| // 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 <math.h> |
| #include <stdbool.h> |
| #include <stddef.h> |
| #include <stdint.h> |
| #include <string.h> |
| |
| #ifdef _WIN32 |
| #include <windows.h> |
| #else |
| #include <pthread.h> |
| #endif |
| |
| #ifdef _MSC_VER |
| #include <intrin.h> |
| #endif |
| |
| #ifndef __EMSCRIPTEN__ |
| #include <cpuinfo.h> |
| #endif |
| |
| #include <xnnpack.h> |
| #include <xnnpack/allocator.h> |
| #include <xnnpack/argmaxpool.h> |
| #include <xnnpack/avgpool.h> |
| #include <xnnpack/common.h> |
| #include <xnnpack/conv.h> |
| #include <xnnpack/dwconv.h> |
| #include <xnnpack/depthtospace.h> |
| #include <xnnpack/gavgpool.h> |
| #include <xnnpack/gemm.h> |
| #include <xnnpack/fill.h> |
| #include <xnnpack/ibilinear.h> |
| #include <xnnpack/igemm.h> |
| #include <xnnpack/log.h> |
| #include <xnnpack/lut.h> |
| #include <xnnpack/maxpool.h> |
| #include <xnnpack/pad.h> |
| #include <xnnpack/params.h> |
| #include <xnnpack/params-init.h> |
| #include <xnnpack/pavgpool.h> |
| #include <xnnpack/prelu.h> |
| #include <xnnpack/raddstoreexpminusmax.h> |
| #include <xnnpack/rmax.h> |
| #include <xnnpack/spmm.h> |
| #include <xnnpack/unpool.h> |
| #include <xnnpack/vaddsub.h> |
| #include <xnnpack/vbinary.h> |
| #include <xnnpack/vcvt.h> |
| #include <xnnpack/vmul.h> |
| #include <xnnpack/vmulcaddc.h> |
| #include <xnnpack/vunary.h> |
| #include <xnnpack/zip.h> |
| |
| #ifndef XNN_ENABLE_ASSEMBLY |
| #define XNN_ENABLE_ASSEMBLY 1 |
| #endif |
| |
| #ifdef _WIN32 |
| static INIT_ONCE init_guard = INIT_ONCE_STATIC_INIT; |
| #else |
| static pthread_once_t init_guard = PTHREAD_ONCE_INIT; |
| #endif |
| |
| static const struct xnn_allocator* volatile init_allocator = NULL; |
| |
| struct xnn_parameters xnn_params = { |
| .init_flags = 0 |
| }; |
| |
| static void init(void) { |
| #if XNN_ARCH_WASM || XNN_ARCH_WASMSIMD |
| // Unlike most other architectures, on x86/x86-64 when floating-point instructions |
| // have no NaN arguments, but produce NaN output, the output NaN has sign bit set. |
| // We use it to distinguish x86/x86-64 from other architectures, by doing subtraction |
| // of two infinities (must produce NaN per IEEE 754 standard). |
| static const volatile float inf = INFINITY; |
| const bool is_wasm_x86 = signbit(inf - inf); |
| #endif |
| uint32_t init_flags = XNN_INIT_FLAG_XNNPACK; |
| |
| #if XNN_ARCH_ARM |
| #if XNN_PLATFORM_MOBILE |
| if (!cpuinfo_has_arm_neon()) { |
| xnn_log_error("XNNPACK initialization failed: NEON is not supported"); |
| return; |
| } |
| #else |
| if (!cpuinfo_has_arm_vfpv2() && !cpuinfo_has_arm_vfpv3()) { |
| xnn_log_error("XNNPACK initialization failed: VFP is not supported"); |
| return; |
| } |
| #endif |
| |
| if (cpuinfo_has_arm_neon()) { |
| /**************************** QC8 micro-kernels ****************************/ |
| #ifndef XNN_NO_QC8_OPERATORS |
| init_flags |= XNN_INIT_FLAG_QC8; |
| |
| if (!XNN_PLATFORM_IOS && cpuinfo_has_arm_neon_dot()) { |
| xnn_params.qc8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_4x8c4__neondot); |
| xnn_params.qc8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_1x8c4__neondot); |
| xnn_params.qc8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_4x8c4__neondot); |
| xnn_params.qc8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_1x8c4__neondot); |
| xnn_params.qc8.gemm.init.qc8 = xnn_init_qs8_minmax_neon_params; |
| xnn_params.qc8.gemm.mr = 4; |
| xnn_params.qc8.gemm.nr = 8; |
| xnn_params.qc8.gemm.log2_kr = 2; |
| } else { |
| xnn_params.qc8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_2x8c2__neon_mlal_padal_dup); |
| xnn_params.qc8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_2x8c2__neon_mlal_padal_dup); |
| xnn_params.qc8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_1x8c2__neon_mlal_padal_dup); |
| xnn_params.qc8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_1x8c2__neon_mlal_padal_dup); |
| xnn_params.qc8.gemm.init.qc8 = xnn_init_qs8_minmax_neon_fp32_params; |
| xnn_params.qc8.gemm.mr = 2; |
| xnn_params.qc8.gemm.nr = 8; |
| xnn_params.qc8.gemm.log2_kr = 1; |
| } |
| |
| xnn_params.qc8.dwconv[0].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qc8_dwconv_minmax_fp32_ukernel_up16x9__neon_mla8_ld64; |
| xnn_params.qc8.dwconv[0].init.qc8 = xnn_init_qs8_minmax_neon_fp32_params; |
| xnn_params.qc8.dwconv[0].channel_tile = 16; |
| xnn_params.qc8.dwconv[0].primary_tile = 9; |
| xnn_params.qc8.dwconv[1].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qc8_dwconv_minmax_fp32_ukernel_up8x25__neon_mla8_ld64; |
| xnn_params.qc8.dwconv[1].init.qc8 = xnn_init_qs8_minmax_neon_fp32_params; |
| xnn_params.qc8.dwconv[1].channel_tile = 8; |
| xnn_params.qc8.dwconv[1].primary_tile = 25; |
| #endif // XNN_NO_QC8_OPERATORS |
| |
| /**************************** QS8 micro-kernels ****************************/ |
| #ifndef XNN_NO_QS8_OPERATORS |
| init_flags |= XNN_INIT_FLAG_QS8; |
| |
| if (!XNN_PLATFORM_IOS && cpuinfo_has_arm_neon_dot()) { |
| xnn_params.qs8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_rndnu_ukernel_4x8c4__neondot); |
| xnn_params.qs8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_rndnu_ukernel_1x8c4__neondot); |
| xnn_params.qs8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_rndnu_ukernel_4x8c4__neondot); |
| xnn_params.qs8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_rndnu_ukernel_1x8c4__neondot); |
| xnn_params.qs8.gemm.init.qs8 = xnn_init_qs8_conv_minmax_rndnu_neon_params; |
| xnn_params.qs8.gemm.mr = 4; |
| xnn_params.qs8.gemm.nr = 8; |
| xnn_params.qs8.gemm.log2_kr = 2; |
| } else { |
| xnn_params.qs8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_rndnu_ukernel_2x8s4c2__neon_mlal_padal); |
| xnn_params.qs8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_rndnu_ukernel_2x8s4c2__neon_mlal_padal); |
| xnn_params.qs8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_rndnu_ukernel_1x8s4c2__neon_mlal_padal); |
| xnn_params.qs8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_rndnu_ukernel_1x8s4c2__neon_mlal_padal); |
| xnn_params.qs8.gemm.init.qs8 = xnn_init_qs8_conv_minmax_rndnu_neon_params; |
| xnn_params.qs8.gemm.mr = 2; |
| xnn_params.qs8.gemm.nr = 8; |
| xnn_params.qs8.gemm.log2_kr = 1; |
| xnn_params.qs8.gemm.log2_sr = 2; |
| } |
| |
| xnn_params.qs8.dwconv[0].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qs8_dwconv_minmax_rndnu_ukernel_up16x9__neon_mla8_ld64; |
| xnn_params.qs8.dwconv[0].init.qs8 = xnn_init_qs8_conv_minmax_rndnu_neon_params; |
| xnn_params.qs8.dwconv[0].channel_tile = 16; |
| xnn_params.qs8.dwconv[0].primary_tile = 9; |
| xnn_params.qs8.dwconv[1].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qs8_dwconv_minmax_rndnu_ukernel_up8x25__neon_mla8_ld64; |
| xnn_params.qs8.dwconv[1].init.qs8 = xnn_init_qs8_conv_minmax_rndnu_neon_params; |
| xnn_params.qs8.dwconv[1].channel_tile = 8; |
| xnn_params.qs8.dwconv[1].primary_tile = 25; |
| |
| xnn_params.qs8.gavgpool = (struct gavgpool_parameters) { |
| .up = (xnn_gavgpool_unipass_ukernel_function) xnn_qs8_gavgpool_minmax_ukernel_7x__neon_c8_acc2, |
| .mp = (xnn_gavgpool_multipass_ukernel_function) xnn_qs8_gavgpool_minmax_ukernel_7p7x__neon_c8_acc2, |
| .mr = 7, |
| }; |
| |
| xnn_params.qs8.vadd = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vadd_minmax_ukernel__neon_ld64_x16, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vaddc_minmax_ukernel__neon_ld64_x16, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vaddc_minmax_ukernel__neon_ld64_x16, |
| .init.qs8_addsub = xnn_init_qs8_add_minmax_neon_params, |
| .element_tile = 16, |
| }; |
| if (cpuinfo_has_arm_neon_v8()) { |
| xnn_params.qs8.vmul = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vmul_minmax_fp32_ukernel__neonv8_ld64_x16, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vmulc_minmax_fp32_ukernel__neonv8_ld64_x16, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vmulc_minmax_fp32_ukernel__neonv8_ld64_x16, |
| .init.qs8_mul = xnn_init_qs8_mul_minmax_fp32_neonv8_params, |
| .element_tile = 16, |
| }; |
| } else { |
| xnn_params.qs8.vmul = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vmul_minmax_fp32_ukernel__neon_ld64_x16, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vmulc_minmax_fp32_ukernel__neon_ld64_x16, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vmulc_minmax_fp32_ukernel__neon_ld64_x16, |
| .init.qs8_mul = xnn_init_qs8_mul_minmax_fp32_neon_params, |
| .element_tile = 16, |
| }; |
| } |
| #endif // XNN_NO_QS8_OPERATORS |
| |
| /*************************** QU8 micro-kernels ***************************/ |
| #ifndef XNN_NO_QU8_OPERATORS |
| init_flags |= XNN_INIT_FLAG_QU8; |
| |
| if (cpuinfo_has_arm_neon_dot()) { |
| xnn_params.qu8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qu8_gemm_minmax_rndnu_ukernel_2x16c4__neondot); |
| xnn_params.qu8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qu8_igemm_minmax_rndnu_ukernel_2x16c4__neondot); |
| xnn_params.qu8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qu8_gemm_minmax_rndnu_ukernel_1x16c4__neondot); |
| xnn_params.qu8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qu8_igemm_minmax_rndnu_ukernel_1x16c4__neondot); |
| xnn_params.qu8.gemm.init.qu8 = xnn_init_qu8_conv_minmax_rndnu_neon_params; |
| xnn_params.qu8.gemm.mr = 2; |
| xnn_params.qu8.gemm.nr = 16; |
| xnn_params.qu8.gemm.log2_kr = 2; |
| } else { |
| xnn_params.qu8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qu8_gemm_minmax_rndnu_ukernel_4x8__neon_mlal_lane); |
| xnn_params.qu8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qu8_igemm_minmax_rndnu_ukernel_4x8__neon_mlal_lane); |
| xnn_params.qu8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qu8_gemm_minmax_rndnu_ukernel_1x8__neon_mlal_lane); |
| xnn_params.qu8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qu8_igemm_minmax_rndnu_ukernel_1x8__neon_mlal_lane); |
| xnn_params.qu8.gemm.init.qu8 = xnn_init_qu8_conv_minmax_rndnu_neon_params; |
| xnn_params.qu8.gemm.mr = 4; |
| xnn_params.qu8.gemm.nr = 8; |
| } |
| xnn_params.qu8.dwconv[0].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qu8_dwconv_minmax_rndnu_ukernel_up16x9__neon_mul8; |
| xnn_params.qu8.dwconv[0].init.qu8 = xnn_init_qu8_conv_minmax_rndnu_neon_params; |
| xnn_params.qu8.dwconv[0].channel_tile = 16; |
| xnn_params.qu8.dwconv[0].primary_tile = 9; |
| xnn_params.qu8.dwconv[1].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qu8_dwconv_minmax_rndnu_ukernel_up8x25__neon_mul8; |
| xnn_params.qu8.dwconv[1].init.qu8 = xnn_init_qu8_conv_minmax_rndnu_neon_params; |
| xnn_params.qu8.dwconv[1].channel_tile = 8; |
| xnn_params.qu8.dwconv[1].primary_tile = 25; |
| |
| xnn_params.qu8.avgpool = (struct avgpool_parameters) { |
| .up = (xnn_avgpool_unipass_ukernel_function) xnn_qu8_avgpool_minmax_ukernel_9x__neon_c8, |
