fix: reenable macos builds (#3399)

* fix: reenable macos builds

Also, add debug function that prints local state if deadlocks occure.

* fix: free cold memory for non-cache mode as well

* chore: disable FreeMemWithEvictionStep again

Because it heavily affects the performance when performing evictions.

---------

Signed-off-by: Roman Gershman <roman@dragonflydb.io>

.github/workflows/daily-builds.yml

@@ -68,7 +68,6 @@ jobs:
             ctest -V -L DFLY
 
   build-macos:
-    if: false
     runs-on: macos-13
     timeout-minutes: 45
     steps:
@@ -94,8 +93,7 @@ jobs:
           rm /usr/local/bin/idle3* || :
           rm /usr/local/bin/pydoc* || :
           rm /usr/local/bin/python3* || :
-          brew update && brew install ninja boost openssl automake gcc zstd bison c-ares \
-              autoconf libtool automake
+          brew update && brew install ninja boost automake zstd bison
 
           mkdir -p $GITHUB_WORKSPACE/build
 
@@ -110,7 +108,7 @@ jobs:
 
           echo "*************************** START BUILDING **************************************"
           CC=gcc-12 CXX=g++-12 cmake .. -DCMAKE_BUILD_TYPE=Debug -GNinja -DWITH_UNWIND=OFF \
-            -DCMAKE_C_COMPILER_LAUNCHER=sccache
+            -DCMAKE_C_COMPILER_LAUNCHER=sccache -DCMAKE_CXX_FLAGS="-Wl,-ld_classic"
 
           ninja src/all
 

helio

@@ -1 +1 @@
-Subproject commit 288cb312971f07cd35fa3a75fcfe047788d7d0fb
+Subproject commit fb46b481e3eb82ecbc7bbf1d22b2fda7f5fac409

src/server/db_slice.cc

@@ -8,11 +8,11 @@
 
 #include "base/flags.h"
 #include "base/logging.h"
-#include "generic_family.h"
 #include "server/channel_store.h"
 #include "server/cluster/cluster_defs.h"
 #include "server/engine_shard_set.h"
 #include "server/error.h"
+#include "server/generic_family.h"
 #include "server/journal/journal.h"
 #include "server/server_state.h"
 #include "server/tiered_storage.h"
@@ -40,7 +40,7 @@ namespace dfly {
 using namespace std;
 using namespace util;
 using absl::GetFlag;
-using facade::OpStatus;
+using namespace facade;
 using Payload = journal::Entry::Payload;
 
 namespace {
@@ -574,6 +574,14 @@ OpResult<DbSlice::AddOrFindResult> DbSlice::AddOrFindInternal(const Context& cnt
   // It's a new entry.
   CallChangeCallbacks(cntx.db_index, key, {key});
 
+  // If we are low on memory due to cold storage, free some memory.
+  if (memory_budget_ < ssize_t(key.size()) && owner_->tiered_storage()) {
+    // At least 40KB bytes to cover potential segment split.
+    size_t goal = std::max<size_t>(key.size() * 2 - memory_budget_, 40_KB);
+    size_t reclaimed = owner_->tiered_storage()->ReclaimMemory(goal);
+    memory_budget_ += reclaimed;
+  }
+
   // In case we are loading from rdb file or replicating we want to disable conservative memory
   // checks (inside PrimeEvictionPolicy::CanGrow) and reject insertions only after we pass max
   // memory limit. When loading a snapshot created by the same server configuration (memory and
@@ -621,11 +629,15 @@ OpResult<DbSlice::AddOrFindResult> DbSlice::AddOrFindInternal(const Context& cnt
     // and we add new objects or update the existing ones and our memory usage grows.
     // We do not require for a single operation to unload the whole negative debt.
     // Instead, we create a positive, converging force that should help with freeing enough memory.
-    // Free at least 256 bytes or 3% of the total debt.
-    size_t evict_goal = std::max<size_t>(256, (-evp.mem_budget()) / 32);
+    // Free at least K bytes or 3% of the total debt.
+    // TODO: to reenable and optimize this - this call significantly slows down server
+    // when evictions are running.
+#if 0
+    size_t evict_goal = std::max<size_t>(512, (-evp.mem_budget()) / 32);
     auto [items, bytes] = FreeMemWithEvictionStep(cntx.db_index, it.segment_id(), evict_goal);
     evicted_obj_bytes = bytes;
     events_.hard_evictions += items;
+#endif
   }
 
