onednn-src 0.1.13

/*******************************************************************************
* Copyright 2020 Intel Corporation
* Copyright 2020 Codeplay Software Limited
*
* 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 "gpu/nvidia/cudnn_convolution.hpp"
#include "gpu/nvidia/stream.hpp"
#include "gpu/nvidia/sycl_cuda_scoped_context.hpp"
#include "gpu/nvidia/sycl_cuda_utils.hpp"
#include "xpu/sycl/memory_storage_helper.hpp"

namespace dnnl {
namespace impl {
namespace gpu {
namespace nvidia {

status_t cudnn_convolution_fwd_t::execute_convolution(
        const exec_ctx_t &ctx, bool with_bias, bool with_scratchpad) const {
    nvidia::stream_t *cuda_stream
            = utils::downcast<nvidia::stream_t *>(ctx.stream());

    return cuda_stream->interop_task([&](::sycl::handler &cgh) {
        auto arg_src = CTX_IN_SYCL_MEMORY(DNNL_ARG_SRC);
        auto arg_weights = CTX_IN_SYCL_MEMORY(DNNL_ARG_WEIGHTS);
        auto arg_dst = CTX_OUT_SYCL_MEMORY(DNNL_ARG_DST);
        auto arg_bias = CTX_IN_SYCL_MEMORY(DNNL_ARG_BIAS);
        auto arg_scratch = CTX_SCRATCH_SYCL_MEMORY(
                memory_tracking::names::key_conv_cudnn_algo);
        auto arg_filter_scratch = CTX_SCRATCH_SYCL_MEMORY(
                memory_tracking::names::key_conv_cudnn_filter);
        auto arg_src_scale
                = CTX_IN_SYCL_MEMORY(DNNL_ARG_ATTR_SCALES | DNNL_ARG_SRC);
        auto arg_wei_scale
                = CTX_IN_SYCL_MEMORY(DNNL_ARG_ATTR_SCALES | DNNL_ARG_WEIGHTS);
        auto arg_dst_scale
                = CTX_IN_SYCL_MEMORY(DNNL_ARG_ATTR_SCALES | DNNL_ARG_DST);

        xpu::sycl::interop_memory_arg_t<::sycl::access::mode::read_write>
                temp_dst;

        if (pd()->use_temp_dst()) {
            memory_storage_t *temp_dst_mem = scratch_storage.get();
            temp_dst = xpu::sycl::interop_memory_arg_t<
                    ::sycl::access::mode::read_write>(temp_dst_mem, cgh);
        }

        xpu::sycl::interop_memory_arg_t<::sycl::access::mode::read_write>
                y_fp32_data;

        if (!arg_dst_scale.empty() || !arg_src_scale.empty()
                || !arg_wei_scale.empty()) {
            memory_storage_t *y_fp32_data_mem = scratch_storage_3.get();
            y_fp32_data = xpu::sycl::interop_memory_arg_t<
                    ::sycl::access::mode::read_write>(y_fp32_data_mem, cgh);
        }

        compat::host_task(cgh, [=, this](const compat::interop_handle &ih) {
            auto &sycl_engine = *utils::downcast<nvidia::engine_t *>(
                    cuda_stream->engine());
            auto sc = cuda_sycl_scoped_context_handler_t(sycl_engine);
            auto handle = cuda_stream->get_cudnn_handle();

            std::vector<void *> args;
            args.push_back(arg_src.get_native_pointer(ih));
            args.push_back(arg_weights.get_native_pointer(ih));
            args.push_back(arg_dst.get_native_pointer(ih));
            args.push_back(arg_bias.get_native_pointer(ih));
            args.push_back(arg_scratch.get_native_pointer(ih));
            args.push_back(arg_filter_scratch.get_native_pointer(ih));
            args.push_back(temp_dst.get_native_pointer(ih));
            args.push_back(arg_src_scale.get_native_pointer(ih));
            args.push_back(arg_wei_scale.get_native_pointer(ih));
            args.push_back(arg_dst_scale.get_native_pointer(ih));
            args.push_back(y_fp32_data.get_native_pointer(ih));

            pd()->impl_->execute(handle, args);
        });
    });
}

status_t cudnn_convolution_bwd_data_t::execute_convolution(
        const exec_ctx_t &ctx, bool with_bias, bool with_scratchpad) const {
    nvidia::stream_t *cuda_stream
            = utils::downcast<nvidia::stream_t *>(ctx.stream());

