[dxvk] Create state cache threads on demand

src/dxvk/dxvk_pipemanager.cpp

@@ -5,7 +5,7 @@
 namespace dxvk {
   
   DxvkPipelineManager::DxvkPipelineManager(
-    const DxvkDevice*         device,
+          DxvkDevice*         device,
           DxvkRenderPassPool* passManager)
   : m_device    (device),
     m_cache     (new DxvkPipelineCache(device->vkd())) {

src/dxvk/dxvk_pipemanager.h

@@ -38,7 +38,7 @@ namespace dxvk {
   public:
     
     DxvkPipelineManager(
-      const DxvkDevice*         device,
+            DxvkDevice*         device,
             DxvkRenderPassPool* passManager);
     
     ~DxvkPipelineManager();
@@ -97,7 +97,7 @@ namespace dxvk {
     
   private:
     
-    const DxvkDevice*         m_device;
+    DxvkDevice*               m_device;
     Rc<DxvkPipelineCache>     m_cache;
     Rc<DxvkStateCache>        m_stateCache;
 

src/dxvk/dxvk_state_cache.cpp

@@ -150,11 +150,12 @@ namespace dxvk {
 
 
   DxvkStateCache::DxvkStateCache(
-    const DxvkDevice*           device,
+          DxvkDevice*           device,
           DxvkPipelineManager*  pipeManager,
           DxvkRenderPassPool*   passManager)
-  : m_pipeManager(pipeManager),
-    m_passManager(passManager) {
+  : m_device      (device),
+    m_pipeManager (pipeManager),
+    m_passManager (passManager) {
     bool newFile = !readCacheFile();
 
     if (newFile) {
@@ -184,28 +185,6 @@ namespace dxvk {
       for (auto& e : m_entries)
         writeCacheEntry(file, e);
     }
-
-    // Use half the available CPU cores for pipeline compilation
-    uint32_t numCpuCores = dxvk::thread::hardware_concurrency();
-    uint32_t numWorkers  = ((std::max(1u, numCpuCores) - 1) * 5) / 7;
-
-    if (numWorkers <  1) numWorkers =  1;
-    if (numWorkers > 32) numWorkers = 32;
-
-    if (device->config().numCompilerThreads > 0)
-      numWorkers = device->config().numCompilerThreads;
-    
-    Logger::info(str::format("DXVK: Using ", numWorkers, " compiler threads"));
-    
-    // Start the worker threads and the file writer
-    m_workerBusy.store(numWorkers);
-
-    for (uint32_t i = 0; i < numWorkers; i++) {
-      m_workerThreads.emplace_back([this] () { workerFunc(); });
-      m_workerThreads[i].set_priority(ThreadPriority::Lowest);
-    }
-    
-    m_writerThread = dxvk::thread([this] () { writerFunc(); });
   }
   
 
@@ -238,6 +217,8 @@ namespace dxvk {
       DxvkComputePipelineStateInfo(),
       format, g_nullHash });
     m_writerCond.notify_one();
+
+    createWriter();
   }
 
 
@@ -262,6 +243,8 @@ namespace dxvk {
       DxvkGraphicsPipelineStateInfo(), state,
       DxvkRenderPassFormat(), g_nullHash });
     m_writerCond.notify_one();
+
+    createWriter();
   }
 
 
@@ -297,8 +280,10 @@ namespace dxvk {
       m_workerQueue.push(item);
     }
 
-    if (workerLock)
+    if (workerLock) {
       m_workerCond.notify_all();
+      createWorkers();
+    }
   }
 
 
@@ -316,7 +301,8 @@ namespace dxvk {
     for (auto& worker : m_workerThreads)
       worker.join();
     
-    m_writerThread.join();
+    if (m_writerThread.joinable())
+      m_writerThread.join();
   }
 
 
@@ -987,6 +973,37 @@ namespace dxvk {
   }
 
 
+  void DxvkStateCache::createWorkers() {
+    if (m_workerThreads.empty()) {
+      // Use half the available CPU cores for pipeline compilation
+      uint32_t numCpuCores = dxvk::thread::hardware_concurrency();
+      uint32_t numWorkers  = ((std::max(1u, numCpuCores) - 1) * 5) / 7;
+
+      if (numWorkers <  1) numWorkers =  1;
+      if (numWorkers > 32) numWorkers = 32;
+
+      if (m_device->config().numCompilerThreads > 0)
+        numWorkers = m_device->config().numCompilerThreads;
+
+      Logger::info(str::format("DXVK: Using ", numWorkers, " compiler threads"));
+
+      // Start the worker threads and the file writer
+      m_workerBusy.store(numWorkers);
+
+      for (uint32_t i = 0; i < numWorkers; i++) {
+        m_workerThreads.emplace_back([this] () { workerFunc(); });
+        m_workerThreads[i].set_priority(ThreadPriority::Lowest);
+      }
+    }
+  }
+
+
+  void DxvkStateCache::createWriter() {
+    if (!m_writerThread.joinable())
+      m_writerThread = dxvk::thread([this] () { writerFunc(); });
+  }
+
+
   std::wstring DxvkStateCache::getCacheFileName() const {
     std::string path = getCacheDir();
 

src/dxvk/dxvk_state_cache.h

@@ -28,7 +28,7 @@ namespace dxvk {
   public:
 
     DxvkStateCache(
-      const DxvkDevice*           device,
+            DxvkDevice*           device,
             DxvkPipelineManager*  pipeManager,
             DxvkRenderPassPool*   passManager);
     
@@ -93,6 +93,7 @@ namespace dxvk {
       DxvkComputePipelineShaders  cp;
     };
 
+    DxvkDevice*                       m_device;
     DxvkPipelineManager*              m_pipeManager;
     DxvkRenderPassPool*               m_passManager;
 
@@ -181,6 +182,10 @@ namespace dxvk {
 
     void writerFunc();
 
+    void createWorkers();
+
+    void createWriter();
+
     std::wstring getCacheFileName() const;
     
     std::string getCacheDir() const;