Calling OpenGL from the renderer process.

chrome/browser/gpu_process_host.h

 // Copyright (c) 2010 The Chromium Authors. All rights reserved.
 // Use of this source code is governed by a BSD-style license that can be
 // found in the LICENSE file.
 
 #ifndef CHROME_BROWSER_GPU_PROCESS_HOST_H_
 #define CHROME_BROWSER_GPU_PROCESS_HOST_H_
 
 #include <queue>
 
 #include "base/basictypes.h"
 #include "base/scoped_ptr.h"
 #include "base/singleton.h"
 #include "chrome/browser/child_process_launcher.h"
 #include "chrome/common/gpu_native_window_handle.h"
 #include "chrome/common/message_router.h"
+#include "gfx/native_widget_types.h"
 #include "ipc/ipc_channel_handle.h"
 #include "ipc/ipc_channel_proxy.h"
 
 class ChildProcessLauncher;
 class CommandBufferProxy;
 
 class GpuProcessHost : public IPC::Channel::Sender,
                        public IPC::Channel::Listener,
                        public ChildProcessLauncher::Client {
  public:
@@ skipping 17 matching lines @@
   // ChildProcessLauncher::Client implementation.
   virtual void OnProcessLaunched();
 
   // See documentation on MessageRouter for AddRoute and RemoveRoute
   void AddRoute(int32 routing_id, IPC::Channel::Listener* listener);
   void RemoveRoute(int32 routing_id);
 
   // Tells the GPU process to create a new channel for communication with a
   // renderer. Will asynchronously send message to object with given routing id
   // on completion.
-  void EstablishGpuChannel(int renderer_id, int routing_id);
+  void EstablishGpuChannel(int renderer_id);
 
  private:
   friend struct DefaultSingletonTraits<GpuProcessHost>;
 
   // Used to queue pending channel requests.
   struct ChannelRequest {
-    ChannelRequest(int renderer_id,
-                   int routing_id) :
-        renderer_id(renderer_id),
-        routing_id(routing_id) {}
+    explicit ChannelRequest(int renderer_id) : renderer_id(renderer_id) {}
     // Used to identify the renderer. The ID is used instead of a pointer to
     // the RenderProcessHost in case it is destroyed while the request is
     // pending.
     // TODO(apatrick): investigate whether these IDs are used for future
     // render processes.
     int renderer_id;
-
-    // Routing ID of object to receive reply message.
-    int routing_id;
   };
 
   GpuProcessHost();
   virtual ~GpuProcessHost();
 
   void OnControlMessageReceived(const IPC::Message& message);
 
   // Message handlers.
   void OnChannelEstablished(const IPC::ChannelHandle& channel_handle);
 
   void ReplyToRenderer(int renderer_id,
-                       int routing_id,
                        const IPC::ChannelHandle& channel);
 
   // These are the channel requests that we have already sent to
   // the GPU process, but haven't heard back about yet.
   std::queue<ChannelRequest> sent_requests_;
 
   // Copies applicable command line switches from the given |browser_cmd| line
   // flags to the output |gpu_cmd| line flags. Not all switches will be
   // copied over.
   void PropagateBrowserCommandLineToGpu(const CommandLine& browser_cmd,
@@ skipping 12 matching lines @@
   // Messages we queue while waiting for the process handle.  We queue them here
   // instead of in the channel so that we ensure they're sent after init related
   // messages that are sent once the process handle is available.  This is
   // because the queued messages may have dependencies on the init messages.
   std::queue<IPC::Message*> queued_messages_;
 
   DISALLOW_COPY_AND_ASSIGN(GpuProcessHost);
 };
 
 #endif  // CHROME_BROWSER_GPU_PROCESS_HOST_H_