Completed the implementation for PPB_Graphics3D interface. Mostly copied from the implementations...

ppapi/proxy/ppb_graphics_3d_proxy.cc

Index: ppapi/proxy/ppb_graphics_3d_proxy.cc
===================================================================
--- ppapi/proxy/ppb_graphics_3d_proxy.cc	(revision 0)
+++ ppapi/proxy/ppb_graphics_3d_proxy.cc	(revision 0)
@@ -0,0 +1,612 @@
+// Copyright (c) 2011 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.
+
+#include "ppapi/proxy/ppb_graphics_3d_proxy.h"
+
+#include "gpu/command_buffer/client/gles2_implementation.h"
+#include "ppapi/c/pp_errors.h"
+#include "ppapi/proxy/enter_proxy.h"
+#include "ppapi/proxy/plugin_dispatcher.h"
+#include "ppapi/proxy/ppapi_messages.h"
+#include "ppapi/thunk/enter.h"
+#include "ppapi/thunk/resource_creation_api.h"
+#include "ppapi/thunk/thunk.h"
+
+using ppapi::thunk::EnterFunctionNoLock;
+using ppapi::thunk::EnterResourceNoLock;
+using ppapi::thunk::PPB_Graphics3D_API;
+using ppapi::thunk::ResourceCreationAPI;
+
+namespace pp {
+namespace proxy {
+
+namespace {
+const int32 kCommandBufferSize = 1024 * 1024;
+const int32 kTransferBufferSize = 1024 * 1024;
+
+class CommandBuffer : public gpu::CommandBuffer {
+ public:
+  CommandBuffer(const HostResource& resource, PluginDispatcher* dispatcher);
+  virtual ~CommandBuffer();
+
+  // gpu::CommandBuffer implementation:
+  virtual bool Initialize(int32 size);
+  virtual bool Initialize(base::SharedMemory* buffer, int32 size);
+  virtual gpu::Buffer GetRingBuffer();
+  virtual State GetState();
+  virtual void Flush(int32 put_offset);
+  virtual State FlushSync(int32 put_offset, int32 last_known_get);
+  virtual void SetGetOffset(int32 get_offset);
+  virtual int32 CreateTransferBuffer(size_t size, int32 id_request);
+  virtual int32 RegisterTransferBuffer(base::SharedMemory* shared_memory,
+                                       size_t size,
+                                       int32 id_request);
+  virtual void DestroyTransferBuffer(int32 id);
+  virtual gpu::Buffer GetTransferBuffer(int32 handle);
+  virtual void SetToken(int32 token);
+  virtual void SetParseError(gpu::error::Error error);
+  virtual void SetContextLostReason(gpu::error::ContextLostReason reason);
+
+ private:
+  bool Send(IPC::Message* msg);
+  void UpdateState(const gpu::CommandBuffer::State& state);
+
+  int32 num_entries_;
+  scoped_ptr<base::SharedMemory> ring_buffer_;
+
+  typedef base::hash_map<int32, gpu::Buffer> TransferBufferMap;
+  TransferBufferMap transfer_buffers_;
+
+  State last_state_;
+
+  HostResource resource_;
+  PluginDispatcher* dispatcher_;
+
+  DISALLOW_COPY_AND_ASSIGN(CommandBuffer);
+};
+
+CommandBuffer::CommandBuffer(const HostResource& resource,
+                             PluginDispatcher* dispatcher)
+    : num_entries_(0),
+      resource_(resource),
+      dispatcher_(dispatcher) {
+}
+
+CommandBuffer::~CommandBuffer() {
+  // Delete all the locally cached shared memory objects, closing the handle
+  // in this process.
+  for (TransferBufferMap::iterator it = transfer_buffers_.begin();
+       it != transfer_buffers_.end(); ++it) {
+    delete it->second.shared_memory;
+    it->second.shared_memory = NULL;
+  }
+}
+
+bool CommandBuffer::Initialize(int32 size) {
+  DCHECK(!ring_buffer_.get());
+
+  // Initialize the service. Assuming we are sandboxed, the GPU
+  // process is responsible for duplicating the handle. This might not be true
+  // for NaCl.