   table_memory_ += (db.table_memory() - table_before);
@@ -1233,8 +1245,6 @@ int32_t DbSlice::GetNextSegmentForEviction(int32_t segment_id, DbIndex db_ind) c
 pair<uint64_t, size_t> DbSlice::FreeMemWithEvictionStep(DbIndex db_ind, size_t starting_segment_id,
                                                         size_t increase_goal_bytes) {
   DCHECK(!owner_->IsReplica());
-  if ((!caching_mode_) || !expire_allowed_)
-    return {0, 0};
 
   size_t evicted_items = 0, evicted_bytes = 0;
 
@@ -1244,6 +1254,9 @@ pair<uint64_t, size_t> DbSlice::FreeMemWithEvictionStep(DbIndex db_ind, size_t s
       return {0, evicted_bytes};
   }
 
+  if ((!caching_mode_) || !expire_allowed_)
+    return {0, 0};
+
   auto max_eviction_per_hb = GetFlag(FLAGS_max_eviction_per_heartbeat);
   auto max_segment_to_consider = GetFlag(FLAGS_max_segment_to_consider);
 

src/server/tiered_storage.h

@@ -133,7 +133,7 @@ class TieredStorage {
   // Min sizes of values taking up full page on their own
   const static size_t kMinOccupancySize = tiering::kPageSize / 2;
 
-  explicit TieredStorage(DbSlice* db_slice, size_t max_size) {
+  explicit TieredStorage(size_t max_size, DbSlice* db_slice) {
   }
 
   TieredStorage(TieredStorage&& other) = delete;
@@ -150,6 +150,10 @@ class TieredStorage {
     return {};
   }
 
+  void Read(DbIndex dbid, std::string_view key, const PrimeValue& value,
+            std::function<void(const std::string&)> readf) {
+  }
+
   template <typename T>
   util::fb2::Future<T> Modify(DbIndex dbid, std::string_view key, const PrimeValue& value,
                               std::function<T(std::string*)> modf) {
@@ -159,7 +163,19 @@ class TieredStorage {
   void TryStash(DbIndex dbid, std::string_view key, PrimeValue* value) {
   }
 
-  void Delete(PrimeValue* value) {
+  void Delete(DbIndex dbid, PrimeValue* value) {
+  }
+
+  size_t ReclaimMemory(size_t goal) {
+    return 0;
+  }
+
+  float WriteDepthUsage() const {
+    return 0;
+  }
+
+  size_t CoolMemoryUsage() const {
+    return 0;
   }
 
   void CancelStash(DbIndex dbid, std::string_view key, PrimeValue* value) {

src/server/tiered_storage_test.cc

@@ -310,7 +310,10 @@ TEST_F(TieredStorageTest, MemoryPressure) {
   constexpr size_t kNum = 10000;
   for (size_t i = 0; i < kNum; i++) {
     auto resp = Run({"SET", absl::StrCat("k", i), BuildString(10000)});
-    ASSERT_EQ(resp, "OK") << i;
+    if (resp != "OK"sv) {
+      resp = Run({"INFO", "ALL"});
+      ASSERT_FALSE(true) << i << "\nInfo ALL:\n" << resp.GetString();
+    }
     // TODO: to remove it once used_mem_current is updated frequently.
     ThisFiber::SleepFor(10us);
   }

src/server/transaction.cc

@@ -718,6 +718,13 @@ void Transaction::ScheduleInternal() {
       run_barrier_.Dec();
     } else {
       IterateActiveShards([cb](const auto& sd, ShardId i) { shard_set->Add(i, cb); });
+
+      // Add this debugging function to print more information when we experience deadlock
+      // during tests.
+      ThisFiber::PrintLocalsCallback locals([&] {
+        return absl::StrCat("unique_shard_cnt_: ", unique_shard_cnt_,
+                            " run_barrier_cnt: ", run_barrier_.DEBUG_Count(), "\n");
+      });
       run_barrier_.Wait();
     }