| .mp = (xnn_avgpool_multipass_ukernel_function) xnn_qu8_avgpool_minmax_ukernel_9p8x__neon_c8, |
| .mr = 9, |
| .qr = 8, |
| }; |
| xnn_params.qu8.gavgpool = (struct gavgpool_parameters) { |
| .up = (xnn_gavgpool_unipass_ukernel_function) xnn_qu8_gavgpool_minmax_ukernel_7x__neon_c8, |
| .mp = (xnn_gavgpool_multipass_ukernel_function) xnn_qu8_gavgpool_minmax_ukernel_7p7x__neon_c8, |
| .mr = 7, |
| }; |
| xnn_params.qu8.vadd = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vadd_minmax_ukernel__neon_ld64_x16, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vaddc_minmax_ukernel__neon_ld64_x16, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vaddc_minmax_ukernel__neon_ld64_x16, |
| .init.qu8_addsub = xnn_init_qu8_add_minmax_neon_params, |
| .element_tile = 8, |
| }; |
| if (cpuinfo_has_arm_neon_v8()) { |
| xnn_params.qu8.vmul = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vmul_minmax_fp32_ukernel__neonv8_ld64_x16, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vmulc_minmax_fp32_ukernel__neonv8_ld64_x16, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vmulc_minmax_fp32_ukernel__neonv8_ld64_x16, |
| .init.qu8_mul = xnn_init_qu8_mul_minmax_fp32_neonv8_params, |
| .element_tile = 16, |
| }; |
| } else { |
| xnn_params.qu8.vmul = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vmul_minmax_fp32_ukernel__neon_ld64_x16, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vmulc_minmax_fp32_ukernel__neon_ld64_x16, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vmulc_minmax_fp32_ukernel__neon_ld64_x16, |
| .init.qu8_mul = xnn_init_qu8_mul_minmax_fp32_neon_params, |
| .element_tile = 16, |
| }; |
| } |
| #endif // XNN_NO_QU8_OPERATORS |
| |
| /**************************** S8 micro-kernels ****************************/ |
| #ifndef XNN_NO_S8_OPERATORS |
| init_flags |= XNN_INIT_FLAG_S8; |
| |
| xnn_params.s8.clamp = (struct vunary_parameters) { |
| .ukernel = (xnn_univector_ukernel_function) xnn_s8_vclamp_ukernel__neon_x64, |
| .init.s8_minmax = xnn_init_s8_minmax_neon_params, |
| .element_tile = 64, |
| }; |
| xnn_params.s8.maxpool = (struct maxpool_parameters) { |
| .ukernel = (xnn_maxpool_ukernel_function) xnn_s8_maxpool_minmax_ukernel_9p8x__neon_c16, |
| .init.s8 = xnn_init_s8_minmax_neon_params, |
| .mr = 9, |
| .qr = 8, |
| }; |
| #endif // XNN_NO_S8_OPERATORS |
| |
| /**************************** U8 micro-kernels ****************************/ |
| #ifndef XNN_NO_U8_OPERATORS |
| init_flags |= XNN_INIT_FLAG_U8; |
| |
| xnn_params.u8.clamp = (struct vunary_parameters) { |
| .ukernel = (xnn_univector_ukernel_function) xnn_u8_vclamp_ukernel__neon_x64, |
| .init.u8_minmax = xnn_init_u8_minmax_neon_params, |
| .element_tile = 64, |
| }; |
| xnn_params.u8.maxpool = (struct maxpool_parameters) { |
| .ukernel = (xnn_maxpool_ukernel_function) xnn_u8_maxpool_minmax_ukernel_9p8x__neon_c16, |
| .init.u8 = xnn_init_u8_minmax_neon_params, |
| .mr = 9, |
| .qr = 8, |
| }; |
| xnn_params.u8.rmax = xnn_u8_rmax_ukernel__neon; |
| xnn_params.u8.lut32norm = xnn_u8_lut32norm_ukernel__scalar; |
| #endif // XNN_NO_U8_OPERATORS |
| |
| /**************************** X8 micro-kernels ****************************/ |
| #ifndef XNN_NO_X8_OPERATORS |
| init_flags |= XNN_INIT_FLAG_X8; |
| |
| xnn_params.x8.lut = xnn_x8_lut_ukernel__scalar_x4; |
| xnn_params.x8.zip = (struct zip_parameters) { |
| .x2 = (xnn_zipc_ukernel_function) xnn_x8_zip_x2_ukernel__neon, |
| .x3 = (xnn_zipc_ukernel_function) xnn_x8_zip_x3_ukernel__neon, |
| .x4 = (xnn_zipc_ukernel_function) xnn_x8_zip_x4_ukernel__neon, |
| .xm = (xnn_zipv_ukernel_function) xnn_x8_zip_xm_ukernel__neon, |
| }; |
| #endif // XNN_NO_X8_OPERATORS |
| |
| /**************************** F32 micro-kernels ****************************/ |
| #ifndef XNN_NO_F32_OPERATORS |
| init_flags |= XNN_INIT_FLAG_F32; |
| |
| #if XNN_ENABLE_ASSEMBLY |
| switch (cpuinfo_get_uarch(0)->uarch) { |
| case cpuinfo_uarch_cortex_a5: |
| case cpuinfo_uarch_cortex_a7: |
| xnn_params.f32.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_4x8__aarch32_neon_cortex_a7); |
| xnn_params.f32.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_4x8__aarch32_neon_cortex_a7); |
| xnn_params.f32.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_1x8__neon_lane_ld64); |
| xnn_params.f32.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_1x8__neon_lane_ld64); |
| xnn_params.f32.gemm.init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.gemm.mr = 4; |
| xnn_params.f32.gemm.nr = 8; |
| break; |
| |
| case cpuinfo_uarch_cortex_a53: |
| case cpuinfo_uarch_cortex_a55r0: |
| xnn_params.f32.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_4x8__aarch32_neon_cortex_a53); |
| xnn_params.f32.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_4x8__aarch32_neon_cortex_a53); |
| xnn_params.f32.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_1x8__neon_lane_ld64); |
| xnn_params.f32.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_1x8__neon_lane_ld64); |
| xnn_params.f32.gemm.init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.gemm.mr = 4; |
| xnn_params.f32.gemm.nr = 8; |
| break; |
| |
| case cpuinfo_uarch_cortex_a55: |
| xnn_params.f32.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_4x8__aarch32_neon_cortex_a55); |
| xnn_params.f32.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_4x8__aarch32_neon_cortex_a55); |
| xnn_params.f32.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_1x8__neon_lane_ld64); |
| xnn_params.f32.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_1x8__neon_lane_ld64); |
| xnn_params.f32.gemm.init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.gemm.mr = 4; |
| xnn_params.f32.gemm.nr = 8; |
| break; |
| |
| case cpuinfo_uarch_cortex_a57: |
| case cpuinfo_uarch_cortex_a72: |
| case cpuinfo_uarch_cortex_a73: |
| xnn_params.f32.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_4x8__aarch32_neon_pld_cortex_a75); |
| xnn_params.f32.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_4x8__aarch32_neon_pld_cortex_a75); |
| xnn_params.f32.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_1x8__neon_lane_ld64); |
| xnn_params.f32.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_1x8__neon_lane_ld64); |
| xnn_params.f32.gemm.init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.gemm.mr = 4; |
| xnn_params.f32.gemm.nr = 8; |
| break; |
| |
| case cpuinfo_uarch_krait: |
| default: |
| xnn_params.f32.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_4x8__aarch32_neon_cortex_a75); |
| xnn_params.f32.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_4x8__aarch32_neon_cortex_a75); |
| xnn_params.f32.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_1x8__neon_lane_ld64); |
| xnn_params.f32.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_1x8__neon_lane_ld64); |
| xnn_params.f32.gemm.init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.gemm.mr = 4; |
| xnn_params.f32.gemm.nr = 8; |
| break; |
| } |
| #if XNN_MAX_UARCH_TYPES > 1 |
| { |
| /* Choose micro-kernels for little cores according to micro-kernel specification for the big core */ |
| const uint32_t mr = xnn_params.f32.gemm.mr; |
| const uint32_t nr = xnn_params.f32.gemm.nr; |
| const uint32_t log2_sr = xnn_params.f32.gemm.log2_sr; |
| for (size_t i = 1; i < XNN_MAX_UARCH_TYPES; i++) { |
| const struct cpuinfo_uarch_info* uarch_info = cpuinfo_get_uarch(i); |
| if (uarch_info == NULL) { |
| /* No more microarchitectures in the system */ |
| break; |
| } |
| |
| switch (uarch_info->uarch) { |
| case cpuinfo_uarch_cortex_a53: |
| case cpuinfo_uarch_cortex_a55r0: |
| if (mr == 4 && nr == 8 && log2_sr == 0) { |
| xnn_params.f32.gemm.minmax.gemm.function[i] = (xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_4x8__aarch32_neon_cortex_a53; |
| xnn_params.f32.gemm.minmax.igemm.function[i] = (xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_4x8__aarch32_neon_cortex_a53; |
| xnn_params.f32.gemm.minmax.gemm1.function[i] = (xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_1x8__neon_lane_ld64; |
| xnn_params.f32.gemm.minmax.igemm1.function[i] = (xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_1x8__neon_lane_ld64; |
| } |
| break; |
| case cpuinfo_uarch_cortex_a55: |
| if (mr == 4 && nr == 8 && log2_sr == 0) { |
| xnn_params.f32.gemm.minmax.gemm.function[i] = (xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_4x8__aarch32_neon_cortex_a55; |
| xnn_params.f32.gemm.minmax.igemm.function[i] = (xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_4x8__aarch32_neon_cortex_a55; |
| xnn_params.f32.gemm.minmax.gemm1.function[i] = (xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_1x8__neon_lane_ld64; |
| xnn_params.f32.gemm.minmax.igemm1.function[i] = (xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_1x8__neon_lane_ld64; |
| } |
| break; |
| default: |
| break; |
| } |
| } |
| } |
| #endif // XNN_MAX_UARCH_TYPES > 1 |
| #else // XNN_ENABLE_ASSEMBLY |
| xnn_params.f32.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_4x8__neon_lane_ld128); |
| xnn_params.f32.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_4x8__neon_lane_ld128); |
| xnn_params.f32.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_1x8__neon_lane_ld64); |
| xnn_params.f32.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_1x8__neon_lane_ld64); |
| xnn_params.f32.gemm.init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.gemm.mr = 4; |
| xnn_params.f32.gemm.nr = 8; |
| #endif // XNN_ENABLE_ASSEMBLY |
| xnn_params.f32.gemm2.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_4x2__neon_lane_ld64); |
| xnn_params.f32.gemm2.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_4x2__neon_lane_ld64); |
| xnn_params.f32.gemm2.init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.gemm2.mr = 4; |
| xnn_params.f32.gemm2.nr = 2; |
| |
| xnn_params.f32.dwconv[0].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_minmax_ukernel_up8x3__neon; |
| xnn_params.f32.dwconv[0].init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.dwconv[0].channel_tile = 8, |
| xnn_params.f32.dwconv[0].primary_tile = 3, |
| |
| xnn_params.f32.dwconv[1].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_minmax_ukernel_up8x4__neon; |
| xnn_params.f32.dwconv[1].init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.dwconv[1].channel_tile = 8, |
| xnn_params.f32.dwconv[1].primary_tile = 4, |
| |
| xnn_params.f32.dwconv[2].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_minmax_ukernel_up8x9__neon; |
| xnn_params.f32.dwconv[2].init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.dwconv[2].channel_tile = 8; |
| xnn_params.f32.dwconv[2].primary_tile = 9; |
| |
| xnn_params.f32.dwconv[3].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_minmax_ukernel_up8x25__neon_acc2; |