+  base::SharedMemoryHandle handle;
+  if (Send(new PpapiHostMsg_PPBGraphics3D_InitCommandBuffer(
+          INTERFACE_ID_PPB_GRAPHICS_3D, resource_, size, &handle)) &&
+      base::SharedMemory::IsHandleValid(handle)) {
+    ring_buffer_.reset(new base::SharedMemory(handle, false));
+    if (ring_buffer_->Map(size)) {
+      num_entries_ = size / sizeof(gpu::CommandBufferEntry);
+      return true;
+    }
+
+    ring_buffer_.reset();
+  }
+
+  return false;
+}
+
+bool CommandBuffer::Initialize(base::SharedMemory* buffer, int32 size) {
+  // Not implemented in proxy.
+  NOTREACHED();
+  return false;
+}
+
+gpu::Buffer CommandBuffer::GetRingBuffer() {
+  // Return locally cached ring buffer.
+  gpu::Buffer buffer;
+  buffer.ptr = ring_buffer_->memory();
+  buffer.size = num_entries_ * sizeof(gpu::CommandBufferEntry);
+  buffer.shared_memory = ring_buffer_.get();
+  return buffer;
+}
+
+gpu::CommandBuffer::State CommandBuffer::GetState() {
+  // Send will flag state with lost context if IPC fails.
+  if (last_state_.error == gpu::error::kNoError) {
+    gpu::CommandBuffer::State state;
+    if (Send(new PpapiHostMsg_PPBGraphics3D_GetState(
+             INTERFACE_ID_PPB_GRAPHICS_3D, resource_, &state)))
+      UpdateState(state);
+  }
+
+  return last_state_;
+}
+
+void CommandBuffer::Flush(int32 put_offset) {
+  if (last_state_.error != gpu::error::kNoError)
+    return;
+
+  IPC::Message* message = new PpapiHostMsg_PPBGraphics3D_AsyncFlush(
+      INTERFACE_ID_PPB_GRAPHICS_3D, resource_, put_offset);
+
+  // Do not let a synchronous flush hold up this message. If this handler is
+  // deferred until after the synchronous flush completes, it will overwrite the
+  // cached last_state_ with out-of-date data.
+  message->set_unblock(true);
+  Send(message);
+}
+
+gpu::CommandBuffer::State CommandBuffer::FlushSync(int32 put_offset,
+                                                   int32 last_known_get) {
+  if (last_known_get == last_state_.get_offset) {
+    // Send will flag state with lost context if IPC fails.
+    if (last_state_.error == gpu::error::kNoError) {
+      gpu::CommandBuffer::State state;
+      if (Send(new PpapiHostMsg_PPBGraphics3D_Flush(
+              INTERFACE_ID_PPB_GRAPHICS_3D, resource_, put_offset,
+              last_known_get, &state)))
+        UpdateState(state);
+    }
+  } else {
+    Flush(put_offset);
+  }
+
+  return last_state_;
+}
+
+void CommandBuffer::SetGetOffset(int32 get_offset) {
+  // Not implemented in proxy.
+  NOTREACHED();
+}
+
+int32 CommandBuffer::CreateTransferBuffer(size_t size, int32 id_request) {
+  if (last_state_.error == gpu::error::kNoError) {
+    int32 id;
+    if (Send(new PpapiHostMsg_PPBGraphics3D_CreateTransferBuffer(
+            INTERFACE_ID_PPB_GRAPHICS_3D, resource_, size, &id))) {
+      return id;
+    }
+  }
+
+  return -1;
+}
+
+int32 CommandBuffer::RegisterTransferBuffer(
+    base::SharedMemory* shared_memory,
+    size_t size,
+    int32 id_request) {
+  // Not implemented in proxy.
+  NOTREACHED();
+  return -1;
+}
+
+void CommandBuffer::DestroyTransferBuffer(int32 id) {
+  if (last_state_.error != gpu::error::kNoError)
+    return;
+
+  // Remove the transfer buffer from the client side4 cache.
+  TransferBufferMap::iterator it = transfer_buffers_.find(id);
+  DCHECK(it != transfer_buffers_.end());
+
+  // Delete the shared memory object, closing the handle in this process.
