dawn-cmake/src/dawn_wire/server/ServerBuffer.cpp
Austin Eng cac0442277 dawn_wire: Support chunked commands
This CL adds support for chunking large commands by first serializing
large commands first into a separate buffer, and then sending the
buffer data chunk by chunk.

This code path is used for large writeBuffer and writeTexture, as well
as the inline memory transfer service for buffer mapping. The transfer
for writeBuffer and writeTexture will be optimized further in Chrome,
and the inline memory transfer service is currently used only in tests.

Bug: chromium:1123861, chromium:951558
Change-Id: I02491a44e653e2383174958d9c3d4a4db6fd7bde
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/28882
Commit-Queue: Austin Eng <enga@chromium.org>
Reviewed-by: Stephen White <senorblanco@chromium.org>
2020-10-13 22:35:34 +00:00

264 lines
11 KiB
C++

// Copyright 2019 The Dawn Authors
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
// http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.
#include "common/Assert.h"
#include "dawn_wire/server/Server.h"
#include <memory>
namespace dawn_wire { namespace server {
bool Server::PreHandleBufferUnmap(const BufferUnmapCmd& cmd) {
auto* buffer = BufferObjects().Get(cmd.selfId);
DAWN_ASSERT(buffer != nullptr);
// The buffer was unmapped. Clear the Read/WriteHandle.
buffer->readHandle = nullptr;
buffer->writeHandle = nullptr;
buffer->mapWriteState = BufferMapWriteState::Unmapped;
return true;
}
bool Server::PreHandleBufferDestroy(const BufferDestroyCmd& cmd) {
// Destroying a buffer does an implicit unmapping.
auto* buffer = BufferObjects().Get(cmd.selfId);
DAWN_ASSERT(buffer != nullptr);
// The buffer was destroyed. Clear the Read/WriteHandle.
buffer->readHandle = nullptr;
buffer->writeHandle = nullptr;
buffer->mapWriteState = BufferMapWriteState::Unmapped;
return true;
}
bool Server::DoBufferMapAsync(ObjectId bufferId,
uint32_t requestSerial,
WGPUMapModeFlags mode,
size_t offset,
size_t size,
uint64_t handleCreateInfoLength,
const uint8_t* handleCreateInfo) {
// These requests are just forwarded to the buffer, with userdata containing what the
// client will require in the return command.
// The null object isn't valid as `self`
if (bufferId == 0) {
return false;
}
auto* buffer = BufferObjects().Get(bufferId);
if (buffer == nullptr) {
return false;
}
// The server only knows how to deal with write XOR read. Validate that.
bool isReadMode = mode & WGPUMapMode_Read;
bool isWriteMode = mode & WGPUMapMode_Write;
if (!(isReadMode ^ isWriteMode)) {
return false;
}
if (handleCreateInfoLength > std::numeric_limits<size_t>::max()) {
// This is the size of data deserialized from the command stream, which must be
// CPU-addressable.
return false;
}
std::unique_ptr<MapUserdata> userdata = std::make_unique<MapUserdata>();
userdata->server = this;
userdata->buffer = ObjectHandle{bufferId, buffer->generation};
userdata->bufferObj = buffer->handle;
userdata->requestSerial = requestSerial;
userdata->offset = offset;
userdata->size = size;
userdata->mode = mode;
// The handle will point to the mapped memory or staging memory for the mapping.
// Store it on the map request.
if (isWriteMode) {
// Deserialize metadata produced from the client to create a companion server handle.
MemoryTransferService::WriteHandle* writeHandle = nullptr;
if (!mMemoryTransferService->DeserializeWriteHandle(
handleCreateInfo, static_cast<size_t>(handleCreateInfoLength), &writeHandle)) {
return false;
}
ASSERT(writeHandle != nullptr);
userdata->writeHandle =
std::unique_ptr<MemoryTransferService::WriteHandle>(writeHandle);
} else {
ASSERT(isReadMode);
// Deserialize metadata produced from the client to create a companion server handle.
MemoryTransferService::ReadHandle* readHandle = nullptr;
if (!mMemoryTransferService->DeserializeReadHandle(
handleCreateInfo, static_cast<size_t>(handleCreateInfoLength), &readHandle)) {
return false;
}
ASSERT(readHandle != nullptr);
userdata->readHandle = std::unique_ptr<MemoryTransferService::ReadHandle>(readHandle);
}
mProcs.bufferMapAsync(buffer->handle, mode, offset, size, ForwardBufferMapAsync,
userdata.release());
return true;
}
bool Server::DoDeviceCreateBuffer(WGPUDevice device,
const WGPUBufferDescriptor* descriptor,
ObjectHandle bufferResult,
uint64_t handleCreateInfoLength,
const uint8_t* handleCreateInfo) {
// Create and register the buffer object.
auto* resultData = BufferObjects().Allocate(bufferResult.id);
if (resultData == nullptr) {
return false;
}
resultData->generation = bufferResult.generation;