| xnn_params.f32.dwconv[3].init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.dwconv[3].channel_tile = 8; |
| xnn_params.f32.dwconv[3].primary_tile = 25; |
| |
| xnn_params.f32.avgpool = (struct avgpool_parameters) { |
| .up = (xnn_avgpool_unipass_ukernel_function) xnn_f32_avgpool_minmax_ukernel_9x__neon_c4, |
| .mp = (xnn_avgpool_multipass_ukernel_function) xnn_f32_avgpool_minmax_ukernel_9p8x__neon_c4, |
| .mr = 9, |
| .qr = 8, |
| }; |
| xnn_params.f32.pavgpool = (struct pavgpool_parameters) { |
| .up = (xnn_pavgpool_unipass_ukernel_function) xnn_f32_pavgpool_minmax_ukernel_9x__neon_c4, |
| .mp = (xnn_pavgpool_multipass_ukernel_function) xnn_f32_pavgpool_minmax_ukernel_9p8x__neon_c4, |
| .mr = 9, |
| .qr = 8, |
| }; |
| xnn_params.f32.gavgpool = (struct gavgpool_parameters) { |
| .up = (xnn_gavgpool_unipass_ukernel_function) xnn_f32_gavgpool_minmax_ukernel_7x__neon_c4, |
| .mp = (xnn_gavgpool_multipass_ukernel_function) xnn_f32_gavgpool_minmax_ukernel_7p7x__neon_c4, |
| .mr = 7, |
| }; |
| xnn_params.f32.maxpool = (struct maxpool_parameters) { |
| .ukernel = (xnn_maxpool_ukernel_function) xnn_f32_maxpool_minmax_ukernel_9p8x__neon_c4, |
| .init.f32 = xnn_init_f32_minmax_scalar_params, |
| .mr = 9, |
| .qr = 8, |
| }; |
| xnn_params.f32.argmaxpool[0] = (struct argmaxpool_parameters) { |
| .up = (xnn_argmaxpool_unipass_ukernel_function) xnn_f32_argmaxpool_ukernel_4x__neon_c4, |
| .mr = 4, |
| }; |
| xnn_params.f32.argmaxpool[1] = (struct argmaxpool_parameters) { |
| .up = (xnn_argmaxpool_unipass_ukernel_function) xnn_f32_argmaxpool_ukernel_9x__neon_c4, |
| .mr = 9, |
| }; |
| xnn_params.f32.argmaxpool[2] = (struct argmaxpool_parameters) { |
| .mp = (xnn_argmaxpool_multipass_ukernel_function) xnn_f32_argmaxpool_ukernel_9p8x__neon_c4, |
| .mr = 9, |
| .qr = 8, |
| }; |
| xnn_params.f32.ibilinear = (struct ibilinear_parameters) { |
| .ukernel = (xnn_ibilinear_ukernel_function) xnn_f32_ibilinear_ukernel__neon_c8, |
| .pixel_tile = 1, |
| .channel_tile = 8, |
| }; |
| xnn_params.f32.abs = (xnn_univector_ukernel_function) xnn_f32_vabs_ukernel__neon_x8; |
| xnn_params.f32.clamp = (struct vunary_parameters) { |
| .ukernel = (xnn_univector_ukernel_function) xnn_f32_vclamp_ukernel__neon_x8, |
| .init.f32_minmax = xnn_init_f32_minmax_scalar_params, |
| .element_tile = 8, |
| }; |
| if (cpuinfo_has_arm_neon_fma()) { |
| xnn_params.f32.elu = (xnn_univector_ukernel_function) xnn_f32_velu_ukernel__neonfma_rr1_p6_x8; |
| } else { |
| xnn_params.f32.elu = (xnn_univector_ukernel_function) xnn_f32_velu_ukernel__neon_rr2_lut16_p3_x8; |
| } |
| xnn_params.f32.hswish = (xnn_univector_ukernel_function) xnn_f32_vhswish_ukernel__neon_x16; |
| xnn_params.f32.lrelu = (xnn_univector_ukernel_function) xnn_f32_vlrelu_ukernel__neon_x8; |
| xnn_params.f32.neg = (xnn_univector_ukernel_function) xnn_f32_vneg_ukernel__neon_x8; |
| if (cpuinfo_has_arm_neon_v8()) { |
| xnn_params.f32.rndne = (xnn_univector_ukernel_function) xnn_f32_vrndne_ukernel__neonv8_x8; |
| xnn_params.f32.rndz = (xnn_univector_ukernel_function) xnn_f32_vrndz_ukernel__neonv8_x8; |
| xnn_params.f32.rndu = (xnn_univector_ukernel_function) xnn_f32_vrndu_ukernel__neonv8_x8; |
| xnn_params.f32.rndd = (xnn_univector_ukernel_function) xnn_f32_vrndd_ukernel__neonv8_x8; |
| } else { |
| xnn_params.f32.rndne = (xnn_univector_ukernel_function) xnn_f32_vrndne_ukernel__neon_x8; |
| xnn_params.f32.rndz = (xnn_univector_ukernel_function) xnn_f32_vrndz_ukernel__neon_x8; |
| xnn_params.f32.rndu = (xnn_univector_ukernel_function) xnn_f32_vrndu_ukernel__neon_x8; |
| xnn_params.f32.rndd = (xnn_univector_ukernel_function) xnn_f32_vrndd_ukernel__neon_x8; |
| } |
| xnn_params.f32.sigmoid = (xnn_univector_ukernel_function) xnn_f32_vsigmoid_ukernel__neon_rr2_lut64_p2_nr2recps_x8; |
| xnn_params.f32.sqr = (xnn_univector_ukernel_function) xnn_f32_vsqr_ukernel__neon_x8; |
| xnn_params.f32.sqrt = (xnn_univector_ukernel_function) xnn_f32_vsqrt_ukernel__scalar_sqrt_x1; |
| xnn_params.f32.prelu = (struct prelu_parameters) { |
| .ukernel = (xnn_prelu_ukernel_function) xnn_f32_prelu_ukernel__neon_2x8, |
| .row_tile = 2, |
| .channel_tile = 8, |
| }; |
| xnn_params.f32.raddstoreexpminusmax = xnn_f32_raddstoreexpminusmax_ukernel__neon_lut64_p2_x8; |
| xnn_params.f32.rmax = xnn_f32_rmax_ukernel__neon; |
| xnn_params.f32.vadd = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vadd_minmax_ukernel__neon_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vaddc_minmax_ukernel__neon_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vaddc_minmax_ukernel__neon_x8, |
| .element_tile = 8, |
| }; |
| xnn_params.f32.vdiv = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vdiv_minmax_ukernel__scalar_x2, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vdivc_minmax_ukernel__scalar_x2, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vrdivc_minmax_ukernel__scalar_x2, |
| .element_tile = 2, |
| }; |
| xnn_params.f32.vmax = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmax_ukernel__neon_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmaxc_ukernel__neon_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmaxc_ukernel__neon_x8, |
| .element_tile = 8, |
| }; |
| xnn_params.f32.vmin = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmin_ukernel__neon_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vminc_ukernel__neon_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vminc_ukernel__neon_x8, |
| .element_tile = 8, |
| }; |
| xnn_params.f32.vmul = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmul_minmax_ukernel__neon_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmulc_minmax_ukernel__neon_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmulc_minmax_ukernel__neon_x8, |
| .element_tile = 8, |
| }; |
| xnn_params.f32.vsub = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsub_minmax_ukernel__neon_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsubc_minmax_ukernel__neon_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vrsubc_minmax_ukernel__neon_x8, |
| .element_tile = 8, |
| }; |
| xnn_params.f32.vsqrdiff = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsqrdiff_ukernel__neon_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsqrdiffc_ukernel__neon_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsqrdiffc_ukernel__neon_x8, |
| .element_tile = 8, |
| }; |
| xnn_params.f32.vmulcaddc = (struct vmulcaddc_parameters) { |
| .ukernel = (xnn_vmulcaddc_ukernel_function) xnn_f32_vmulcaddc_minmax_ukernel_c4__neon_2x, |
| .init.f32 = xnn_init_f32_minmax_scalar_params, |
| .channel_tile = 4, |
| .row_tile = 2, |
| }; |
| #ifndef XNN_NO_NCHW_OPERATORS |
| init_flags |= XNN_INIT_FLAG_CHW_OPT; |
| |
| xnn_params.f32.spmm = (struct spmm_parameters) { |
| .ukernel = (xnn_spmm_ukernel_function) xnn_f32_spmm_minmax_ukernel_32x1__neon, |
| .mr = 32, |
| .nr = 1, |
| }; |
| xnn_params.f32.conv_hwc2chw_3x3c3s2 = (struct conv_hwc2chw_parameters) { |
| .ukernel_with_symm_padding = |
| (xnn_conv_hwc2chw_ukernel_function) xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__neon_2x2, |
| .output_channel_tile = 4, |
| .output_height_tile = 2, |
| .output_width_tile = 2, |
| }; |
| xnn_params.f32.dwconv2d_chw_3x3 = (struct dwconv2d_chw_parameters) { |
| .ukernel = (xnn_dwconv2d_chw_ukernel_function) xnn_f32_dwconv2d_chw_ukernel_3x3p1__neon_2x4, |
| .output_width_tile = 4, |
| .output_height_tile = 2, |
| }; |
| xnn_params.f32.dwconv2d_chw_3x3s2 = (struct dwconv2d_chw_parameters) { |
| .ukernel = (xnn_dwconv2d_chw_ukernel_function) xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__neon_1x4, |
| .output_width_tile = 4, |
| .output_height_tile = 1, |
| }; |
| xnn_params.f32.dwconv2d_chw_5x5 = (struct dwconv2d_chw_parameters) { |
| .ukernel = (xnn_dwconv2d_chw_ukernel_function) xnn_f32_dwconv2d_chw_ukernel_5x5p2__neon_1x4, |
| .output_width_tile = 4, |
| .output_height_tile = 1, |
| }; |
| xnn_params.f32.dwconv2d_chw_5x5s2 = (struct dwconv2d_chw_parameters) { |
| .ukernel = (xnn_dwconv2d_chw_ukernel_function) xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neon_1x4, |
| .output_width_tile = 4, |
| .output_height_tile = 1, |
| }; |
| xnn_params.f32.gavgpool_cw = (struct gavgpool_cw_parameters) { |
| .ukernel = (xnn_gavgpool_cw_ukernel_function) xnn_f32_gavgpool_cw_ukernel__neon_x4, |
| .channel_tile = 4, |
| }; |
| xnn_params.f32.ibilinear_chw = (struct ibilinear_chw_parameters) { |
| .ukernel = (xnn_ibilinear_chw_ukernel_function) xnn_f32_ibilinear_chw_ukernel__neon_p8, |
| .channel_tile = 1, |
| .pixel_tile = 8, |
| }; |
| #endif // XNN_NO_NCHW_OPERATORS |
| #endif // XNN_NO_F32_OPERATORS |
| |
| /*************************** VCVT micro-kernels ***************************/ |
| #ifndef XNN_NO_VCVT_OPERATORS |
| init_flags |= XNN_INIT_FLAG_VCVT; |
| |
| if (cpuinfo_has_arm_neon_fp16()) { |
| xnn_params.vcvt.f16_to_f32 = (xnn_univector_ukernel_function) xnn_f16_f32_vcvt_ukernel__neonfp16_x16; |
| } else { |
| xnn_params.vcvt.f16_to_f32 = (xnn_univector_ukernel_function) xnn_f16_f32_vcvt_ukernel__neon_int16_x16; |
| } |
| #endif // XNN_NO_VCVT_OPERATORS |
| |
| /**************************** X32 micro-kernels ****************************/ |
| #ifndef XNN_NO_X32_OPERATORS |
| init_flags |= XNN_INIT_FLAG_X32; |
| |
| xnn_params.x32.unpool = (xnn_unpool_ukernel_function) xnn_x32_unpool_ukernel__neon; |
| xnn_params.x32.zip = (struct zip_parameters) { |
| .x2 = (xnn_zipc_ukernel_function) xnn_x32_zip_x2_ukernel__neon, |
| .x3 = (xnn_zipc_ukernel_function) xnn_x32_zip_x3_ukernel__neon, |
| .x4 = (xnn_zipc_ukernel_function) xnn_x32_zip_x4_ukernel__neon, |
| .xm = (xnn_zipv_ukernel_function) xnn_x32_zip_xm_ukernel__neon, |
| }; |
| #ifndef XNN_NO_NCHW_OPERATORS |
| xnn_params.x32.depthtospace2d_chw2hwc = (struct depthtospace2d_chw2hwc_parameters) { |
| .ukernel = (xnn_depthtospace2d_chw2hwc_ukernel_function) xnn_x32_depthtospace2d_chw2hwc_ukernel__scalar, |
| .channel_tile = 1, |
| .pixel_tile = 1, |
| }; |
| #endif // XNN_NO_NCHW_OPERATORS |
| #endif // XNN_NO_X32_OPERATORS |
| |
| /**************************** XX micro-kernels ****************************/ |
| #ifndef XNN_NO_XX_OPERATORS |
| init_flags |= XNN_INIT_FLAG_XX; |
| |
| xnn_params.xx.copy = (xnn_univector_ukernel_function) xnn_xx_copy_ukernel__memcpy; |
| xnn_params.xx.fill = (struct fill_parameters) { |
| .ukernel = (xnn_fill_ukernel_function) xnn_xx_fill_ukernel__neon_x64, |
| .row_tile = 1, |
| }; |
| xnn_params.xx.pad = (struct pad_parameters) { |
| .ukernel = (xnn_pad_ukernel_function) xnn_xx_pad_ukernel__neon, |
| .row_tile = 1, |
| }; |
| #endif // XNN_NO_XX_OPERATORS |
| |
| } else if (!XNN_PLATFORM_MOBILE) { |
| |
| /*************************** QS8 micro-kernels ***************************/ |
| #ifndef XNN_NO_QS8_OPERATORS |
| init_flags |= XNN_INIT_FLAG_QS8; |
| |