+  delete it->second.shared_memory;
+
+  transfer_buffers_.erase(it);
+
+  Send(new PpapiHostMsg_PPBGraphics3D_DestroyTransferBuffer(
+      INTERFACE_ID_PPB_GRAPHICS_3D, resource_, id));
+}
+
+gpu::Buffer CommandBuffer::GetTransferBuffer(int32 id) {
+  if (last_state_.error != gpu::error::kNoError)
+    return gpu::Buffer();
+
+  // Check local cache to see if there is already a client side shared memory
+  // object for this id.
+  TransferBufferMap::iterator it = transfer_buffers_.find(id);
+  if (it != transfer_buffers_.end()) {
+    return it->second;
+  }
+
+  // Assuming we are in the renderer process, the service is responsible for
+  // duplicating the handle. This might not be true for NaCl.
+  base::SharedMemoryHandle handle;
+  uint32 size;
+  if (!Send(new PpapiHostMsg_PPBGraphics3D_GetTransferBuffer(
+          INTERFACE_ID_PPB_GRAPHICS_3D, resource_, id, &handle, &size))) {
+    return gpu::Buffer();
+  }
+
+  // Cache the transfer buffer shared memory object client side.
+  scoped_ptr<base::SharedMemory> shared_memory(
+      new base::SharedMemory(handle, false));
+
+  // Map the shared memory on demand.
+  if (!shared_memory->memory()) {
+    if (!shared_memory->Map(size)) {
+      return gpu::Buffer();
+    }
+  }
+
+  gpu::Buffer buffer;
+  buffer.ptr = shared_memory->memory();
+  buffer.size = size;
+  buffer.shared_memory = shared_memory.release();
+  transfer_buffers_[id] = buffer;
+
+  return buffer;
+}
+
+void CommandBuffer::SetToken(int32 token) {
+  NOTREACHED();
+}
+
+void CommandBuffer::SetParseError(gpu::error::Error error) {
+  NOTREACHED();
+}
+
+void CommandBuffer::SetContextLostReason(
+    gpu::error::ContextLostReason reason) {
+  NOTREACHED();
+}
+
+bool CommandBuffer::Send(IPC::Message* msg) {
+  DCHECK(last_state_.error == gpu::error::kNoError);
+
+  if (dispatcher_->Send(msg))
+    return true;
+
+  last_state_.error = gpu::error::kLostContext;
+  return false;
+}
+
+void CommandBuffer::UpdateState(const gpu::CommandBuffer::State& state) {
+  // Handle wraparound. It works as long as we don't have more than 2B state
+  // updates in flight across which reordering occurs.
+  if (state.generation - last_state_.generation < 0x80000000U)
+    last_state_ = state;
+}
+
+base::SharedMemoryHandle TransportSHMHandleFromInt(Dispatcher* dispatcher,
+                                                   int shm_handle) {
+  // TODO(piman): Change trusted interface to return a PP_FileHandle, those
+  // casts are ugly.
+  base::PlatformFile source =
+#if defined(OS_WIN)
+      reinterpret_cast<HANDLE>(static_cast<intptr_t>(shm_handle));
+#elif defined(OS_POSIX)
+      shm_handle;
+#else
+  #error Not implemented.
+#endif
+  // Don't close the handle, it doesn't belong to us.
