fix(test): correct test logic and semantics in multiple test cases

1. ChannelTest::MultipleSendersReceivers
   - Add C++14-compatible latch implementation (similar to C++20 std::latch)
   - Ensure receivers are ready before senders start sending messages
   - This prevents race condition where senders might send before receivers are listening

2. RWLockTest::ReadWriteReadPattern
   - Fix test logic: lock_shared allows multiple concurrent readers
   - Previous test had race condition where both threads could read same value
   - New test: each thread writes based on thread id (1 or 2), then reads
   - Expected result: 1*20 + 2*20 = 60

3. ShmTest::ReleaseMemory
   - Correct return value semantics: release() returns ref count before decrement, or -1 on error
   - Must call get_mem() to map memory and set ref count to 1 before release
   - Expected: release() returns 1 (ref count before decrement)

4. ShmTest::ReferenceCount
   - Correct semantics: ref count is 0 after acquire (memory not mapped)
   - get_mem() maps memory and sets ref count to 1
   - Second acquire+get_mem increases ref count to 2
   - Test now properly validates reference counting behavior

5. ShmTest::SubtractReference
   - sub_ref() only works after get_mem() has mapped the memory
   - Must call get_mem() first to initialize ref count to 1
   - sub_ref() then decrements it to 0
   - Test now follows correct API usage pattern

test/test_ipc_channel.cpp

@@ -22,6 +22,8 @@
 #include <vector>
 #include <string>
 #include <cstring>
+#include <mutex>
+#include <condition_variable>
 #include "libipc/ipc.h"
 #include "libipc/buffer.h"
 
@@ -29,6 +31,29 @@ using namespace ipc;
 
 namespace {
 
+// Simple latch implementation for C++14 (similar to C++20 std::latch)
+class latch {
+public:
+  explicit latch(std::ptrdiff_t count) : count_(count) {}
+  
+  void count_down() {
+      std::unique_lock<std::mutex> lock(mutex_);
+      if (--count_ <= 0) {
+          cv_.notify_all();
+      }
+  }
+  
+  void wait() {
+      std::unique_lock<std::mutex> lock(mutex_);
+      cv_.wait(lock, [this] { return count_ <= 0; });
+  }
+
+private:
+  std::ptrdiff_t count_;
+  std::mutex mutex_;
+  std::condition_variable cv_;
+};
+
 std::string generate_unique_ipc_name(const char* prefix) {
   static int counter = 0;
   return std::string(prefix) + "_ipc_" + std::to_string(++counter);
@@ -516,6 +541,27 @@ TEST_F(ChannelTest, MultipleSendersReceivers) {
   std::atomic<int> sent_count{0};
   std::atomic<int> received_count{0};
   
+  // Use latch to ensure receivers are ready before senders start
+  latch receivers_ready(num_receivers);
+  
+  std::vector<std::thread> receivers;
+  for (int i = 0; i < num_receivers; ++i) {
+      receivers.emplace_back([&, i]() {
+          channel ch(name.c_str(), receiver);
+          receivers_ready.count_down();  // Signal this receiver is ready
+          
+          for (int j = 0; j < messages_per_sender; ++j) {
+              buffer buf = ch.recv(2000);
+              if (!buf.empty()) {
+                  ++received_count;
+              }
+          }
+      });
+  }
+  
+  // Wait for all receivers to be ready
+  receivers_ready.wait();
+  
   std::vector<std::thread> senders;
   for (int i = 0; i < num_senders; ++i) {
       senders.emplace_back([&, i]() {
@@ -530,19 +576,6 @@ TEST_F(ChannelTest, MultipleSendersReceivers) {
       });
   }
   
-  std::vector<std::thread> receivers;
-  for (int i = 0; i < num_receivers; ++i) {
-      receivers.emplace_back([&, i]() {
-          channel ch(name.c_str(), receiver);
-          for (int j = 0; j < messages_per_sender; ++j) {
-              buffer buf = ch.recv(2000);
-              if (!buf.empty()) {
-                  ++received_count;
-              }
-          }
-      });
-  }
-  
   for (auto& t : senders) {
       t.join();
   }

test/test_locks.cpp

@@ -383,20 +383,23 @@ TEST_F(RWLockTest, ReadersWritersNoOverlap) {
 TEST_F(RWLockTest, ReadWriteReadPattern) {
   rw_lock lock;
   int data = 0;
+  std::atomic<int> iterations{0};
   
   auto pattern_task = [&](int id) {
       for (int i = 0; i < 20; ++i) {
-          // Read
-          lock.lock_shared();
-          int read_val = data;
-          lock.unlock_shared();
+          // Write: increment based on thread id
+          lock.lock();
+          data += id;
+          lock.unlock();
           
