Route IPC through browser when creating a viewable command buffer.

chrome/browser/renderer_host/render_message_filter.cc

Index: chrome/browser/renderer_host/render_message_filter.cc
diff --git a/chrome/browser/renderer_host/render_message_filter.cc b/chrome/browser/renderer_host/render_message_filter.cc
index e00a6b8dd3730c37662c1980a594ed0fabdf65ab..338c64fca97f8a61149f131dbde7075e4a3fb992 100644
--- a/chrome/browser/renderer_host/render_message_filter.cc
+++ b/chrome/browser/renderer_host/render_message_filter.cc
@@ -48,6 +48,7 @@
 #include "chrome/common/chrome_switches.h"
 #include "chrome/common/extensions/extension_file_util.h"
 #include "chrome/common/extensions/extension_message_bundle.h"
+#include "chrome/common/gpu_create_command_buffer_config.h"
 #include "chrome/common/notification_service.h"
 #include "chrome/common/render_messages.h"
 #include "chrome/common/render_messages_params.h"
@@ -397,6 +398,8 @@ bool RenderMessageFilter::OnMessageReceived(const IPC::Message& message,
     IPC_MESSAGE_HANDLER(ViewHostMsg_EstablishGpuChannel, OnEstablishGpuChannel)
     IPC_MESSAGE_HANDLER_DELAY_REPLY(ViewHostMsg_SynchronizeGpu,
                                     OnSynchronizeGpu)
+    IPC_MESSAGE_HANDLER_DELAY_REPLY(ViewHostMsg_CreateViewCommandBuffer,
+                                    OnCreateViewCommandBuffer)
     IPC_MESSAGE_HANDLER(ViewHostMsg_AsyncOpenFile, OnAsyncOpenFile)
     IPC_MESSAGE_UNHANDLED(handled = false)
   IPC_END_MESSAGE_MAP_EX()
@@ -1423,6 +1426,14 @@ void RenderMessageFilter::OnSynchronizeGpu(IPC::Message* reply) {
   GpuProcessHost::Get()->Synchronize(reply, this);
 }
 
+void RenderMessageFilter::OnCreateViewCommandBuffer(

[jam, 2011/01/25 17:30:21]

There are 3 GPU related messages now being filtered, it may be time to create a
separate BrowserMessageFilter to filter them, so that RenderMessageFilter
doesn't have to know about GPU related code.  Do you expect more messages in the
future?

If you do this, the renderer<>browser GPU related messages can also move to
gpu_messages

[jonathan.backer, 2011/01/26 20:42:17]

I think that this is a good idea. We probably will want more such messages for
sandboxing the gpu process (e.g. have the creation of offscreen command buffers
go through the browser so that we can allocate SHM for them as well).

However, this patch it getting long already. Could I do that in a separate
patch?

+    int32 render_view_id,
+    const GPUCreateCommandBufferConfig& init_params,
+    IPC::Message* reply) {
+  GpuProcessHost::Get()->CreateViewCommandBuffer(
+      render_view_id, render_process_id_, init_params, reply, this);
+}
+
 void RenderMessageFilter::OnGetExtensionMessageBundle(
     const std::string& extension_id, IPC::Message* reply_msg) {
   ChromeURLRequestContext* context = static_cast<ChromeURLRequestContext*>(