+  return dispatcher->ShareHandleWithRemote(source, false);
+}
+
+PP_Graphics3DTrustedState GetErrorState() {
+  PP_Graphics3DTrustedState error_state = { 0 };
+  error_state.error = PPB_GRAPHICS3D_TRUSTED_ERROR_GENERICERROR;
+  return error_state;
+}
+
+gpu::CommandBuffer::State GPUStateFromPPState(
+    const PP_Graphics3DTrustedState& s) {
+  gpu::CommandBuffer::State state;
+  state.num_entries = s.num_entries;
+  state.get_offset = s.get_offset;
+  state.put_offset = s.put_offset;
+  state.token = s.token;
+  state.error = static_cast<gpu::error::Error>(s.error);
+  state.generation = s.generation;
+  return state;
+}
+
+InterfaceProxy* CreateGraphics3DProxy(Dispatcher* dispatcher,
+                                      const void* target_interface) {
+  return new PPB_Graphics3D_Proxy(dispatcher, target_interface);
+}
+}  // namespace
+
+Graphics3D::Graphics3D(const HostResource& resource)
+    : PluginResource(resource) {
+}
+
+Graphics3D::~Graphics3D() {
+  DestroyGLES2Impl();
+}
+
+bool Graphics3D::Init() {
+  PluginDispatcher* dispatcher = PluginDispatcher::GetForInstance(instance());
+  if (!dispatcher)
+    return false;
+
+  command_buffer_.reset(new CommandBuffer(host_resource(), dispatcher));
+
+  return CreateGLES2Impl(kCommandBufferSize, kTransferBufferSize);
+}
+
+PP_Bool Graphics3D::InitCommandBuffer(int32_t size) {
+  return PP_FALSE;
+}
+
+PP_Bool Graphics3D::GetRingBuffer(int* shm_handle, uint32_t* shm_size) {
+  return PP_FALSE;
+}
+
+PP_Graphics3DTrustedState Graphics3D::GetState() {
+  return GetErrorState();
+}
+
+PP_Bool Graphics3D::Flush(int32_t put_offset) {
+  return PP_FALSE;
+}
+
+PP_Graphics3DTrustedState Graphics3D::FlushSync(int32_t put_offset) {
+  return GetErrorState();
+}
+
+int32_t Graphics3D::CreateTransferBuffer(uint32_t size) {
+  return PP_FALSE;
+}
+
+PP_Bool Graphics3D::DestroyTransferBuffer(int32_t id) {
+  return PP_FALSE;
+}
+
+PP_Bool Graphics3D::GetTransferBuffer(int32_t id,
+                                      int* shm_handle,
+                                      uint32_t* shm_size) {
+  return PP_FALSE;
+}
+
+PP_Graphics3DTrustedState Graphics3D::FlushSyncFast(int32_t put_offset,
+                                                    int32_t last_known_get) {
+  return GetErrorState();
+}
+
+gpu::CommandBuffer* Graphics3D::GetCommandBuffer() {
+  return command_buffer_.get();
+}
+
+int32 Graphics3D::DoSwapBuffers() {
+  IPC::Message* msg = new PpapiHostMsg_PPBGraphics3D_SwapBuffers(
+      INTERFACE_ID_PPB_GRAPHICS_3D, host_resource());
+  msg->set_unblock(true);
+  GetDispatcher()->Send(msg);
+
+  gles2_impl()->SwapBuffers();
+  return PP_OK_COMPLETIONPENDING;
+}
+
+PPB_Graphics3D_Proxy::PPB_Graphics3D_Proxy(Dispatcher* dispatcher,
+                                           const void* target_interface)
+    : InterfaceProxy(dispatcher, target_interface) {
+}
+
+PPB_Graphics3D_Proxy::~PPB_Graphics3D_Proxy() {
+}
+
+// static
+const InterfaceProxy::Info* PPB_Graphics3D_Proxy::GetInfo() {
+  static const Info info = {
+    ::ppapi::thunk::GetPPB_Graphics3D_Thunk(),
+    PPB_GRAPHICS_3D_DEV_INTERFACE,
+    INTERFACE_ID_PPB_GRAPHICS_3D,
+    false,
+    &CreateGraphics3DProxy,
+  };
+  return &info;
+}
+
+// static
+PP_Resource PPB_Graphics3D_Proxy::CreateProxyResource(
+    PP_Instance instance,
+    PP_Config3D_Dev config,
+    PP_Resource share_context,
+    const int32_t* attrib_list) {
+  PluginDispatcher* dispatcher = PluginDispatcher::GetForInstance(instance);
+  if (!dispatcher)
+    return PP_ERROR_BADARGUMENT;
+
+  // TODO(alokp): Support shared context.
