| // Copyright (c) Facebook, Inc. and its affiliates. |
| // All rights reserved. |
| // |
| // Copyright 2019 Google LLC |
| // |
| // This source code is licensed under the BSD-style license found in the |
| // LICENSE file in the root directory of this source tree. |
| |
| #include <assert.h> |
| #include <math.h> |
| #include <stddef.h> |
| #include <stdint.h> |
| #include <stdlib.h> |
| |
| #include <xnnpack.h> |
| #include <xnnpack/allocator.h> |
| #include <xnnpack/log.h> |
| #include <xnnpack/operator.h> |
| #include <xnnpack/params-init.h> |
| #include <xnnpack/params.h> |
| |
| |
| enum xnn_status xnn_create_add_nc_q8( |
| size_t channels, |
| size_t a_stride, |
| size_t b_stride, |
| size_t sum_stride, |
| uint8_t a_zero_point, |
| float a_scale, |
| uint8_t b_zero_point, |
| float b_scale, |
| uint8_t sum_zero_point, |
| float sum_scale, |
| uint8_t sum_min, |
| uint8_t sum_max, |
| uint32_t flags, |
| xnn_operator_t* add_op_out) |
| { |
| xnn_operator_t add_op = NULL; |
| enum xnn_status status = xnn_status_uninitialized; |
| |
| if (!xnn_params.initialized) { |
| xnn_log_error("failed to create %s operator: XNNPACK is not initialized", |
| xnn_operator_type_to_string(xnn_operator_type_add_nc_q8)); |
| goto error; |
| } |
| |
| status = xnn_status_invalid_parameter; |
| |
| if (channels == 0) { |
| xnn_log_error( |
| "failed to create %s operator with %zu channels: number of channels must be non-zero", |
| xnn_operator_type_to_string(xnn_operator_type_add_nc_q8), channels); |
| goto error; |
| } |
| |
| if (a_stride < channels) { |
| xnn_log_error( |
| "failed to create %s operator with A element stride of %zu: " |
| "stride must be at least as large as the number of channels (%zu)", |
| xnn_operator_type_to_string(xnn_operator_type_add_nc_q8), a_stride, channels); |
| goto error; |
| } |
| |
| if (b_stride < channels) { |
| xnn_log_error( |
| "failed to create %s operator with B element stride of %zu: " |
| "stride must be at least as large as the number of channels (%zu)", |
| xnn_operator_type_to_string(xnn_operator_type_add_nc_q8), b_stride, channels); |
| goto error; |
| } |
| |
| if (sum_stride < channels) { |
| xnn_log_error( |
| "failed to create %s operator with Sum element stride of %zu: " |
| "stride must be at least as large as the number of channels (%zu)", |
| xnn_operator_type_to_string(xnn_operator_type_add_nc_q8), sum_stride, channels); |
| goto error; |
| } |
| |
| if (a_scale <= 0.0f || !isnormal(a_scale)) { |
| xnn_log_error( |
| "failed to create %s operator with %.7g A scale: scale must be finite, normalized, and positive", |
| xnn_operator_type_to_string(xnn_operator_type_add_nc_q8), a_scale); |
| goto error; |
| } |
| |
| if (b_scale <= 0.0f || !isnormal(b_scale)) { |
| xnn_log_error( |
| "failed to create %s operator with %.7g B scale: scale must be finite, normalized, and positive", |
| xnn_operator_type_to_string(xnn_operator_type_add_nc_q8), b_scale); |
| goto error; |
| } |
| |
| if (sum_scale <= 0.0f || !isnormal(sum_scale)) { |
| xnn_log_error( |
| "failed to create %s operator with %.7g output scale: scale must be finite, normalized, and positive", |
| xnn_operator_type_to_string(xnn_operator_type_add_nc_q8), sum_scale); |
| goto error; |
| } |
| |
| if (sum_min >= sum_max) { |
| xnn_log_error( |