| xnn_params.qs8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_fp32_ukernel_2x2__scalar_magic); |
| xnn_params.qs8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_fp32_ukernel_2x2__scalar_magic); |
| xnn_params.qs8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_fp32_ukernel_1x2__scalar_magic); |
| xnn_params.qs8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_fp32_ukernel_1x2__scalar_magic); |
| xnn_params.qs8.gemm.init.qs8 = xnn_init_qs8_conv_minmax_fp32_scalar_magic_params; |
| xnn_params.qs8.gemm.mr = 2; |
| xnn_params.qs8.gemm.nr = 2; |
| |
| xnn_params.qs8.dwconv[0].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qs8_dwconv_minmax_fp32_ukernel_up1x9__scalar_magic; |
| xnn_params.qs8.dwconv[0].init.qs8 = xnn_init_qs8_conv_minmax_fp32_scalar_magic_params; |
| xnn_params.qs8.dwconv[0].channel_tile = 1; |
| xnn_params.qs8.dwconv[0].primary_tile = 9; |
| xnn_params.qs8.dwconv[1].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qs8_dwconv_minmax_fp32_ukernel_up1x25__scalar_magic; |
| xnn_params.qs8.dwconv[1].init.qs8 = xnn_init_qs8_conv_minmax_fp32_scalar_magic_params; |
| xnn_params.qs8.dwconv[1].channel_tile = 1; |
| xnn_params.qs8.dwconv[1].primary_tile = 25; |
| |
| xnn_params.qs8.gavgpool = (struct gavgpool_parameters) { |
| .up = (xnn_gavgpool_unipass_ukernel_function) xnn_qs8_gavgpool_minmax_ukernel_7x__scalar_c1, |
| .mp = (xnn_gavgpool_multipass_ukernel_function) xnn_qs8_gavgpool_minmax_ukernel_7p7x__scalar_c1, |
| .mr = 7, |
| }; |
| xnn_params.qs8.vadd = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vadd_minmax_ukernel__scalar_x1, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vaddc_minmax_ukernel__scalar_x1, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vaddc_minmax_ukernel__scalar_x1, |
| .init.qs8_addsub = xnn_init_qs8_add_minmax_scalar_params, |
| .element_tile = 1, |
| }; |
| xnn_params.qs8.vmul = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vmul_minmax_fp32_ukernel__scalar_x4, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vmulc_minmax_fp32_ukernel__scalar_x4, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vmulc_minmax_fp32_ukernel__scalar_x4, |
| .init.qs8_mul = xnn_init_qs8_mul_minmax_fp32_scalar_params, |
| .element_tile = 4, |
| }; |
| #endif // XNN_NO_QS8_OPERATORS |
| |
| /*************************** QU8 micro-kernels ***************************/ |
| #ifndef XNN_NO_QU8_OPERATORS |
| init_flags |= XNN_INIT_FLAG_QU8; |
| |
| xnn_params.qu8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qu8_gemm_minmax_fp32_ukernel_2x2__scalar_magic); |
| xnn_params.qu8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qu8_igemm_minmax_fp32_ukernel_2x2__scalar_magic); |
| xnn_params.qu8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qu8_gemm_minmax_fp32_ukernel_1x2__scalar_magic); |
| xnn_params.qu8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qu8_igemm_minmax_fp32_ukernel_1x2__scalar_magic); |
| xnn_params.qu8.gemm.init.qu8 = xnn_init_qu8_conv_minmax_fp32_scalar_magic_params; |
| xnn_params.qu8.gemm.mr = 2; |
| xnn_params.qu8.gemm.nr = 2; |
| |
| xnn_params.qu8.dwconv[0].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qu8_dwconv_minmax_fp32_ukernel_up1x9__scalar_magic; |
| xnn_params.qu8.dwconv[0].init.qu8 = xnn_init_qu8_conv_minmax_fp32_scalar_magic_params; |
| xnn_params.qu8.dwconv[0].channel_tile = 1; |
| xnn_params.qu8.dwconv[0].primary_tile = 9; |
| xnn_params.qu8.dwconv[1].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qu8_dwconv_minmax_fp32_ukernel_up1x25__scalar_magic; |
| xnn_params.qu8.dwconv[1].init.qu8 = xnn_init_qu8_conv_minmax_fp32_scalar_magic_params; |
| xnn_params.qu8.dwconv[1].channel_tile = 1; |
| xnn_params.qu8.dwconv[1].primary_tile = 25; |
| |
| xnn_params.qu8.avgpool = (struct avgpool_parameters) { |
| .up = (xnn_avgpool_unipass_ukernel_function) xnn_qu8_avgpool_minmax_ukernel_9x__scalar_c1, |
| .mp = (xnn_avgpool_multipass_ukernel_function) xnn_qu8_avgpool_minmax_ukernel_9p8x__scalar_c1, |
| .mr = 9, |
| .qr = 8, |
| }; |
| xnn_params.qu8.gavgpool = (struct gavgpool_parameters) { |
| .up = (xnn_gavgpool_unipass_ukernel_function) xnn_qu8_gavgpool_minmax_ukernel_7x__scalar_c1, |
| .mp = (xnn_gavgpool_multipass_ukernel_function) xnn_qu8_gavgpool_minmax_ukernel_7p7x__scalar_c1, |
| .mr = 7, |
| }; |
| xnn_params.qu8.vadd = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vadd_minmax_ukernel__scalar_x1, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vaddc_minmax_ukernel__scalar_x1, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vaddc_minmax_ukernel__scalar_x1, |
| .init.qu8_addsub = xnn_init_qu8_add_minmax_scalar_params, |
| .element_tile = 1, |
| }; |
| xnn_params.qu8.vmul = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vmul_minmax_fp32_ukernel__scalar_x4, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vmulc_minmax_fp32_ukernel__scalar_x4, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vmulc_minmax_fp32_ukernel__scalar_x4, |
| .init.qu8_mul = xnn_init_qu8_mul_minmax_fp32_scalar_params, |
| .element_tile = 4, |
| }; |
| #endif // XNN_NO_QU8_OPERATORS |
| |
| /**************************** S8 micro-kernels ****************************/ |
| #ifndef XNN_NO_S8_OPERATORS |
| init_flags |= XNN_INIT_FLAG_S8; |
| |
| xnn_params.s8.clamp = (struct vunary_parameters) { |
| .ukernel = (xnn_univector_ukernel_function) xnn_s8_vclamp_ukernel__scalar_x4, |
| .init.s8_minmax = xnn_init_s8_minmax_scalar_params, |
| .element_tile = 4, |
| }; |
| xnn_params.s8.maxpool = (struct maxpool_parameters) { |
| .ukernel = (xnn_maxpool_ukernel_function) xnn_s8_maxpool_minmax_ukernel_9p8x__scalar_c1, |
| .init.s8 = xnn_init_s8_minmax_scalar_params, |
| .mr = 9, |
| .qr = 8, |
| }; |
| #endif // XNN_NO_S8_OPERATORS |
| |
| /**************************** U8 micro-kernels ****************************/ |
| #ifndef XNN_NO_U8_OPERATORS |
| init_flags |= XNN_INIT_FLAG_U8; |
| |
| xnn_params.u8.clamp = (struct vunary_parameters) { |
| .ukernel = (xnn_univector_ukernel_function) xnn_u8_vclamp_ukernel__scalar_x4, |
| .init.u8_minmax = xnn_init_u8_minmax_scalar_params, |
| .element_tile = 4, |
| }; |
| xnn_params.u8.maxpool = (struct maxpool_parameters) { |
| .ukernel = (xnn_maxpool_ukernel_function) xnn_u8_maxpool_minmax_ukernel_9p8x__scalar_c1, |
| .init.u8 = xnn_init_u8_minmax_scalar_params, |
| .mr = 9, |
| .qr = 8, |
| }; |
| xnn_params.u8.lut32norm = xnn_u8_lut32norm_ukernel__scalar; |
| xnn_params.u8.rmax = xnn_u8_rmax_ukernel__scalar; |
| #endif // XNN_NO_U8_OPERATORS |
| |
| /**************************** X8 micro-kernels ****************************/ |
| #ifndef XNN_NO_X8_OPERATORS |
| init_flags |= XNN_INIT_FLAG_X8; |
| |
| xnn_params.x8.lut = xnn_x8_lut_ukernel__scalar_x4; |
| xnn_params.x8.zip = (struct zip_parameters) { |
| .x2 = (xnn_zipc_ukernel_function) xnn_x8_zip_x2_ukernel__scalar, |
| .x3 = (xnn_zipc_ukernel_function) xnn_x8_zip_x3_ukernel__scalar, |
| .x4 = (xnn_zipc_ukernel_function) xnn_x8_zip_x4_ukernel__scalar, |
| .xm = (xnn_zipv_ukernel_function) xnn_x8_zip_xm_ukernel__scalar, |
| }; |
| #endif // XNN_NO_X8_OPERATORS |
| |
| /**************************** F32 micro-kernels ****************************/ |
| #ifndef XNN_NO_F32_OPERATORS |
| init_flags |= XNN_INIT_FLAG_F32; |
| |
| xnn_params.f32.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_4x4__scalar); |
| xnn_params.f32.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_4x4__scalar); |
| xnn_params.f32.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_1x4__scalar); |
| xnn_params.f32.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_1x4__scalar); |
| xnn_params.f32.gemm.relu.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_relu_ukernel_4x4__scalar); |
| xnn_params.f32.gemm.relu.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_relu_ukernel_4x4__scalar); |
| xnn_params.f32.gemm.relu.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_relu_ukernel_1x4__scalar); |
| xnn_params.f32.gemm.relu.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_relu_ukernel_1x4__scalar); |
| xnn_params.f32.gemm.linear.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_ukernel_4x4__scalar); |
| xnn_params.f32.gemm.linear.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_ukernel_4x4__scalar); |
| xnn_params.f32.gemm.linear.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_ukernel_1x4__scalar); |
| xnn_params.f32.gemm.linear.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_ukernel_1x4__scalar); |
| xnn_params.f32.gemm.init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.gemm.mr = 4; |
| xnn_params.f32.gemm.nr = 4; |
| |
| xnn_params.f32.gemm2.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_4x2__scalar); |
| xnn_params.f32.gemm2.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_4x2__scalar), |
| xnn_params.f32.gemm2.linear.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_ukernel_4x2__scalar); |
| xnn_params.f32.gemm2.linear.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_ukernel_4x2__scalar), |
| xnn_params.f32.gemm2.init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.gemm2.mr = 4; |
| xnn_params.f32.gemm2.nr = 2; |
| |
| xnn_params.f32.dwconv[0].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_minmax_ukernel_up1x3__scalar_acc2; |
| xnn_params.f32.dwconv[0].linear.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_ukernel_up1x3__scalar_acc2; |
| xnn_params.f32.dwconv[0].init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.dwconv[0].channel_tile = 1; |
| xnn_params.f32.dwconv[0].primary_tile = 3; |
| |
| xnn_params.f32.dwconv[1].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_minmax_ukernel_up1x4__scalar_acc2; |
| xnn_params.f32.dwconv[1].linear.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_ukernel_up1x4__scalar_acc2; |
| xnn_params.f32.dwconv[1].init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.dwconv[1].channel_tile = 1; |
| xnn_params.f32.dwconv[1].primary_tile = 4; |
| |
| xnn_params.f32.dwconv[2].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_minmax_ukernel_up1x9__scalar_acc2; |
| xnn_params.f32.dwconv[2].linear.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_ukernel_up1x9__scalar_acc2; |
| xnn_params.f32.dwconv[2].init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.dwconv[2].channel_tile = 1; |
| xnn_params.f32.dwconv[2].primary_tile = 9; |
| |
| xnn_params.f32.dwconv[3].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_minmax_ukernel_up1x25__scalar_acc2; |
| xnn_params.f32.dwconv[3].linear.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_ukernel_up1x25__scalar_acc2; |
| xnn_params.f32.dwconv[3].init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.dwconv[3].channel_tile = 1; |
| xnn_params.f32.dwconv[3].primary_tile = 25; |
| |