+  DCHECK_EQ(0, share_context);
+  if (share_context != 0)
+    return 0;
+
+  std::vector<int32_t> attribs;
+  if (attrib_list) {
+    for (const int32_t* attr = attrib_list;
+         *attr != PP_GRAPHICS3DATTRIB_NONE;
+         ++attr) {
+      attribs.push_back(*attr);
+    }
+    attribs.push_back(PP_GRAPHICS3DATTRIB_NONE);
+  }
+
+  HostResource result;
+  dispatcher->Send(new PpapiHostMsg_PPBGraphics3D_Create(
+      INTERFACE_ID_PPB_GRAPHICS_3D, instance, config, attribs, &result));
+  if (result.is_null())
+    return 0;
+
+  linked_ptr<Graphics3D> graphics_3d(new Graphics3D(result));
+  if (!graphics_3d->Init())
+    return 0;
+
+  return PluginResourceTracker::GetInstance()->AddResource(graphics_3d);
+}
+
+bool PPB_Graphics3D_Proxy::OnMessageReceived(const IPC::Message& msg) {
+  bool handled = true;
+  IPC_BEGIN_MESSAGE_MAP(PPB_Graphics3D_Proxy, msg)
+    IPC_MESSAGE_HANDLER(PpapiHostMsg_PPBGraphics3D_Create,
+                        OnMsgCreate)
+    IPC_MESSAGE_HANDLER(PpapiHostMsg_PPBGraphics3D_InitCommandBuffer,
+                        OnMsgInitCommandBuffer)
+    IPC_MESSAGE_HANDLER(PpapiHostMsg_PPBGraphics3D_GetState,
+                        OnMsgGetState)
+    IPC_MESSAGE_HANDLER(PpapiHostMsg_PPBGraphics3D_Flush,
+                        OnMsgFlush)
+    IPC_MESSAGE_HANDLER(PpapiHostMsg_PPBGraphics3D_AsyncFlush,
+                        OnMsgAsyncFlush)
+    IPC_MESSAGE_HANDLER(PpapiHostMsg_PPBGraphics3D_CreateTransferBuffer,
+                        OnMsgCreateTransferBuffer)
+    IPC_MESSAGE_HANDLER(PpapiHostMsg_PPBGraphics3D_DestroyTransferBuffer,
+                        OnMsgDestroyTransferBuffer)
+    IPC_MESSAGE_HANDLER(PpapiHostMsg_PPBGraphics3D_GetTransferBuffer,
+                        OnMsgGetTransferBuffer)
+    IPC_MESSAGE_HANDLER(PpapiHostMsg_PPBGraphics3D_SwapBuffers,
+                        OnMsgSwapBuffers)
+
+    IPC_MESSAGE_HANDLER(PpapiMsg_PPBGraphics3D_SwapBuffersACK,
+                        OnMsgSwapBuffersACK)
+    IPC_MESSAGE_UNHANDLED(handled = false)
+
+  IPC_END_MESSAGE_MAP()
+  // FIXME(brettw) handle bad messages!
+  return handled;
+}
+
+void PPB_Graphics3D_Proxy::OnMsgCreate(PP_Instance instance,
+                                       PP_Config3D_Dev config,
+                                       const std::vector<int32_t>& attribs,
+                                       HostResource* result) {
+  if (attribs.empty() || attribs.back() != 0)
+    return;  // Bad message.
+
+  EnterFunctionNoLock<ResourceCreationAPI> enter(instance, true);
+  if (enter.succeeded()) {
+    result->SetHostResource(
+        instance,
+        enter.functions()->CreateGraphics3DRaw(instance, config, 0,
+                                               &attribs.front()));
+  }
+}
+
+void PPB_Graphics3D_Proxy::OnMsgInitCommandBuffer(
+    const HostResource& context,
+    int32 size,
+    base::SharedMemoryHandle* ring_buffer) {
+  *ring_buffer = base::SharedMemory::NULLHandle();
+  EnterHostFromHostResource<PPB_Graphics3D_API> enter(context);
+  if (enter.failed())
+    return;
+
+  if (!enter.object()->InitCommandBuffer(size))
+    return;
+
+  int shm_handle;
+  uint32_t shm_size;
+  if (!enter.object()->GetRingBuffer(&shm_handle, &shm_size))
+    return;
+  *ring_buffer = TransportSHMHandleFromInt(dispatcher(), shm_handle);
+}
+
+void PPB_Graphics3D_Proxy::OnMsgGetState(const HostResource& context,
+                                         gpu::CommandBuffer::State* state) {
+  EnterHostFromHostResource<PPB_Graphics3D_API> enter(context);