| "failed to create %s operator with [%" PRIu8 ", %" PRIu8 "] output range: range min must be below range max", |
| xnn_operator_type_to_string(xnn_operator_type_add_nc_q8), sum_min, sum_max); |
| goto error; |
| } |
| |
| status = xnn_status_unsupported_parameter; |
| |
| const float a_output_scale = a_scale / sum_scale; |
| if (a_output_scale < 0x1.0p-14f || a_output_scale >= 0x1.0p+8f) { |
| xnn_log_error( |
| "failed to create %s operator with %.7g A-to-output scale ratio: scale ratio must be in [2**-14, 2**8) range", |
| xnn_operator_type_to_string(xnn_operator_type_add_nc_q8), a_output_scale); |
| goto error; |
| } |
| |
| const float b_output_scale = b_scale / sum_scale; |
| if (b_output_scale < 0x1.0p-14f || b_output_scale >= 0x1.0p+8f) { |
| xnn_log_error( |
| "failed to create %s operator with %.7g A-to-output scale ratio: scale ratio must be in [2**-14, 2**8) range", |
| xnn_operator_type_to_string(xnn_operator_type_add_nc_q8), b_output_scale); |
| goto error; |
| } |
| |
| status = xnn_status_out_of_memory; |
| |
| add_op = xnn_allocate_zero_simd_memory(sizeof(struct xnn_operator)); |
| if (add_op == NULL) { |
| xnn_log_error( |
| "failed to allocate %zu bytes for %s operator descriptor", |
| sizeof(struct xnn_operator), xnn_operator_type_to_string(xnn_operator_type_add_nc_q8)); |
| goto error; |
| } |
| |
| add_op->channels = channels; |
| add_op->input_pixel_stride = a_stride; |
| add_op->input2_pixel_stride = b_stride; |
| add_op->output_pixel_stride = sum_stride; |
| add_op->params.q8_add = |
| xnn_init_q8_add_params( |
| a_zero_point, b_zero_point, sum_zero_point, |
| a_scale / sum_scale, b_scale / sum_scale, |
| sum_min, sum_max); |
| |
| add_op->type = xnn_operator_type_add_nc_q8; |
| add_op->ukernel.type = xnn_ukernel_type_add; |
| |
| add_op->state = xnn_run_state_invalid; |
| |
| *add_op_out = add_op; |
| return xnn_status_success; |
| |
| error: |
| xnn_delete_operator(add_op); |
| return status; |
| } |
| |
| enum xnn_status xnn_create_add_nc_f32( |
| size_t channels, |
| size_t a_stride, |
| size_t b_stride, |
| size_t sum_stride, |
| float sum_min, |
| float sum_max, |
| uint32_t flags, |
| xnn_operator_t* add_op_out) |
| { |
| xnn_operator_t add_op = NULL; |
| enum xnn_status status = xnn_status_uninitialized; |
| |
| if (!xnn_params.initialized) { |
| xnn_log_error("failed to create %s operator: XNNPACK is not initialized", |
| xnn_operator_type_to_string(xnn_operator_type_add_nc_f32)); |
| goto error; |
| } |
| |
| status = xnn_status_invalid_parameter; |
| |
| if (channels == 0) { |
| xnn_log_error( |
| "failed to create %s operator with %zu channels: number of channels must be non-zero", |
| xnn_operator_type_to_string(xnn_operator_type_add_nc_f32), channels); |
| goto error; |
| } |
| |
| if (a_stride < channels) { |
| xnn_log_error( |
| "failed to create %s operator with A element stride of %zu: " |
| "stride must be at least as large as the number of channels (%zu)", |
| xnn_operator_type_to_string(xnn_operator_type_add_nc_f32), a_stride, channels); |
| goto error; |
| } |
| |
| if (b_stride < channels) { |
| xnn_log_error( |
| "failed to create %s operator with B element stride of %zu: " |
| "stride must be at least as large as the number of channels (%zu)", |
| xnn_operator_type_to_string(xnn_operator_type_add_nc_f32), b_stride, channels); |
| goto error; |
| } |
| |