| xnn_params.f32.avgpool = (struct avgpool_parameters) { |
| .up = (xnn_avgpool_unipass_ukernel_function) xnn_f32_avgpool_minmax_ukernel_9x__scalar_c1, |
| .mp = (xnn_avgpool_multipass_ukernel_function) xnn_f32_avgpool_minmax_ukernel_9p8x__scalar_c1, |
| .mr = 9, |
| .qr = 8, |
| }; |
| xnn_params.f32.pavgpool = (struct pavgpool_parameters) { |
| .up = (xnn_pavgpool_unipass_ukernel_function) xnn_f32_pavgpool_minmax_ukernel_9x__scalar_c1, |
| .mp = (xnn_pavgpool_multipass_ukernel_function) xnn_f32_pavgpool_minmax_ukernel_9p8x__scalar_c1, |
| .mr = 9, |
| .qr = 8, |
| }; |
| xnn_params.f32.gavgpool = (struct gavgpool_parameters) { |
| .up = (xnn_gavgpool_unipass_ukernel_function) xnn_f32_gavgpool_minmax_ukernel_7x__scalar_c1, |
| .mp = (xnn_gavgpool_multipass_ukernel_function) xnn_f32_gavgpool_minmax_ukernel_7p7x__scalar_c1, |
| .mr = 7, |
| }; |
| xnn_params.f32.maxpool = (struct maxpool_parameters) { |
| .ukernel = (xnn_maxpool_ukernel_function) xnn_f32_maxpool_minmax_ukernel_9p8x__scalar_c1, |
| .init.f32 = xnn_init_f32_minmax_scalar_params, |
| .mr = 9, |
| .qr = 8, |
| }; |
| xnn_params.f32.argmaxpool[0] = (struct argmaxpool_parameters) { |
| .up = (xnn_argmaxpool_unipass_ukernel_function) xnn_f32_argmaxpool_ukernel_4x__scalar_c1, |
| .mr = 4, |
| }; |
| xnn_params.f32.argmaxpool[1] = (struct argmaxpool_parameters) { |
| .up = (xnn_argmaxpool_unipass_ukernel_function) xnn_f32_argmaxpool_ukernel_9x__scalar_c1, |
| .mr = 9, |
| }; |
| xnn_params.f32.argmaxpool[2] = (struct argmaxpool_parameters) { |
| .mp = (xnn_argmaxpool_multipass_ukernel_function) xnn_f32_argmaxpool_ukernel_9p8x__scalar_c1, |
| .mr = 9, |
| .qr = 8, |
| }; |
| xnn_params.f32.ibilinear = (struct ibilinear_parameters) { |
| .ukernel = (xnn_ibilinear_ukernel_function) xnn_f32_ibilinear_ukernel__scalar_c2, |
| .pixel_tile = 1, |
| .channel_tile = 2, |
| }; |
| xnn_params.f32.abs = (xnn_univector_ukernel_function) xnn_f32_vabs_ukernel__scalar_x4; |
| xnn_params.f32.clamp = (struct vunary_parameters) { |
| .ukernel = (xnn_univector_ukernel_function) xnn_f32_vclamp_ukernel__scalar_x4, |
| .init.f32_minmax = xnn_init_f32_minmax_scalar_params, |
| .element_tile = 4, |
| }; |
| xnn_params.f32.elu = (xnn_univector_ukernel_function) xnn_f32_velu_ukernel__scalar_rr2_lut16_p3_x4; |
| xnn_params.f32.hswish = (xnn_univector_ukernel_function) xnn_f32_vhswish_ukernel__scalar_x4; |
| xnn_params.f32.lrelu = (xnn_univector_ukernel_function) xnn_f32_vlrelu_ukernel__scalar_x4; |
| xnn_params.f32.neg = (xnn_univector_ukernel_function) xnn_f32_vneg_ukernel__scalar_x4; |
| xnn_params.f32.rndne = (xnn_univector_ukernel_function) xnn_f32_vrndne_ukernel__scalar_libm_x1; |
| xnn_params.f32.rndz = (xnn_univector_ukernel_function) xnn_f32_vrndz_ukernel__scalar_libm_x1; |
| xnn_params.f32.rndu = (xnn_univector_ukernel_function) xnn_f32_vrndu_ukernel__scalar_libm_x1; |
| xnn_params.f32.rndd = (xnn_univector_ukernel_function) xnn_f32_vrndd_ukernel__scalar_libm_x1; |
| xnn_params.f32.sigmoid = (xnn_univector_ukernel_function) xnn_f32_vsigmoid_ukernel__scalar_lut64_p2_div_x2; |
| xnn_params.f32.sqr = (xnn_univector_ukernel_function) xnn_f32_vsqr_ukernel__scalar_x4; |
| xnn_params.f32.sqrt = (xnn_univector_ukernel_function) xnn_f32_vsqrt_ukernel__scalar_sqrt_x1; |
| xnn_params.f32.prelu = (struct prelu_parameters) { |
| .ukernel = (xnn_prelu_ukernel_function) xnn_f32_prelu_ukernel__scalar_2x4, |
| .row_tile = 4, |
| .channel_tile = 4, |
| }; |
| xnn_params.f32.raddstoreexpminusmax = xnn_f32_raddstoreexpminusmax_ukernel__scalar_p5_x4_acc2; |
| xnn_params.f32.rmax = xnn_f32_rmax_ukernel__scalar; |
| xnn_params.f32.vadd = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vadd_minmax_ukernel__scalar_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vaddc_minmax_ukernel__scalar_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vaddc_minmax_ukernel__scalar_x8, |
| .element_tile = 8, |
| }; |
| xnn_params.f32.vdiv = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vdiv_minmax_ukernel__scalar_x2, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vdivc_minmax_ukernel__scalar_x2, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vrdivc_minmax_ukernel__scalar_x2, |
| .element_tile = 2, |
| }; |
| xnn_params.f32.vmax = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmax_ukernel__scalar_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmaxc_ukernel__scalar_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmaxc_ukernel__scalar_x8, |
| .element_tile = 8, |
| }; |
| xnn_params.f32.vmin = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmin_ukernel__scalar_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vminc_ukernel__scalar_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vminc_ukernel__scalar_x8, |
| .element_tile = 8, |
| }; |
| xnn_params.f32.vmul = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmul_minmax_ukernel__scalar_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmulc_minmax_ukernel__scalar_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmulc_minmax_ukernel__scalar_x8, |
| .element_tile = 8, |
| }; |
| xnn_params.f32.vsub = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsub_minmax_ukernel__scalar_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsubc_minmax_ukernel__scalar_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vrsubc_minmax_ukernel__scalar_x8, |
| .element_tile = 8, |
| }; |
| xnn_params.f32.vsqrdiff = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsqrdiff_ukernel__scalar_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsqrdiffc_ukernel__scalar_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsqrdiffc_ukernel__scalar_x8, |
| .element_tile = 8, |
| }; |
| xnn_params.f32.vmulcaddc = (struct vmulcaddc_parameters) { |
| .ukernel = (xnn_vmulcaddc_ukernel_function) xnn_f32_vmulcaddc_minmax_ukernel_c1__scalar_2x, |
| .init.f32 = xnn_init_f32_minmax_scalar_params, |
| .channel_tile = 1, |
| .row_tile = 2, |
| }; |
| #ifndef XNN_NO_NCHW_OPERATORS |
| init_flags |= XNN_INIT_FLAG_CHW_OPT; |
| |
| xnn_params.f32.spmm = (struct spmm_parameters) { |
| .ukernel = (xnn_spmm_ukernel_function) xnn_f32_spmm_minmax_ukernel_8x1__scalar, |
| .mr = 8, |
| .nr = 1, |
| }; |
| xnn_params.f32.spmm2 = (struct spmm_parameters) { |
| .ukernel = (xnn_spmm_ukernel_function) xnn_f32_spmm_minmax_ukernel_8x2__scalar, |
| .mr = 8, |
| .nr = 2, |
| }; |
| xnn_params.f32.spmm4 = (struct spmm_parameters) { |
| .ukernel = (xnn_spmm_ukernel_function) xnn_f32_spmm_minmax_ukernel_8x4__scalar, |
| .mr = 8, |
| .nr = 4, |
| }; |
| xnn_params.f32.conv_hwc2chw_3x3c3s2 = (struct conv_hwc2chw_parameters) { |
| .ukernel_with_symm_padding = |
| (xnn_conv_hwc2chw_ukernel_function) xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__scalar_1x1, |
| .output_channel_tile = 4, |
| .output_height_tile = 1, |
| .output_width_tile = 1, |
| }; |
| xnn_params.f32.dwconv2d_chw_3x3 = (struct dwconv2d_chw_parameters) { |
| .ukernel = (xnn_dwconv2d_chw_ukernel_function) xnn_f32_dwconv2d_chw_ukernel_3x3p1__scalar_4x1, |
| .output_width_tile = 1, |
| .output_height_tile = 4, |
| }; |
| xnn_params.f32.dwconv2d_chw_3x3s2 = (struct dwconv2d_chw_parameters) { |
| .ukernel = (xnn_dwconv2d_chw_ukernel_function) xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__scalar_2x1_acc2, |
| .output_width_tile = 1, |
| .output_height_tile = 2, |
| }; |
| xnn_params.f32.dwconv2d_chw_5x5 = (struct dwconv2d_chw_parameters) { |
| .ukernel = (xnn_dwconv2d_chw_ukernel_function) xnn_f32_dwconv2d_chw_ukernel_5x5p2__scalar_2x1_acc2, |
| .output_width_tile = 1, |
| .output_height_tile = 2, |
| }; |
| xnn_params.f32.dwconv2d_chw_5x5s2 = (struct dwconv2d_chw_parameters) { |
| .ukernel = (xnn_dwconv2d_chw_ukernel_function) xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__scalar_2x1_acc2, |
| .output_width_tile = 1, |
| .output_height_tile = 2, |
| }; |
| xnn_params.f32.gavgpool_cw = (struct gavgpool_cw_parameters) { |
| .ukernel = (xnn_gavgpool_cw_ukernel_function) xnn_f32_gavgpool_cw_ukernel__scalar_x1, |
| .channel_tile = 1, |
| }; |
| xnn_params.f32.ibilinear_chw = (struct ibilinear_chw_parameters) { |
| .ukernel = (xnn_ibilinear_chw_ukernel_function) xnn_f32_ibilinear_chw_ukernel__scalar_p4, |
| .channel_tile = 1, |
| .pixel_tile = 4, |
| }; |
| #endif // XNN_NO_NCHW_OPERATORS |
| #endif // XNN_NO_F32_OPERATORS |
| |
| /*************************** VCVT micro-kernels ***************************/ |
| #ifndef XNN_NO_VCVT_OPERATORS |
| init_flags |= XNN_INIT_FLAG_VCVT; |
| |
| xnn_params.vcvt.f16_to_f32 = (xnn_univector_ukernel_function) xnn_f16_f32_vcvt_ukernel__scalar_float_x4; |
| #endif // XNN_NO_VCVT_OPERATORS |
| |
| /**************************** X32 micro-kernels ****************************/ |
| #ifndef XNN_NO_X32_OPERATORS |
| init_flags |= XNN_INIT_FLAG_X32; |
| |
| xnn_params.x32.unpool = (xnn_unpool_ukernel_function) xnn_x32_unpool_ukernel__scalar; |
| xnn_params.x32.zip = (struct zip_parameters) { |
| .x2 = (xnn_zipc_ukernel_function) xnn_x32_zip_x2_ukernel__scalar, |
| .x3 = (xnn_zipc_ukernel_function) xnn_x32_zip_x3_ukernel__scalar, |
| .x4 = (xnn_zipc_ukernel_function) xnn_x32_zip_x4_ukernel__scalar, |
| .xm = (xnn_zipv_ukernel_function) xnn_x32_zip_xm_ukernel__scalar, |
| }; |
| #ifndef XNN_NO_NCHW_OPERATORS |
| xnn_params.x32.depthtospace2d_chw2hwc = (struct depthtospace2d_chw2hwc_parameters) { |
| .ukernel = (xnn_depthtospace2d_chw2hwc_ukernel_function) xnn_x32_depthtospace2d_chw2hwc_ukernel__scalar, |
| .channel_tile = 1, |
| .pixel_tile = 1, |
| }; |
| #endif // XNN_NO_NCHW_OPERATORS |
| #endif // XNN_NO_X32_OPERATORS |
| |
| /**************************** XX micro-kernels ****************************/ |
| #ifndef XNN_NO_XX_OPERATORS |
| init_flags |= XNN_INIT_FLAG_XX; |
| |
| xnn_params.xx.copy = (xnn_univector_ukernel_function) xnn_xx_copy_ukernel__memcpy; |
| xnn_params.xx.fill = (struct fill_parameters) { |
| .ukernel = (xnn_fill_ukernel_function) xnn_xx_fill_ukernel__scalar_x16, |
| .row_tile = 1, |
| }; |
| xnn_params.xx.pad = (struct pad_parameters) { |
| .ukernel = (xnn_pad_ukernel_function) xnn_xx_pad_ukernel__scalar, |
| .row_tile = 1, |
| }; |
| #endif // XNN_NO_XX_OPERATORS |
| } |
| |
| #elif XNN_ARCH_ARM64 |
| |
| /**************************** QC8 micro-kernels ****************************/ |
| #ifndef XNN_NO_QC8_OPERATORS |
| init_flags |= XNN_INIT_FLAG_QC8; |
| |
| #if XNN_PLATFORM_IOS || XNN_PLATFORM_MAC |
| #if XNN_ENABLE_ASSEMBLY |
| if (cpuinfo_has_arm_neon_dot()) { |
| xnn_params.qc8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_4x16c4__aarch64_neondot_ld128); |
| xnn_params.qc8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_1x16c4__neondot); |
| xnn_params.qc8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_4x16c4__aarch64_neondot_ld128); |
