOldValueLog(valuelog_name);
+
+    mutex_.Unlock();
 
 
     Log(options_.info_log, "remove file during gc %s", valuelog_name.c_str());
diff --git a/db/version_set.cc b/db/version_set.cc
index dcce922..b2df47d 100644
--- a/db/version_set.cc
+++ b/db/version_set.cc
@@ -82,6 +82,10 @@ Version::~Version() {
       }
     }
   }
+
+  for(auto valuelog_name:old_valuelog_names){
+    vset_->env_->RemoveFile(valuelog_name);
+  }
 }
 
 int FindFile(const InternalKeyComparator& icmp,
@@ -461,6 +465,10 @@ void Version::Unref() {
   }
 }
 
+void Version::addOldValueLog(std::string valuelog_name){
+  old_valuelog_names.push_back(valuelog_name);
+}
+
 bool Version::OverlapInLevel(int level, const Slice* smallest_user_key,
                              const Slice* largest_user_key) {
   return SomeFileOverlapsRange(vset_->icmp_, (level > 0), files_[level],
diff --git a/db/version_set.h b/db/version_set.h
index ea0c925..0568b6c 100644
--- a/db/version_set.h
+++ b/db/version_set.h
@@ -90,6 +90,9 @@ class Version {
   // under live iterators)
   void Ref();
   void Unref();
+  
+  // Add old valuelog to current version, delete them when current version is no more used
+  void addOldValueLog(std::string valuelog_name);
 
   void GetOverlappingInputs(
       int level,
@@ -153,6 +156,8 @@ class Version {
   // List of files per level
   std::vector<FileMetaData*> files_[config::kNumLevels];
 
+  std::vector<std::string> old_valuelog_names;
+
   // Next file to compact based on seek stats.
   FileMetaData* file_to_compact_;
   int file_to_compact_level_;