resultData->handle = mProcs.deviceCreateBuffer(device, descriptor);
// If the buffer isn't mapped at creation, we are done.
if (!descriptor->mappedAtCreation) {
return handleCreateInfoLength == 0;
}
// This is the size of data deserialized from the command stream to create the write handle,
// which must be CPU-addressable.
if (handleCreateInfoLength > std::numeric_limits<size_t>::max()) {
return false;
}
void* mapping = mProcs.bufferGetMappedRange(resultData->handle, 0, descriptor->size);
if (mapping == nullptr) {
// A zero mapping is used to indicate an allocation error of an error buffer. This is a
// valid case and isn't fatal. Remember the buffer is an error so as to skip subsequent
// mapping operations.
resultData->mapWriteState = BufferMapWriteState::MapError;
return true;
}
// Deserialize metadata produced from the client to create a companion server handle.
MemoryTransferService::WriteHandle* writeHandle = nullptr;
if (!mMemoryTransferService->DeserializeWriteHandle(
handleCreateInfo, static_cast<size_t>(handleCreateInfoLength), &writeHandle)) {
return false;
}
// Set the target of the WriteHandle to the mapped GPU memory.
ASSERT(writeHandle != nullptr);
writeHandle->SetTarget(mapping, descriptor->size);
resultData->mapWriteState = BufferMapWriteState::Mapped;
resultData->writeHandle.reset(writeHandle);
return true;
}
bool Server::DoBufferUpdateMappedData(ObjectId bufferId,
uint64_t writeFlushInfoLength,
const uint8_t* writeFlushInfo) {
// The null object isn't valid as `self`
if (bufferId == 0) {
return false;
}
if (writeFlushInfoLength > std::numeric_limits<size_t>::max()) {
return false;
}
auto* buffer = BufferObjects().Get(bufferId);
if (buffer == nullptr) {
return false;
}
switch (buffer->mapWriteState) {
case BufferMapWriteState::Unmapped:
return false;
case BufferMapWriteState::MapError:
// The buffer is mapped but there was an error allocating mapped data.
// Do not perform the memcpy.
return true;
case BufferMapWriteState::Mapped:
break;
}
if (!buffer->writeHandle) {
// This check is performed after the check for the MapError state. It is permissible
// to Unmap and attempt to update mapped data of an error buffer.
return false;
}
// Deserialize the flush info and flush updated data from the handle into the target
// of the handle. The target is set via WriteHandle::SetTarget.
return buffer->writeHandle->DeserializeFlush(writeFlushInfo,
static_cast<size_t>(writeFlushInfoLength));
}
void Server::ForwardBufferMapAsync(WGPUBufferMapAsyncStatus status, void* userdata) {
auto data = static_cast<MapUserdata*>(userdata);
data->server->OnBufferMapAsyncCallback(status, data);
}
void Server::OnBufferMapAsyncCallback(WGPUBufferMapAsyncStatus status, MapUserdata* userdata) {
std::unique_ptr<MapUserdata> data(userdata);
// Skip sending the callback if the buffer has already been destroyed.
auto* bufferData = BufferObjects().Get(data->buffer.id);
if (bufferData == nullptr || bufferData->generation != data->buffer.generation) {
return;
}
bool isRead = data->mode & WGPUMapMode_Read;
bool isSuccess = status == WGPUBufferMapAsyncStatus_Success;
ReturnBufferMapAsyncCallbackCmd cmd;
cmd.buffer = data->buffer;
cmd.requestSerial = data->requestSerial;
cmd.status = status;
cmd.readInitialDataInfoLength = 0;
cmd.readInitialDataInfo = nullptr;
const void* readData = nullptr;
if (isSuccess && isRead) {
// Get the serialization size of the message to initialize ReadHandle data.
readData = mProcs.bufferGetConstMappedRange(data->bufferObj, data->offset, data->size);
cmd.readInitialDataInfoLength =
data->readHandle->SerializeInitialDataSize(readData, data->size);
}
SerializeCommand(cmd, cmd.readInitialDataInfoLength, [&](char* cmdSpace) {
if (isSuccess) {
if (isRead) {
// Serialize the initialization message into the space after the command.
data->readHandle->SerializeInitialData(readData, data->size, cmdSpace);
// The in-flight map request returned successfully.
// Move the ReadHandle so it is owned by the buffer.
bufferData->readHandle = std::move(data->readHandle);
} else {
// The in-flight map request returned successfully.
// Move the WriteHandle so it is owned by the buffer.
bufferData->writeHandle = std::move(data->writeHandle);
bufferData->mapWriteState = BufferMapWriteState::Mapped;
// Set the target of the WriteHandle to the mapped buffer data.
bufferData->writeHandle->SetTarget(
mProcs.bufferGetMappedRange(data->bufferObj, data->offset, data->size),
data->size);
}
}
});
}
}} // namespace dawn_wire::server