| xnn_params.qc8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_1x16c4__neondot); |
| xnn_params.qc8.gemm.init.qc8 = xnn_init_qs8_minmax_neon_params; |
| xnn_params.qc8.gemm.mr = 4; |
| xnn_params.qc8.gemm.nr = 16; |
| xnn_params.qc8.gemm.log2_kr = 2; |
| } else { |
| xnn_params.qc8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_2x8c8__aarch64_neon_mlal_padal); |
| xnn_params.qc8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_2x8c8__aarch64_neon_mlal_padal); |
| xnn_params.qc8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_1x8c8__aarch64_neon_mlal_padal); |
| xnn_params.qc8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_1x8c8__aarch64_neon_mlal_padal); |
| xnn_params.qc8.gemm.init.qc8 = xnn_init_qs8_minmax_neon_params; |
| xnn_params.qc8.gemm.mr = 2; |
| xnn_params.qc8.gemm.nr = 8; |
| xnn_params.qc8.gemm.log2_kr = 3; |
| } |
| #else // !XNN_ENABLE_ASSEMBLY |
| if (cpuinfo_has_arm_neon_dot()) { |
| xnn_params.qc8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_4x16c4__neondot); |
| xnn_params.qc8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_1x16c4__neondot); |
| xnn_params.qc8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_4x16c4__neondot); |
| xnn_params.qc8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_1x16c4__neondot); |
| xnn_params.qc8.gemm.init.qc8 = xnn_init_qs8_minmax_neon_params; |
| xnn_params.qc8.gemm.mr = 4; |
| xnn_params.qc8.gemm.nr = 16; |
| xnn_params.qc8.gemm.log2_kr = 2; |
| } else { |
| xnn_params.qc8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_2x8c2__neonv8_mlal_padal_dup); |
| xnn_params.qc8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_2x8c2__neonv8_mlal_padal_dup); |
| xnn_params.qc8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_1x8c2__neonv8_mlal_padal_dup); |
| xnn_params.qc8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_1x8c2__neonv8_mlal_padal_dup); |
| xnn_params.qc8.gemm.init.qc8 = xnn_init_qs8_minmax_neon_params; |
| xnn_params.qc8.gemm.mr = 2; |
| xnn_params.qc8.gemm.nr = 8; |
| xnn_params.qc8.gemm.log2_kr = 1; |
| } |
| #endif // XNN_ENABLE_ASSEMBLY |
| #else // !XNN_PLATFORM_IOS && !XNN_PLATFORM_MAC |
| #if XNN_ENABLE_ASSEMBLY |
| if (cpuinfo_has_arm_neon_dot()) { |
| switch (cpuinfo_get_core(0)->uarch) { |
| case cpuinfo_uarch_cortex_a55: |
| xnn_params.qc8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_4x16c4__aarch64_neondot_cortex_a55); |
| xnn_params.qc8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_4x16c4__aarch64_neondot_cortex_a55); |
| break; |
| case cpuinfo_uarch_cortex_x1: |
| case cpuinfo_uarch_cortex_a78: |
| xnn_params.qc8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_4x16c4__aarch64_neondot_ld128); |
| xnn_params.qc8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_4x16c4__aarch64_neondot_ld128); |
| break; |
| default: |
| xnn_params.qc8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_4x16c4__aarch64_neondot_ld64); |
| xnn_params.qc8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_4x16c4__aarch64_neondot_ld64); |
| break; |
| } |
| xnn_params.qc8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_1x16c4__neondot); |
| xnn_params.qc8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_1x16c4__neondot); |
| xnn_params.qc8.gemm.init.qc8 = xnn_init_qs8_minmax_neon_params; |
| xnn_params.qc8.gemm.mr = 4; |
| xnn_params.qc8.gemm.nr = 16; |
| xnn_params.qc8.gemm.log2_kr = 2; |
| } else { |
| switch (cpuinfo_get_core(0)->uarch) { |
| case cpuinfo_uarch_cortex_a53: |
| case cpuinfo_uarch_cortex_a55r0: |
| xnn_params.qc8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_4x16__aarch64_neon_mlal_lane_cortex_a53); |
| xnn_params.qc8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_4x16__aarch64_neon_mlal_lane_cortex_a53); |
| xnn_params.qc8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_1x16__neonv8_mlal_lane); |
| xnn_params.qc8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_1x16__neonv8_mlal_lane); |
| xnn_params.qc8.gemm.init.qc8 = xnn_init_qs8_minmax_neon_params; |
| xnn_params.qc8.gemm.mr = 4; |
| xnn_params.qc8.gemm.nr = 16; |
| break; |
| |
| case cpuinfo_uarch_cortex_a72: |
| case cpuinfo_uarch_cortex_a73: |
| case cpuinfo_uarch_kryo: |
| xnn_params.qc8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_2x8c8__aarch64_neon_mlal_padal_prfm); |
| xnn_params.qc8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_2x8c8__aarch64_neon_mlal_padal_prfm); |
| xnn_params.qc8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_1x8c8__aarch64_neon_mlal_padal_prfm); |
| xnn_params.qc8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_1x8c8__aarch64_neon_mlal_padal_prfm); |
| xnn_params.qc8.gemm.init.qc8 = xnn_init_qs8_minmax_neon_params; |
| xnn_params.qc8.gemm.mr = 2; |
| xnn_params.qc8.gemm.nr = 8; |
| xnn_params.qc8.gemm.log2_kr = 3; |
| break; |
| |
| default: |
| xnn_params.qc8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_2x8c8__aarch64_neon_mlal_padal); |
| xnn_params.qc8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_2x8c8__aarch64_neon_mlal_padal); |
| xnn_params.qc8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_1x8c8__aarch64_neon_mlal_padal); |
| xnn_params.qc8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_1x8c8__aarch64_neon_mlal_padal); |
| xnn_params.qc8.gemm.init.qc8 = xnn_init_qs8_minmax_neon_params; |
| xnn_params.qc8.gemm.mr = 2; |
| xnn_params.qc8.gemm.nr = 8; |
| xnn_params.qc8.gemm.log2_kr = 3; |
| break; |
| } |
| } |
| #if XNN_MAX_UARCH_TYPES > 1 |
| { |
| /* Choose micro-kernels for little cores according to micro-kernel specification for the big core */ |
| const uint32_t mr = xnn_params.qc8.gemm.mr; |
| const uint32_t nr = xnn_params.qc8.gemm.nr; |
| const uint32_t log2_kr = xnn_params.qc8.gemm.log2_kr; |
| for (size_t i = 1; i < XNN_MAX_UARCH_TYPES; i++) { |
| const struct cpuinfo_uarch_info* uarch_info = cpuinfo_get_uarch(i); |
| if (uarch_info == NULL) { |
| /* No more microarchitectures in the system */ |
| break; |
| } |
| |
| switch (uarch_info->uarch) { |
| case cpuinfo_uarch_cortex_a53: |
| case cpuinfo_uarch_cortex_a55r0: |
| if (mr == 2 && nr == 8 && log2_kr == 3) { |
| xnn_params.qc8.gemm.minmax.gemm.function[i] = (xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_2x8c8__aarch64_neon_mlal_padal_prfm_cortex_a53; |
| xnn_params.qc8.gemm.minmax.igemm.function[i] = (xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_2x8c8__aarch64_neon_mlal_padal_prfm_cortex_a53; |
| xnn_params.qc8.gemm.minmax.gemm1.function[i] = (xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_1x8c8__aarch64_neon_mlal_padal_prfm_cortex_a53; |
| xnn_params.qc8.gemm.minmax.igemm1.function[i] = (xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_1x8c8__aarch64_neon_mlal_padal_prfm_cortex_a53; |
| } |
| break; |
| |
| case cpuinfo_uarch_cortex_a55: |
| if (mr == 4 && nr == 16 && log2_kr == 2 && cpuinfo_has_arm_neon_dot()) { |
| xnn_params.qc8.gemm.minmax.gemm.function[i] = (xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_4x16c4__aarch64_neondot_cortex_a55; |
| xnn_params.qc8.gemm.minmax.igemm.function[i] = (xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_4x16c4__aarch64_neondot_cortex_a55; |
| xnn_params.qc8.gemm.minmax.gemm1.function[i] = (xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_1x16c4__neondot; |
| xnn_params.qc8.gemm.minmax.igemm1.function[i] = (xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_1x16c4__neondot; |
| } |
| break; |
| default: |
| break; |
| } |
| } |
| } |
| #endif // XNN_MAX_UARCH_TYPES > 1 |
| #else // !XNN_ENABLE_ASSEMBLY |
| if (cpuinfo_has_arm_neon_dot()) { |
| xnn_params.qc8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_4x16c4__neondot); |
| xnn_params.qc8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_1x16c4__neondot); |
| xnn_params.qc8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_4x16c4__neondot); |
| xnn_params.qc8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_1x16c4__neondot); |
| xnn_params.qc8.gemm.init.qc8 = xnn_init_qs8_minmax_neon_params; |
| xnn_params.qc8.gemm.mr = 4; |
| xnn_params.qc8.gemm.nr = 16; |
| xnn_params.qc8.gemm.log2_kr = 2; |
| } else { |
| xnn_params.qc8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_2x8c2__neonv8_mlal_padal_dup); |
| xnn_params.qc8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_2x8c2__neonv8_mlal_padal_dup); |
| xnn_params.qc8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_1x8c2__neonv8_mlal_padal_dup); |
| xnn_params.qc8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_1x8c2__neonv8_mlal_padal_dup); |
| xnn_params.qc8.gemm.init.qc8 = xnn_init_qs8_minmax_neon_params; |
| xnn_params.qc8.gemm.mr = 2; |
| xnn_params.qc8.gemm.nr = 8; |
| xnn_params.qc8.gemm.log2_kr = 1; |
| } |
| #endif // XNN_ENABLE_ASSEMBLY |
| #endif // XNN_PLATFORM_IOS || XNN_PLATFORM_MAC |
| |
| xnn_params.qc8.dwconv[0].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qc8_dwconv_minmax_fp32_ukernel_up16x9__neonv8_mla8_ld64; |
| xnn_params.qc8.dwconv[0].init.qc8 = xnn_init_qs8_minmax_neon_params; |
| xnn_params.qc8.dwconv[0].channel_tile = 16; |
| xnn_params.qc8.dwconv[0].primary_tile = 9; |
| xnn_params.qc8.dwconv[1].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qc8_dwconv_minmax_fp32_ukernel_up16x25__neonv8_mla8_ld64; |
| xnn_params.qc8.dwconv[1].init.qc8 = xnn_init_qs8_minmax_neon_params; |
| xnn_params.qc8.dwconv[1].channel_tile = 16; |
| xnn_params.qc8.dwconv[1].primary_tile = 25; |
| #endif // XNN_NO_QC8_OPERATORS |
| |
| /**************************** QS8 micro-kernels ****************************/ |
| #ifndef XNN_NO_QS8_OPERATORS |
| init_flags |= XNN_INIT_FLAG_QS8; |
| |
| #if XNN_PLATFORM_IOS || XNN_PLATFORM_MAC |
| #if XNN_ENABLE_ASSEMBLY |
| if (cpuinfo_has_arm_neon_dot()) { |
| xnn_params.qs8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_rndnu_ukernel_4x16c4__aarch64_neondot_ld128); |
| xnn_params.qs8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_rndnu_ukernel_1x16c4__neondot); |
| xnn_params.qs8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_rndnu_ukernel_4x16c4__aarch64_neondot_ld128); |
| xnn_params.qs8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_rndnu_ukernel_1x16c4__neondot); |
| xnn_params.qs8.gemm.init.qs8 = xnn_init_qs8_conv_minmax_rndnu_neon_params; |
| xnn_params.qs8.gemm.mr = 4; |
| xnn_params.qs8.gemm.nr = 16; |
| xnn_params.qs8.gemm.log2_kr = 2; |
| } else { |
| xnn_params.qs8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_rndnu_ukernel_2x8c8__aarch64_neon_mlal_padal); |
| xnn_params.qs8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_rndnu_ukernel_2x8c8__aarch64_neon_mlal_padal); |
| xnn_params.qs8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_rndnu_ukernel_1x8c8__aarch64_neon_mlal_padal); |
