src/dynamic_fusion/sketch/gpu/operators/GpuMatMul.cpp

/*
 * Copyright (c) 2023-2024 Arm Limited.
 *
 * SPDX-License-Identifier: MIT
 *
 * Permission is hereby granted, free of charge, to any person obtaining a copy
 * of this software and associated documentation files (the "Software"), to
 * deal in the Software without restriction, including without limitation the
 * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
 * sell copies of the Software, and to permit persons to whom the Software is
 * furnished to do so, subject to the following conditions:
 *
 * The above copyright notice and this permission notice shall be included in all
 * copies or substantial portions of the Software.
 *
 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
 * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
 * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
 * SOFTWARE.
 */

#include "arm_compute/dynamic_fusion/sketch/gpu/operators/GpuMatMul.h"

#include "arm_compute/core/KernelDescriptors.h"
#include "arm_compute/core/utils/misc/ShapeCalculator.h"
#include "arm_compute/core/Validate.h"
#include "arm_compute/runtime/CL/CLScheduler.h"

#include "src/common/utils/Log.h"
#include "src/core/helpers/AutoConfiguration.h"
#include "src/dynamic_fusion/sketch/ArgumentPack.h"
#include "src/dynamic_fusion/sketch/gpu/components/cl/ClComponentMatMul.h"
#include "src/dynamic_fusion/sketch/gpu/GpuWorkloadSketchImpl.h"

namespace arm_compute
{
namespace experimental
{
namespace dynamic_fusion
{
namespace
{
void calculate_and_init_dst_if_empty(ITensorInfo             *dst,
                                     const ITensorInfo       *lhs,
                                     const ITensorInfo       *rhs,
                                     const MatMulAttributes  &attributes,
                                     const GpuMatMulSettings &settings)
{
    ARM_COMPUTE_UNUSED(attributes);

    if (dst->total_size() == 0U)
    {
        const auto dst_shape = misc::shape_calculator::compute_matmul_shape(
            lhs->tensor_shape(), rhs->tensor_shape(),
            MatMulKernelInfo(attributes.adj_lhs(), attributes.adj_rhs(), settings.m0(), settings.n0(), settings.k0()));

        auto_init_if_empty(*dst, lhs->clone()->set_tensor_shape(dst_shape));
    }
}

/* A helper method to reduce the duplication in dst tensor initialization
*  when calling validate()
*/
Status is_supported_op_helper(const GpuWorkloadContext &context,
                              const ITensorInfo        *lhs,
                              const ITensorInfo        *rhs,
                              const ITensorInfo        *dst,
                              const MatMulAttributes   &attributes,
                              const GpuMatMulSettings  &settings)
{
    ARM_COMPUTE_RETURN_ERROR_ON_NULLPTR(lhs, rhs);

    TensorInfo         dst_info_to_validate;
    const ITensorInfo *dst_info_to_validate_ptr = &dst_info_to_validate;

    if (dst != nullptr)
    {
        dst_info_to_validate_ptr = dst;
    }

    calculate_and_init_dst_if_empty(&dst_info_to_validate, lhs, rhs, attributes, settings);

    // Check support level
    // Data type
    ARM_COMPUTE_RETURN_ERROR_ON_DATA_TYPE_CHANNEL_NOT_IN(lhs, 1, DataType::F16, DataType::F32);

    // Check components
    if (context.gpu_language() == GpuLanguage::OpenCL)
    {
        const auto cl_compile_ctx = context.cl_compile_context();
        ARM_COMPUTE_RETURN_ERROR_ON(cl_compile_ctx == nullptr);
        // Validate MatMul Component
        {
            const auto properties =
                IGpuKernelComponent::Properties().stage(UnitWorkloadStage{UnitWorkloadStage::Stage::Run});

            ArgumentPack<ITensorInfo> arguments;
            arguments.add_const_tensor(ACL_SRC_0, lhs);
            arguments.add_const_tensor(ACL_SRC_1, rhs);
            arguments.add_const_tensor(ACL_DST_0, dst_info_to_validate_ptr);

            ARM_COMPUTE_RETURN_ON_ERROR(ClComponentMatMul::validate(properties, arguments, attributes, settings));
        }
    }
    else
    {
        ARM_COMPUTE_RETURN_ERROR_MSG("Unimplemented Gpu language");
    }
    return Status{};
}

constexpr GpuOperatorType operator_type = GpuOperatorType::Complex;
} // namespace

int GpuMatMulSettings::n0() const
{
    return _n0;
}

GpuMatMulSettings &GpuMatMulSettings::n0(int n0)
{
    _n0 = n0;
    return *this;
}

int GpuMatMulSettings::m0() const
{
    return _m0;
}

GpuMatMulSettings &GpuMatMulSettings::m0(int m0)
{
    _m0 = m0;
    return *this;
}

int GpuMatMulSettings::k0() const
{
    return _k0;
}

GpuMatMulSettings &GpuMatMulSettings::k0(int k0)
{
    _k0 = k0;
    return *this;
}