    return cuda_stream->interop_task([&](::sycl::handler &cgh) {
        auto arg_diff_src = CTX_OUT_SYCL_MEMORY(DNNL_ARG_DIFF_SRC);
        auto arg_weights = CTX_IN_SYCL_MEMORY(DNNL_ARG_WEIGHTS);
        auto arg_diff_dst = CTX_IN_SYCL_MEMORY(DNNL_ARG_DIFF_DST);
        auto arg_bias = CTX_IN_SYCL_MEMORY(DNNL_ARG_BIAS);
        auto arg_scratch = CTX_SCRATCH_SYCL_MEMORY(
                memory_tracking::names::key_conv_cudnn_algo);
        auto arg_filter_scratch = CTX_SCRATCH_SYCL_MEMORY(
                memory_tracking::names::key_conv_cudnn_filter);

        compat::host_task(cgh, [=, this](const compat::interop_handle &ih) {
            auto &sycl_engine = *utils::downcast<nvidia::engine_t *>(
                    cuda_stream->engine());
            auto sc = cuda_sycl_scoped_context_handler_t(sycl_engine);
            auto handle = cuda_stream->get_cudnn_handle();

            std::vector<void *> args;
            args.push_back(arg_diff_src.get_native_pointer(ih));
            args.push_back(arg_weights.get_native_pointer(ih));
            args.push_back(arg_diff_dst.get_native_pointer(ih));
            args.push_back(arg_bias.get_native_pointer(ih));
            args.push_back(arg_scratch.get_native_pointer(ih));
            args.push_back(arg_filter_scratch.get_native_pointer(ih));

            pd()->impl_->execute(handle, args);
        });
    });
}

status_t cudnn_convolution_bwd_weights_t::execute_zero_dims(
        const exec_ctx_t &ctx) const {
    nvidia::stream_t *cuda_stream
            = utils::downcast<nvidia::stream_t *>(ctx.stream());

    return cuda_stream->interop_task([&](::sycl::handler &cgh) {
        auto arg_diff_weights = CTX_OUT_SYCL_MEMORY(DNNL_ARG_DIFF_WEIGHTS);
        auto arg_diff_bias = CTX_OUT_SYCL_MEMORY(DNNL_ARG_DIFF_BIAS);

        compat::host_task(cgh, [=, this](const compat::interop_handle &ih) {
            auto &sycl_engine = *utils::downcast<nvidia::engine_t *>(
                    cuda_stream->engine());
            auto sc = cuda_sycl_scoped_context_handler_t(sycl_engine);
            auto handle = cuda_stream->get_cudnn_handle();

            void *weights = arg_diff_weights.get_native_pointer(ih);
            void *bias = arg_diff_bias.get_native_pointer(ih);

            pd()->impl_->execute_set_weights_bias(handle, weights, bias, 0.f);
        });
    });
}

status_t cudnn_convolution_bwd_weights_t::execute_convolution(
        const exec_ctx_t &ctx, bool with_bias, bool with_scratchpad) const {
    nvidia::stream_t *cuda_stream
            = utils::downcast<nvidia::stream_t *>(ctx.stream());

    return cuda_stream->interop_task([&](::sycl::handler &cgh) {
        auto arg_src = CTX_IN_SYCL_MEMORY(DNNL_ARG_SRC);
        auto arg_diff_weights = CTX_OUT_SYCL_MEMORY(DNNL_ARG_DIFF_WEIGHTS);
        auto arg_diff_dst = CTX_IN_SYCL_MEMORY(DNNL_ARG_DIFF_DST);
        auto arg_scratch = CTX_SCRATCH_SYCL_MEMORY(
                memory_tracking::names::key_conv_cudnn_algo);
        auto arg_filter_scratch = CTX_SCRATCH_SYCL_MEMORY(
                memory_tracking::names::key_conv_cudnn_filter);

        xpu::sycl::interop_memory_arg_t<::sycl::access::mode::write>
                arg_diff_bias;

        if (with_bias) {
            arg_diff_bias = CTX_OUT_SYCL_MEMORY(DNNL_ARG_DIFF_BIAS);
        }

        compat::host_task(cgh, [=, this](const compat::interop_handle &ih) {
            auto &sycl_engine = *utils::downcast<nvidia::engine_t *>(
                    cuda_stream->engine());
            auto sc = cuda_sycl_scoped_context_handler_t(sycl_engine);
            auto handle = cuda_stream->get_cudnn_handle();

            std::vector<void *> args;
            args.push_back(arg_src.get_native_pointer(ih));
            args.push_back(arg_diff_weights.get_native_pointer(ih));
            args.push_back(arg_diff_dst.get_native_pointer(ih));
            args.push_back(arg_diff_bias.get_native_pointer(ih));
            args.push_back(arg_scratch.get_native_pointer(ih));
            args.push_back(arg_filter_scratch.get_native_pointer(ih));

            pd()->impl_->execute(handle, args);
        });
    });
}

} // namespace nvidia
} // namespace gpu
} // namespace impl
} // namespace dnnl