+  if (enter.failed())
+    return;
+  PP_Graphics3DTrustedState pp_state = enter.object()->GetState();
+  *state = GPUStateFromPPState(pp_state);
+}
+
+void PPB_Graphics3D_Proxy::OnMsgFlush(const HostResource& context,
+                                      int32 put_offset,
+                                      int32 last_known_get,
+                                      gpu::CommandBuffer::State* state) {
+  EnterHostFromHostResource<PPB_Graphics3D_API> enter(context);
+  if (enter.failed())
+    return;
+  PP_Graphics3DTrustedState pp_state = enter.object()->FlushSyncFast(
+      put_offset, last_known_get);
+  *state = GPUStateFromPPState(pp_state);
+}
+
+void PPB_Graphics3D_Proxy::OnMsgAsyncFlush(const HostResource& context,
+                                           int32 put_offset) {
+  EnterHostFromHostResource<PPB_Graphics3D_API> enter(context);
+  if (enter.succeeded())
+    enter.object()->Flush(put_offset);
+}
+
+void PPB_Graphics3D_Proxy::OnMsgCreateTransferBuffer(
+    const HostResource& context,
+    int32 size,
+    int32* id) {
+  EnterHostFromHostResource<PPB_Graphics3D_API> enter(context);
+  if (enter.succeeded())
+    *id = enter.object()->CreateTransferBuffer(size);
+  else
+    *id = 0;
+}
+
+void PPB_Graphics3D_Proxy::OnMsgDestroyTransferBuffer(
+    const HostResource& context,
+    int32 id) {
+  EnterHostFromHostResource<PPB_Graphics3D_API> enter(context);
+  if (enter.succeeded())
+    enter.object()->DestroyTransferBuffer(id);
+}
+
+void PPB_Graphics3D_Proxy::OnMsgGetTransferBuffer(
+    const HostResource& context,
+    int32 id,
+    base::SharedMemoryHandle* transfer_buffer,
+    uint32* size) {
+  *transfer_buffer = base::SharedMemory::NULLHandle();
+  *size = 0;
+
+  EnterHostFromHostResource<PPB_Graphics3D_API> enter(context);
+  int shm_handle = 0;
+  uint32_t shm_size = 0;
+  if (enter.succeeded() &&
+      enter.object()->GetTransferBuffer(id, &shm_handle, &shm_size)) {
+    *transfer_buffer = TransportSHMHandleFromInt(dispatcher(), shm_handle);
+    *size = shm_size;
+  }
+}
+
+void PPB_Graphics3D_Proxy::OnMsgSwapBuffers(const HostResource& context) {
+  CompletionCallback callback = callback_factory_.NewOptionalCallback(
+      &PPB_Graphics3D_Proxy::SendSwapBuffersACKToPlugin, context);
+
+  EnterHostFromHostResource<PPB_Graphics3D_API> enter(context);
+  int32_t result = PP_ERROR_BADRESOURCE;
+  if (enter.succeeded())
+    result = enter.object()->SwapBuffers(callback.pp_completion_callback());
+  if (result != PP_OK_COMPLETIONPENDING) {
+    // There was some error, so we won't get a flush callback. We need to now
+    // issue the ACK to the plugin hears about the error. This will also clean
+    // up the data associated with the callback.
+    callback.Run(result);
+  }
+}
+
+void PPB_Graphics3D_Proxy::OnMsgSwapBuffersACK(const HostResource& resource,
+                                              int32_t pp_error) {
+  EnterPluginFromHostResource<PPB_Graphics3D_API> enter(resource);
+  if (enter.succeeded())
+    static_cast<Graphics3D*>(enter.object())->SwapBuffersACK(pp_error);
+}
+
+void PPB_Graphics3D_Proxy::SendSwapBuffersACKToPlugin(
+    int32_t result,
+    const HostResource& context) {
+  dispatcher()->Send(new PpapiMsg_PPBGraphics3D_SwapBuffersACK(
+      INTERFACE_ID_PPB_GRAPHICS_3D, context, result));
+}
+
+}  // namespace proxy
+}  // namespace pp
+
Property changes on: ppapi\proxy\ppb_graphics_3d_proxy.cc
___________________________________________________________________
Added: svn:eol-style
   + LF