| xnn_params.qs8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_rndnu_ukernel_1x8c8__aarch64_neon_mlal_padal); |
| xnn_params.qs8.gemm.init.qs8 = xnn_init_qs8_conv_minmax_rndnu_neon_params; |
| xnn_params.qs8.gemm.mr = 2; |
| xnn_params.qs8.gemm.nr = 8; |
| xnn_params.qs8.gemm.log2_kr = 3; |
| } |
| #else // !XNN_ENABLE_ASSEMBLY |
| if (cpuinfo_has_arm_neon_dot()) { |
| xnn_params.qs8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_rndnu_ukernel_4x16c4__neondot); |
| xnn_params.qs8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_rndnu_ukernel_1x16c4__neondot); |
| xnn_params.qs8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_rndnu_ukernel_4x16c4__neondot); |
| xnn_params.qs8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_rndnu_ukernel_1x16c4__neondot); |
| xnn_params.qs8.gemm.init.qs8 = xnn_init_qs8_conv_minmax_rndnu_neon_params; |
| xnn_params.qs8.gemm.mr = 4; |
| xnn_params.qs8.gemm.nr = 16; |
| xnn_params.qs8.gemm.log2_kr = 2; |
| } else { |
| xnn_params.qs8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_rndnu_ukernel_2x8s4c2__neon_mlal_padal); |
| xnn_params.qs8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_rndnu_ukernel_2x8s4c2__neon_mlal_padal); |
| xnn_params.qs8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_rndnu_ukernel_1x8s4c2__neon_mlal_padal); |
| xnn_params.qs8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_rndnu_ukernel_1x8s4c2__neon_mlal_padal); |
| xnn_params.qs8.gemm.init.qs8 = xnn_init_qs8_conv_minmax_rndnu_neon_params; |
| xnn_params.qs8.gemm.mr = 2; |
| xnn_params.qs8.gemm.nr = 8; |
| xnn_params.qs8.gemm.log2_kr = 1; |
| xnn_params.qs8.gemm.log2_sr = 2; |
| } |
| #endif // XNN_ENABLE_ASSEMBLY |
| #else // !XNN_PLATFORM_IOS && !XNN_PLATFORM_MAC |
| #if XNN_ENABLE_ASSEMBLY |
| if (cpuinfo_has_arm_neon_dot()) { |
| switch (cpuinfo_get_core(0)->uarch) { |
| case cpuinfo_uarch_cortex_a55: |
| xnn_params.qs8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_rndnu_ukernel_4x16c4__aarch64_neondot_cortex_a55); |
| xnn_params.qs8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_rndnu_ukernel_4x16c4__aarch64_neondot_cortex_a55); |
| break; |
| case cpuinfo_uarch_cortex_x1: |
| case cpuinfo_uarch_cortex_a78: |
| xnn_params.qs8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_rndnu_ukernel_4x16c4__aarch64_neondot_ld128); |
| xnn_params.qs8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_rndnu_ukernel_4x16c4__aarch64_neondot_ld128); |
| break; |
| default: |
| xnn_params.qs8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_rndnu_ukernel_4x16c4__aarch64_neondot_ld64); |
| xnn_params.qs8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_rndnu_ukernel_4x16c4__aarch64_neondot_ld64); |
| break; |
| } |
| xnn_params.qs8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_rndnu_ukernel_1x16c4__neondot); |
| xnn_params.qs8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_rndnu_ukernel_1x16c4__neondot); |
| xnn_params.qs8.gemm.init.qs8 = xnn_init_qs8_conv_minmax_rndnu_neon_params; |
| xnn_params.qs8.gemm.mr = 4; |
| xnn_params.qs8.gemm.nr = 16; |
| xnn_params.qs8.gemm.log2_kr = 2; |
| } else { |
| switch (cpuinfo_get_core(0)->uarch) { |
| case cpuinfo_uarch_cortex_a53: |
| case cpuinfo_uarch_cortex_a55r0: |
| xnn_params.qs8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_rndnu_ukernel_4x16__aarch64_neon_mlal_lane_cortex_a53); |
| xnn_params.qs8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_rndnu_ukernel_4x16__aarch64_neon_mlal_lane_cortex_a53); |
| xnn_params.qs8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_rndnu_ukernel_1x16__neon_mlal_lane); |
| xnn_params.qs8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_rndnu_ukernel_1x16__neon_mlal_lane); |
| xnn_params.qs8.gemm.init.qs8 = xnn_init_qs8_conv_minmax_rndnu_neon_params; |
| xnn_params.qs8.gemm.mr = 4; |
| xnn_params.qs8.gemm.nr = 16; |
| break; |
| |
| case cpuinfo_uarch_cortex_a72: |
| case cpuinfo_uarch_cortex_a73: |
| case cpuinfo_uarch_kryo: |
| xnn_params.qs8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_rndnu_ukernel_2x8c8__aarch64_neon_mlal_padal_prfm); |
| xnn_params.qs8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_rndnu_ukernel_2x8c8__aarch64_neon_mlal_padal_prfm); |
| xnn_params.qs8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_rndnu_ukernel_1x8c8__aarch64_neon_mlal_padal_prfm); |
| xnn_params.qs8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_rndnu_ukernel_1x8c8__aarch64_neon_mlal_padal_prfm); |
| xnn_params.qs8.gemm.init.qs8 = xnn_init_qs8_conv_minmax_rndnu_neon_params; |
| xnn_params.qs8.gemm.mr = 2; |
| xnn_params.qs8.gemm.nr = 8; |
| xnn_params.qs8.gemm.log2_kr = 3; |
| break; |
| |
| default: |
| xnn_params.qs8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_rndnu_ukernel_2x8c8__aarch64_neon_mlal_padal); |
| xnn_params.qs8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_rndnu_ukernel_2x8c8__aarch64_neon_mlal_padal); |
| xnn_params.qs8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_rndnu_ukernel_1x8c8__aarch64_neon_mlal_padal); |
| xnn_params.qs8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_rndnu_ukernel_1x8c8__aarch64_neon_mlal_padal); |
| xnn_params.qs8.gemm.init.qs8 = xnn_init_qs8_conv_minmax_rndnu_neon_params; |
| xnn_params.qs8.gemm.mr = 2; |
| xnn_params.qs8.gemm.nr = 8; |
| xnn_params.qs8.gemm.log2_kr = 3; |
| break; |
| } |
| } |
| #if XNN_MAX_UARCH_TYPES > 1 |
| { |
| /* Choose micro-kernels for little cores according to micro-kernel specification for the big core */ |
| const uint32_t mr = xnn_params.qs8.gemm.mr; |
| const uint32_t nr = xnn_params.qs8.gemm.nr; |
| const uint32_t log2_kr = xnn_params.qs8.gemm.log2_kr; |
| for (size_t i = 1; i < XNN_MAX_UARCH_TYPES; i++) { |
| const struct cpuinfo_uarch_info* uarch_info = cpuinfo_get_uarch(i); |
| if (uarch_info == NULL) { |
| /* No more microarchitectures in the system */ |
| break; |
| } |
| |
| switch (uarch_info->uarch) { |
| case cpuinfo_uarch_cortex_a53: |
| case cpuinfo_uarch_cortex_a55r0: |
| if (mr == 2 && nr == 8 && log2_kr == 3) { |
| xnn_params.qs8.gemm.minmax.gemm.function[i] = (xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_rndnu_ukernel_2x8c8__aarch64_neon_mlal_padal_prfm_cortex_a53; |
| xnn_params.qs8.gemm.minmax.igemm.function[i] = (xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_rndnu_ukernel_2x8c8__aarch64_neon_mlal_padal_prfm_cortex_a53; |
| xnn_params.qs8.gemm.minmax.gemm1.function[i] = (xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_rndnu_ukernel_1x8c8__aarch64_neon_mlal_padal_prfm_cortex_a53; |
| xnn_params.qs8.gemm.minmax.igemm1.function[i] = (xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_rndnu_ukernel_1x8c8__aarch64_neon_mlal_padal_prfm_cortex_a53; |
| } |
| break; |
| |
| case cpuinfo_uarch_cortex_a55: |
| if (mr == 4 && nr == 16 && log2_kr == 2 && cpuinfo_has_arm_neon_dot()) { |
| xnn_params.qs8.gemm.minmax.gemm.function[i] = (xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_rndnu_ukernel_4x16c4__aarch64_neondot_cortex_a55; |
| xnn_params.qs8.gemm.minmax.igemm.function[i] = (xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_rndnu_ukernel_4x16c4__aarch64_neondot_cortex_a55; |
| xnn_params.qs8.gemm.minmax.gemm1.function[i] = (xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_rndnu_ukernel_1x16c4__neondot; |
| xnn_params.qs8.gemm.minmax.igemm1.function[i] = (xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_rndnu_ukernel_1x16c4__neondot; |
| } |
| break; |
| default: |
| break; |
| } |
| } |
| } |
| #endif // XNN_MAX_UARCH_TYPES > 1 |
| #else // !XNN_ENABLE_ASSEMBLY |
| if (cpuinfo_has_arm_neon_dot()) { |
| xnn_params.qs8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_rndnu_ukernel_4x16c4__neondot); |
| xnn_params.qs8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_rndnu_ukernel_1x16c4__neondot); |
| xnn_params.qs8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_rndnu_ukernel_4x16c4__neondot); |
| xnn_params.qs8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_rndnu_ukernel_1x16c4__neondot); |
| xnn_params.qs8.gemm.init.qs8 = xnn_init_qs8_conv_minmax_rndnu_neon_params; |
| xnn_params.qs8.gemm.mr = 4; |
| xnn_params.qs8.gemm.nr = 16; |
| xnn_params.qs8.gemm.log2_kr = 2; |
| } else { |
| xnn_params.qs8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_rndnu_ukernel_2x8s4c2__neon_mlal_padal); |
| xnn_params.qs8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_rndnu_ukernel_2x8s4c2__neon_mlal_padal); |
| xnn_params.qs8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_rndnu_ukernel_1x8s4c2__neon_mlal_padal); |
| xnn_params.qs8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_rndnu_ukernel_1x8s4c2__neon_mlal_padal); |
| xnn_params.qs8.gemm.init.qs8 = xnn_init_qs8_conv_minmax_rndnu_neon_params; |
| xnn_params.qs8.gemm.mr = 2; |
| xnn_params.qs8.gemm.nr = 8; |
| xnn_params.qs8.gemm.log2_kr = 1; |
| } |
| #endif // XNN_ENABLE_ASSEMBLY |
| #endif // XNN_PLATFORM_IOS || XNN_PLATFORM_MAC |
| |
| xnn_params.qs8.dwconv[0].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qs8_dwconv_minmax_rndnu_ukernel_up16x9__neon_mla8_ld64; |
| xnn_params.qs8.dwconv[0].init.qs8 = xnn_init_qs8_conv_minmax_rndnu_neon_params; |
| xnn_params.qs8.dwconv[0].channel_tile = 16; |
| xnn_params.qs8.dwconv[0].primary_tile = 9; |
| xnn_params.qs8.dwconv[1].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qs8_dwconv_minmax_rndnu_ukernel_up16x25__neon_mla8_ld64; |
| xnn_params.qs8.dwconv[1].init.qs8 = xnn_init_qs8_conv_minmax_rndnu_neon_params; |
| xnn_params.qs8.dwconv[1].channel_tile = 16; |
| xnn_params.qs8.dwconv[1].primary_tile = 25; |
| |
| xnn_params.qs8.gavgpool = (struct gavgpool_parameters) { |
| .up = (xnn_gavgpool_unipass_ukernel_function) xnn_qs8_gavgpool_minmax_ukernel_7x__neon_c8_acc2, |
| .mp = (xnn_gavgpool_multipass_ukernel_function) xnn_qs8_gavgpool_minmax_ukernel_7p7x__neon_c8_acc2, |
| .mr = 7, |
| }; |
| |
| xnn_params.qs8.vadd = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vadd_minmax_ukernel__neon_ld64_x32, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vaddc_minmax_ukernel__neon_ld64_x32, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vaddc_minmax_ukernel__neon_ld64_x32, |
| .init.qs8_addsub = xnn_init_qs8_add_minmax_neon_params, |
| .element_tile = 32, |
| }; |
| xnn_params.qs8.vmul = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vmul_minmax_fp32_ukernel__neonv8_ld64_x16, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vmulc_minmax_fp32_ukernel__neonv8_ld64_x16, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vmulc_minmax_fp32_ukernel__neonv8_ld64_x16, |
| .init.qs8_mul = xnn_init_qs8_mul_minmax_fp32_neonv8_params, |
| .element_tile = 16, |
| }; |
| #endif // XNN_NO_QS8_OPERATORS |
| |
| /**************************** QU8 micro-kernels ****************************/ |
| #ifndef XNN_NO_QU8_OPERATORS |
