/*
 * Copyright (c) 2019-2024, NVIDIA CORPORATION. All rights reserved.
 *
 * 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.
 */
#version 450 core
#pragma use_vulkan_memory_model
#extension GL_EXT_scalar_block_layout : enable
#extension GL_KHR_memory_scope_semantics : enable
#extension GL_KHR_cooperative_matrix : enable
#extension GL_EXT_shader_explicit_arithmetic_types_float16 : enable
#extension GL_EXT_shader_explicit_arithmetic_types_int8 : enable
#extension GL_EXT_shader_explicit_arithmetic_types_int32 : enable
#extension GL_EXT_buffer_reference : enable
#extension GL_EXT_control_flow_attributes : enable

layout(constant_id = 0) const uint lM = 1;
layout(constant_id = 1) const uint lN = 1;
layout(constant_id = 2) const uint lK = 1;
layout(constant_id = 3) const uint TILE_M = 1;
layout(constant_id = 4) const uint TILE_N = 1;
layout(constant_id = 5) const uint TILE_K = 1;
layout(constant_id = 6) const uint K = 1;

#define A_BITS 16
#define A_TYPE float16_t
#define C_BITS 16
#define C_TYPE float16_t

buffer Output { C_TYPE x[]; } outputO;

shared uvec4 Ash[128];
shared uvec4 Bsh[128];

const uint C_ROWS = 2;
const uint C_COLS = 2;
coopmat<C_TYPE, gl_ScopeSubgroup, lM, lN, gl_MatrixUseAccumulator> result[C_ROWS][C_COLS];

void main()
{
    [[unroll]] for (uint i = 0; i < C_ROWS; ++i) {
        [[unroll]] for (uint j = 0; j < C_COLS; ++j) {
            result[i][j] = coopmat<C_TYPE, gl_ScopeSubgroup, lM, lN, gl_MatrixUseAccumulator>(0.0);
        }
    }

    for (uint chunkK = 0; chunkK < K; chunkK += TILE_K) {
        [[unroll]] for (uint k = 0; k < TILE_K / lK; ++k)
        {
            coopmat<A_TYPE, gl_ScopeSubgroup, lM, lK, gl_MatrixUseA> matA[C_ROWS];
            [[unroll]] for (uint i = 0; i < C_ROWS; ++i) {
                coopMatLoad(matA[i], Ash, 0, 0, gl_CooperativeMatrixLayoutRowMajor);
            }

            coopmat<A_TYPE, gl_ScopeSubgroup, lK, lN, gl_MatrixUseB> matB;
            [[unroll]] for (uint j = 0; j < C_COLS; ++j) {
                coopMatLoad(matB, Bsh, 0, 0, gl_CooperativeMatrixLayoutRowMajor);

                [[unroll]] for (uint i = 0; i < C_ROWS; ++i) {
                    result[i][j] = coopMatMulAdd(matA[i], matB, result[i][j]);
                }
            }
        }
    }

    [[unroll]] for (uint i = 0; i < C_ROWS; ++i) {
        [[unroll]] for (uint j = 0; j < C_COLS; ++j) {
            coopMatStore(result[i][j], outputO.x, 0, 0, gl_CooperativeMatrixLayoutRowMajor);
        }
    }
}