Status GpuMatMul::is_supported_op(const GpuWorkloadContext &context,
                                  const ITensorInfo        *lhs,
                                  const ITensorInfo        *rhs,
                                  const MatMulAttributes   &attributes,
                                  const GpuMatMulSettings  &settings)
{
    return is_supported_op_helper(context, lhs, rhs, nullptr, attributes, settings);
}

Status GpuMatMul::validate_op(const GpuWorkloadSketch &sketch,
                              const ITensorInfo       *lhs,
                              const ITensorInfo       *rhs,
                              const MatMulAttributes  &attributes,
                              const GpuMatMulSettings &settings)
{
    ARM_COMPUTE_RETURN_ERROR_ON_NULLPTR(lhs, rhs);

    // Check if tensors have valid id. I.e. they are created from a sketch
    ARM_COMPUTE_RETURN_ERROR_ON(!lhs->has_valid_id() || !rhs->has_valid_id());

    // Refer to GpuMatmul::validate_op() for id-validness of this TensorInfo object
    TensorInfo dst_info_to_validate;

    // Auto initialize dst tensor info
    calculate_and_init_dst_if_empty(&dst_info_to_validate, lhs, rhs, attributes, settings);

    // Perform fusion test
    // Check if operator meets fusion constraints
    ArgumentPack<ITensorInfo> tensors;
    tensors.add_const_tensor(ACL_SRC_0, lhs);
    tensors.add_const_tensor(ACL_SRC_1, rhs);
    tensors.add_const_tensor(ACL_DST_0, &dst_info_to_validate);
    const auto op = sketch.implementation().operator_group().new_operator(operator_type, tensors);
    ARM_COMPUTE_RETURN_ERROR_ON_MSG(!sketch.implementation().operator_group().try_add_operator(op),
                                    "Operator fusion test failed. This operator cannot be fused into the workload");

    // Check if configuration is supported
    return is_supported_op_helper(*sketch.gpu_context(), lhs, rhs, &dst_info_to_validate, attributes, settings);
}

ITensorInfo *GpuMatMul::create_op(GpuWorkloadSketch &sketch,
                                  ITensorInfo       *lhs,
                                  ITensorInfo       *rhs,
                                  const Attributes  &attributes,
                                  const Settings    &settings)
{
    ARM_COMPUTE_LOG_PARAMS(lhs, rhs, attributes, settings);

    ITensorInfo *dst = sketch.implementation().create_virtual_tensor();

    // Assert validation
    ARM_COMPUTE_ERROR_THROW_ON(GpuMatMul::validate_op(sketch, lhs, rhs, attributes, settings));
    ARM_COMPUTE_ERROR_ON_NULLPTR(lhs, rhs, dst);

    // Auto initialize dst tensor
    calculate_and_init_dst_if_empty(dst, lhs, rhs, attributes, settings);

    // Translate into components and add to component graph
    auto      &comp_graph = sketch.implementation().component_graph();
    const auto sketch_ctx = sketch.implementation().context();

    if (sketch_ctx->gpu_language() == GpuLanguage::OpenCL)
    {
        auto properties = IGpuKernelComponent::Properties();
        properties.stage(UnitWorkloadStage{UnitWorkloadStage::Stage::Run});

        ArgumentPack<ITensorInfo> arguments;
        arguments.add_const_tensor(ACL_SRC_0, lhs);
        arguments.add_const_tensor(ACL_SRC_1, rhs);
        arguments.add_const_tensor(ACL_DST_0, dst);
        comp_graph.add_new_component<ClComponentMatMul>(properties, arguments, attributes, settings);
    }
    else
    {
        ARM_COMPUTE_ERROR("Unimplemented Gpu language");
    }

    // Set up fusion test by adding to the Operator Group
    // Note this has to be performed after all the components have been successfully added to the component graph

    // Pack tensor infos
    ArgumentPack<ITensorInfo> tensors;
    tensors.add_const_tensor(ACL_SRC_0, lhs);
    tensors.add_const_tensor(ACL_SRC_1, rhs);
    tensors.add_const_tensor(ACL_DST_0, dst);

    const auto op = sketch.implementation().operator_group().new_operator(operator_type, tensors);
    sketch.implementation().operator_group().add_operator(op);

    return dst;
}

} // namespace dynamic_fusion
} // namespace experimental
} // namespace arm_compute