| if (sum_stride < channels) { |
| xnn_log_error( |
| "failed to create %s operator with Sum element stride of %zu: " |
| "stride must be at least as large as the number of channels (%zu)", |
| xnn_operator_type_to_string(xnn_operator_type_add_nc_f32), sum_stride, channels); |
| goto error; |
| } |
| |
| if (isnan(sum_min)) { |
| xnn_log_error( |
| "failed to create %s operator with NaN output lower bound: lower bound must be non-NaN", |
| xnn_operator_type_to_string(xnn_operator_type_add_nc_f32)); |
| goto error; |
| } |
| |
| if (isnan(sum_max)) { |
| xnn_log_error( |
| "failed to create %s operator with NaN output upper bound: upper bound must be non-NaN", |
| xnn_operator_type_to_string(xnn_operator_type_add_nc_f32)); |
| goto error; |
| } |
| |
| if (sum_min >= sum_max) { |
| xnn_log_error( |
| "failed to create %s operator with [%.7g, %.7g] output range: lower bound must be below upper bound", |
| xnn_operator_type_to_string(xnn_operator_type_add_nc_f32), sum_min, sum_max); |
| goto error; |
| } |
| |
| status = xnn_status_out_of_memory; |
| |
| add_op = xnn_allocate_zero_simd_memory(sizeof(struct xnn_operator)); |
| if (add_op == NULL) { |
| xnn_log_error( |
| "failed to allocate %zu bytes for %s operator descriptor", |
| sizeof(struct xnn_operator), xnn_operator_type_to_string(xnn_operator_type_add_nc_f32)); |
| goto error; |
| } |
| |
| add_op->channels = channels; |
| add_op->input_pixel_stride = a_stride; |
| add_op->input2_pixel_stride = b_stride; |
| add_op->output_pixel_stride = sum_stride; |
| add_op->params.f32_minmax = xnn_init_f32_minmax_params(sum_min, sum_max); |
| |
| add_op->type = xnn_operator_type_add_nc_f32; |
| add_op->ukernel.type = xnn_ukernel_type_add; |
| |
| add_op->state = xnn_run_state_invalid; |
| |
| *add_op_out = add_op; |
| return xnn_status_success; |
| |
| error: |
| xnn_delete_operator(add_op); |
| return status; |
| } |
| |
| enum xnn_status xnn_setup_add_nc_q8( |
| xnn_operator_t add_op, |
| size_t batch_size, |
| const uint8_t* a, |
| const uint8_t* b, |
| uint8_t* sum, |
| pthreadpool_t threadpool) |
| { |
| if (add_op->type != xnn_operator_type_add_nc_q8) { |
| xnn_log_error("failed to setup operator: operator type mismatch (expected %s, got %s)", |
| xnn_operator_type_to_string(xnn_operator_type_add_nc_q8), |
| xnn_operator_type_to_string(add_op->type)); |
| return xnn_status_invalid_parameter; |
| } |
| add_op->state = xnn_run_state_invalid; |
| |
| if (!xnn_params.initialized) { |
| xnn_log_error("failed to setup %s operator: XNNPACK is not initialized", |
| xnn_operator_type_to_string(xnn_operator_type_add_nc_q8)); |
| return xnn_status_uninitialized; |
| } |
| |
| if (batch_size == 0) { |
| add_op->state = xnn_run_state_skip; |
| return xnn_status_success; |
| } |
| |
| const size_t channels = add_op->channels; |
| const size_t a_stride = add_op->input_pixel_stride; |
| const size_t b_stride = add_op->input2_pixel_stride; |
| const size_t sum_stride = add_op->output_pixel_stride; |
| if ((((a_stride ^ channels) | (b_stride ^ channels) | (sum_stride ^ channels)) == 0) || batch_size == 1) { |
| const size_t block_size = 4096; |
| add_op->context.add_contiguous = (struct add_contiguous_context) { |
| .a = a, |
| .b = b, |
| .y = sum, |
| .params.q8 = add_op->params.q8_add, |
| .ukernel = xnn_params.q8.vadd, |
| }; |