+          iterations.fetch_add(1);
           std::this_thread::yield();
           
-          // Write
-          lock.lock();
-          data = read_val + 1;
-          lock.unlock();
+          // Read: verify data is consistent
+          lock.lock_shared();
+          int read_val = data;
+          EXPECT_GE(read_val, 0);  // Data should be non-negative
+          lock.unlock_shared();
           
           std::this_thread::yield();
       }
@@ -408,7 +411,10 @@ TEST_F(RWLockTest, ReadWriteReadPattern) {
   t1.join();
   t2.join();
   
-  EXPECT_EQ(data, 40);
+  // Each thread increments by its id (1 or 2), 20 times each
+  // Total = 1*20 + 2*20 = 20 + 40 = 60
+  EXPECT_EQ(data, 60);
+  EXPECT_EQ(iterations.load(), 40);
 }
 
 // Test many readers, one writer

test/test_shm.cpp

@@ -126,8 +126,13 @@ TEST_F(ShmTest, ReleaseMemory) {
   shm::id_t id = shm::acquire(name.c_str(), 128, shm::create);
   ASSERT_NE(id, nullptr);
   
+  // Must call get_mem to map memory and set reference count
+  void* mem = shm::get_mem(id, nullptr);
+  ASSERT_NE(mem, nullptr);
+  
+  // release returns the reference count before decrement, or -1 on error
   std::int32_t ref_count = shm::release(id);
-  EXPECT_GE(ref_count, 0);
+  EXPECT_EQ(ref_count, 1);  // Should be 1 (set by get_mem, before decrement)
   
   shm::remove(name.c_str());
 }
@@ -161,14 +166,25 @@ TEST_F(ShmTest, ReferenceCount) {
   shm::id_t id1 = shm::acquire(name.c_str(), 512, shm::create);
   ASSERT_NE(id1, nullptr);
   
-  std::int32_t ref1 = shm::get_ref(id1);
-  EXPECT_GT(ref1, 0);
+  // Reference count is 0 after acquire (memory not mapped yet)
+  std::int32_t ref_before_get_mem = shm::get_ref(id1);
+  EXPECT_EQ(ref_before_get_mem, 0);
   
-  // Acquire again (should increase reference count)
+  // get_mem maps memory and sets reference count to 1
+  void* mem1 = shm::get_mem(id1, nullptr);
+  ASSERT_NE(mem1, nullptr);
+  
+  std::int32_t ref1 = shm::get_ref(id1);
+  EXPECT_EQ(ref1, 1);
+  
+  // Acquire again and get_mem (should increase reference count)
   shm::id_t id2 = shm::acquire(name.c_str(), 512, shm::open);
   if (id2 != nullptr) {
+      void* mem2 = shm::get_mem(id2, nullptr);
+      ASSERT_NE(mem2, nullptr);
+      
       std::int32_t ref2 = shm::get_ref(id2);
-      EXPECT_GE(ref2, ref1);
+      EXPECT_EQ(ref2, 2);  // Should be 2 now
       
       shm::release(id2);
   }
@@ -184,11 +200,17 @@ TEST_F(ShmTest, SubtractReference) {
   shm::id_t id = shm::acquire(name.c_str(), 256, shm::create);
   ASSERT_NE(id, nullptr);
   
-  std::int32_t ref_before = shm::get_ref(id);
-  shm::sub_ref(id);
-  std::int32_t ref_after = shm::get_ref(id);
+  // Must call get_mem first to map memory and initialize reference count
+  void* mem = shm::get_mem(id, nullptr);
+  ASSERT_NE(mem, nullptr);
   
-  EXPECT_EQ(ref_after, ref_before - 1);
+  std::int32_t ref_before = shm::get_ref(id);
+  EXPECT_EQ(ref_before, 1);  // Should be 1 after get_mem
+  
+  shm::sub_ref(id);
+  
+  std::int32_t ref_after = shm::get_ref(id);
+  EXPECT_EQ(ref_after, 0);  // Should be 0 after sub_ref
   
   // Use remove(id) to clean up - it internally calls release()
   shm::remove(id);