| add_op->compute.type = xnn_parallelization_type_1d_tile_1d; |
| add_op->compute.task_1d_tile_1d = (pthreadpool_task_1d_tile_1d_t) xnn_compute_add_contiguous; |
| add_op->compute.range[0] = batch_size * channels * sizeof(uint8_t); |
| add_op->compute.tile[0] = block_size; |
| } else { |
| add_op->context.add_strided = (struct add_strided_context) { |
| .a = a, |
| .a_stride = a_stride * sizeof(uint8_t), |
| .b = b, |
| .b_stride = b_stride * sizeof(uint8_t), |
| .y = sum, |
| .y_stride = sum_stride * sizeof(uint8_t), |
| .n = channels, |
| .params.q8 = add_op->params.q8_add, |
| .ukernel = xnn_params.q8.vadd, |
| }; |
| add_op->compute.type = xnn_parallelization_type_1d_tile_1d; |
| add_op->compute.task_1d_tile_1d = (pthreadpool_task_1d_tile_1d_t) xnn_compute_add_strided; |
| add_op->compute.range[0] = batch_size; |
| add_op->compute.tile[0] = 1; |
| } |
| add_op->state = xnn_run_state_ready; |
| |
| return xnn_status_success; |
| } |
| |
| enum xnn_status xnn_setup_add_nc_f32( |
| xnn_operator_t add_op, |
| size_t batch_size, |
| const float* a, |
| const float* b, |
| float* sum, |
| pthreadpool_t threadpool) |
| { |
| if (add_op->type != xnn_operator_type_add_nc_f32) { |
| xnn_log_error("failed to setup operator: operator type mismatch (expected %s, got %s)", |
| xnn_operator_type_to_string(xnn_operator_type_add_nc_f32), |
| xnn_operator_type_to_string(add_op->type)); |
| return xnn_status_invalid_parameter; |
| } |
| add_op->state = xnn_run_state_invalid; |
| |
| if (!xnn_params.initialized) { |
| xnn_log_error("failed to setup %s operator: XNNPACK is not initialized", |
| xnn_operator_type_to_string(xnn_operator_type_add_nc_f32)); |
| return xnn_status_uninitialized; |
| } |
| |
| if (batch_size == 0) { |
| add_op->state = xnn_run_state_skip; |
| return xnn_status_success; |
| } |
| |
| const size_t channels = add_op->channels; |
| const size_t a_stride = add_op->input_pixel_stride; |
| const size_t b_stride = add_op->input2_pixel_stride; |
| const size_t sum_stride = add_op->output_pixel_stride; |
| if ((((a_stride ^ channels) | (b_stride ^ channels) | (sum_stride ^ channels)) == 0) || batch_size == 1) { |
| const size_t block_size = 4096; |
| add_op->context.add_contiguous = (struct add_contiguous_context) { |
| .a = a, |
| .b = b, |
| .y = sum, |
| .params.f32 = add_op->params.f32_minmax, |
| .ukernel = xnn_params.f32.vadd.op_ukernel, |
| }; |
| add_op->compute.type = xnn_parallelization_type_1d_tile_1d; |
| add_op->compute.task_1d_tile_1d = (pthreadpool_task_1d_tile_1d_t) xnn_compute_add_contiguous; |
| add_op->compute.range[0] = batch_size * channels * sizeof(float); |
| add_op->compute.tile[0] = block_size; |
| } else { |
| add_op->context.add_strided = (struct add_strided_context) { |
| .a = a, |
| .a_stride = a_stride * sizeof(float), |
| .b = b, |
| .b_stride = b_stride * sizeof(float), |
| .y = sum, |
| .y_stride = sum_stride * sizeof(float), |
| .n = channels * sizeof(float), |
| .params.f32 = add_op->params.f32_minmax, |
| .ukernel = xnn_params.f32.vadd.op_ukernel, |
| }; |
| add_op->compute.type = xnn_parallelization_type_1d_tile_1d; |
| add_op->compute.task_1d_tile_1d = (pthreadpool_task_1d_tile_1d_t) xnn_compute_add_strided; |
| add_op->compute.range[0] = batch_size; |
| add_op->compute.tile[0] = 1; |
| } |
| add_op->state = xnn_run_state_ready; |
| |
| return xnn_status_success; |
| } |