| init_flags |= XNN_INIT_FLAG_QU8; |
| |
| #if XNN_ENABLE_ASSEMBLY |
| if (cpuinfo_has_arm_neon_dot()) { |
| switch (cpuinfo_get_core(0)->uarch) { |
| case cpuinfo_uarch_cortex_a55: |
| xnn_params.qu8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qu8_gemm_minmax_rndnu_ukernel_4x16c4__aarch64_neondot_cortex_a55); |
| xnn_params.qu8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qu8_igemm_minmax_rndnu_ukernel_4x16c4__aarch64_neondot_cortex_a55); |
| xnn_params.qu8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qu8_gemm_minmax_rndnu_ukernel_1x16c4__neondot); |
| xnn_params.qu8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qu8_igemm_minmax_rndnu_ukernel_1x16c4__neondot); |
| xnn_params.qu8.gemm.init.qu8 = xnn_init_qu8_conv_minmax_rndnu_neon_params; |
| xnn_params.qu8.gemm.mr = 4; |
| xnn_params.qu8.gemm.nr = 16; |
| xnn_params.qu8.gemm.log2_kr = 2; |
| break; |
| default: |
| xnn_params.qu8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qu8_gemm_minmax_rndnu_ukernel_4x16c4__aarch64_neondot_ld128); |
| xnn_params.qu8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qu8_igemm_minmax_rndnu_ukernel_4x16c4__aarch64_neondot_ld128); |
| xnn_params.qu8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qu8_gemm_minmax_rndnu_ukernel_1x16c4__neondot); |
| xnn_params.qu8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qu8_igemm_minmax_rndnu_ukernel_1x16c4__neondot); |
| xnn_params.qu8.gemm.init.qu8 = xnn_init_qu8_conv_minmax_rndnu_neon_params; |
| xnn_params.qu8.gemm.mr = 4; |
| xnn_params.qu8.gemm.nr = 16; |
| xnn_params.qu8.gemm.log2_kr = 2; |
| break; |
| } |
| } else { |
| switch (cpuinfo_get_core(0)->uarch) { |
| case cpuinfo_uarch_cortex_a53: |
| case cpuinfo_uarch_cortex_a55r0: |
| xnn_params.qu8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qu8_gemm_minmax_rndnu_ukernel_4x16__aarch64_neon_mlal_lane_prfm_cortex_a53); |
| xnn_params.qu8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qu8_igemm_minmax_rndnu_ukernel_4x16__aarch64_neon_mlal_lane_prfm_cortex_a53); |
| xnn_params.qu8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qu8_gemm_minmax_rndnu_ukernel_1x16__neon_mlal_lane); |
| xnn_params.qu8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qu8_igemm_minmax_rndnu_ukernel_1x16__neon_mlal_lane); |
| xnn_params.qu8.gemm.init.qu8 = xnn_init_qu8_conv_minmax_rndnu_neon_params; |
| xnn_params.qu8.gemm.mr = 4; |
| xnn_params.qu8.gemm.nr = 16; |
| break; |
| |
| case cpuinfo_uarch_cortex_a57: |
| case cpuinfo_uarch_cortex_a72: |
| case cpuinfo_uarch_cortex_a73: |
| case cpuinfo_uarch_cortex_a75: |
| case cpuinfo_uarch_cortex_a76: |
| case cpuinfo_uarch_exynos_m1: |
| case cpuinfo_uarch_exynos_m2: |
| case cpuinfo_uarch_exynos_m3: |
| case cpuinfo_uarch_exynos_m4: |
| xnn_params.qu8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qu8_gemm_minmax_rndnu_ukernel_4x16__aarch64_neon_mlal_lane_prfm_cortex_a75); |
| xnn_params.qu8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qu8_igemm_minmax_rndnu_ukernel_4x16__aarch64_neon_mlal_lane_prfm_cortex_a75); |
| xnn_params.qu8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qu8_gemm_minmax_rndnu_ukernel_1x16__neon_mlal_lane); |
| xnn_params.qu8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qu8_igemm_minmax_rndnu_ukernel_1x16__neon_mlal_lane); |
| xnn_params.qu8.gemm.init.qu8 = xnn_init_qu8_conv_minmax_rndnu_neon_params; |
| xnn_params.qu8.gemm.mr = 4; |
| xnn_params.qu8.gemm.nr = 16; |
| break; |
| |
| case cpuinfo_uarch_kryo: |
| xnn_params.qu8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qu8_gemm_minmax_rndnu_ukernel_4x16__neon_mlal_lane); |
| xnn_params.qu8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qu8_igemm_minmax_rndnu_ukernel_4x16__neon_mlal_lane); |
| xnn_params.qu8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qu8_gemm_minmax_rndnu_ukernel_1x16__neon_mlal_lane); |
| xnn_params.qu8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qu8_igemm_minmax_rndnu_ukernel_1x16__neon_mlal_lane); |
| xnn_params.qu8.gemm.init.qu8 = xnn_init_qu8_conv_minmax_rndnu_neon_params; |
| xnn_params.qu8.gemm.mr = 4; |
| xnn_params.qu8.gemm.nr = 16; |
| break; |
| |
| default: |
| xnn_params.qu8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qu8_gemm_minmax_rndnu_ukernel_4x16__aarch64_neon_mlal_lane_cortex_a75); |
| xnn_params.qu8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qu8_igemm_minmax_rndnu_ukernel_4x16__aarch64_neon_mlal_lane_cortex_a75); |
| xnn_params.qu8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qu8_gemm_minmax_rndnu_ukernel_1x16__neon_mlal_lane); |
| xnn_params.qu8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qu8_igemm_minmax_rndnu_ukernel_1x16__neon_mlal_lane); |
| xnn_params.qu8.gemm.init.qu8 = xnn_init_qu8_conv_minmax_rndnu_neon_params; |
| xnn_params.qu8.gemm.mr = 4; |
| xnn_params.qu8.gemm.nr = 16; |
| break; |
| } |
| } |
| #if XNN_MAX_UARCH_TYPES > 1 |
| { |
| /* Choose micro-kernels for little cores according to micro-kernel specification for the big core */ |
| const uint32_t mr = xnn_params.qu8.gemm.mr; |
| const uint32_t nr = xnn_params.qu8.gemm.nr; |
| const uint32_t log2_kr = xnn_params.qu8.gemm.log2_kr; |
| for (size_t i = 1; i < XNN_MAX_UARCH_TYPES; i++) { |
| const struct cpuinfo_uarch_info* uarch_info = cpuinfo_get_uarch(i); |
| if (uarch_info == NULL) { |
| /* No more microarchitectures in the system */ |
| break; |
| } |
| |
| switch (uarch_info->uarch) { |
| case cpuinfo_uarch_cortex_a53: |
| case cpuinfo_uarch_cortex_a55r0: |
| if (mr == 4 && nr == 16 && log2_kr == 0) { |
| xnn_params.qu8.gemm.minmax.gemm.function[i] = (xnn_gemm_ukernel_function) xnn_qu8_gemm_minmax_rndnu_ukernel_4x16__aarch64_neon_mlal_lane_prfm_cortex_a53; |
| xnn_params.qu8.gemm.minmax.igemm.function[i] = (xnn_igemm_ukernel_function) xnn_qu8_igemm_minmax_rndnu_ukernel_4x16__aarch64_neon_mlal_lane_prfm_cortex_a53; |
| } |
| break; |
| |
| case cpuinfo_uarch_cortex_a55: |
| if (mr == 4 && nr == 16 && log2_kr == 2 && cpuinfo_has_arm_neon_dot()) { |
| xnn_params.qu8.gemm.minmax.gemm.function[i] = (xnn_gemm_ukernel_function) xnn_qu8_gemm_minmax_rndnu_ukernel_4x16c4__aarch64_neondot_cortex_a55; |
| xnn_params.qu8.gemm.minmax.igemm.function[i] = (xnn_igemm_ukernel_function) xnn_qu8_igemm_minmax_rndnu_ukernel_4x16c4__aarch64_neondot_cortex_a55; |
| } |
| break; |
| default: |
| break; |
| } |
| } |
| } |
| #endif // XNN_MAX_UARCH_TYPES > 1 |
| #else // !XNN_ENABLE_ASSEMBLY |
| if (cpuinfo_has_arm_neon_dot()) { |
| xnn_params.qu8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qu8_gemm_minmax_rndnu_ukernel_4x16c4__neondot); |
| xnn_params.qu8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qu8_igemm_minmax_rndnu_ukernel_4x16c4__neondot); |
| xnn_params.qu8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qu8_gemm_minmax_rndnu_ukernel_1x16c4__neondot); |
| xnn_params.qu8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qu8_igemm_minmax_rndnu_ukernel_1x16c4__neondot); |
| xnn_params.qu8.gemm.init.qu8 = xnn_init_qu8_conv_minmax_rndnu_neon_params; |
| xnn_params.qu8.gemm.mr = 4; |
| xnn_params.qu8.gemm.nr = 16; |
| xnn_params.qu8.gemm.log2_kr = 2; |
| } else { |
| xnn_params.qu8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qu8_gemm_minmax_rndnu_ukernel_4x16__neon_mlal_lane); |
| xnn_params.qu8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qu8_igemm_minmax_rndnu_ukernel_4x16__neon_mlal_lane); |
| xnn_params.qu8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qu8_gemm_minmax_rndnu_ukernel_1x16__neon_mlal_lane); |
| xnn_params.qu8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qu8_igemm_minmax_rndnu_ukernel_1x16__neon_mlal_lane); |
| xnn_params.qu8.gemm.init.qu8 = xnn_init_qu8_conv_minmax_rndnu_neon_params; |
| xnn_params.qu8.gemm.mr = 4; |
| xnn_params.qu8.gemm.nr = 16; |
| #endif // XNN_ENABLE_ASSEMBLY |
| |
| xnn_params.qu8.dwconv[0].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qu8_dwconv_minmax_rndnu_ukernel_up16x9__neon_mul8; |
| xnn_params.qu8.dwconv[0].init.qu8 = xnn_init_qu8_conv_minmax_rndnu_neon_params; |
| xnn_params.qu8.dwconv[0].channel_tile = 16; |
| xnn_params.qu8.dwconv[0].primary_tile = 9; |
| xnn_params.qu8.dwconv[1].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qu8_dwconv_minmax_rndnu_ukernel_up8x25__neon_mul8; |
| xnn_params.qu8.dwconv[1].init.qu8 = xnn_init_qu8_conv_minmax_rndnu_neon_params; |
| xnn_params.qu8.dwconv[1].channel_tile = 8; |
| xnn_params.qu8.dwconv[1].primary_tile = 25; |
| |
| xnn_params.qu8.avgpool = (struct avgpool_parameters) { |
| .up = (xnn_avgpool_unipass_ukernel_function) xnn_qu8_avgpool_minmax_ukernel_9x__neon_c8, |
| .mp = (xnn_avgpool_multipass_ukernel_function) xnn_qu8_avgpool_minmax_ukernel_9p8x__neon_c8, |
| .mr = 9, |
| .qr = 8, |
| }; |
| xnn_params.qu8.gavgpool = (struct gavgpool_parameters) { |
| .up = (xnn_gavgpool_unipass_ukernel_function) xnn_qu8_gavgpool_minmax_ukernel_7x__neon_c8, |
| .mp = (xnn_gavgpool_multipass_ukernel_function) xnn_qu8_gavgpool_minmax_ukernel_7p7x__neon_c8, |
| .mr = 7, |
| }; |
| xnn_params.qu8.vadd = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vadd_minmax_ukernel__neon_ld64_x32, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vaddc_minmax_ukernel__neon_ld64_x32, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vaddc_minmax_ukernel__neon_ld64_x32, |
| .init.qu8_addsub = xnn_init_qu8_add_minmax_neon_params, |
| .element_tile = 8, |
| }; |
| xnn_params.qu8.vmul = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vmul_minmax_fp32_ukernel__neonv8_ld64_x16, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vmulc_minmax_fp32_ukernel__neonv8_ld64_x16, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vmulc_minmax_fp32_ukernel__neonv8_ld64_x16, |
| .init.qu8_mul = xnn_init_qu8_mul_minmax_fp32_neonv8_params, |
| .element_tile = 16, |
| }; |
| #endif // XNN_NO_QU8_OPERATORS |
| |
| /**************************** S8 micro-kernels ****************************/ |
| #ifndef XNN_NO_S8_OPERATORS |
| init_flags |= XNN_INIT_FLAG_S8; |
| |
| xnn_params.s8.clamp = (struct vunary_parameters) { |
| .ukernel = (xnn_univector_ukernel_function) xnn_s8_vclamp_ukernel__neon_x64, |
| .init.s8_minmax = xnn_init_s8_minmax_neon_params, |
| .element_tile = 64, |
| }; |
| xnn_params.s8.maxpool = (struct maxpool_parameters) { |
| .ukernel = (xnn_maxpool_ukernel_function) xnn_s8_maxpool_minmax_ukernel_9p8x__neon_c16, |
| .init.s8 = xnn_init_s8_minmax_neon_params, |
| .mr = 9, |
| .qr = 8, |
| }; |
| #endif // XNN_NO_S8_OPERATORS |
| |
| /**************************** U8 micro-kernels ****************************/ |
| #ifndef XNN_NO_U8_OPERATORS |
| init_flags |= XNN_INIT_FLAG_U8; |
| |
| xnn_params.u8.clamp = (struct vunary_parameters) { |
| .ukernel = (xnn_univector_ukernel_function) xnn_u8_vclamp_ukernel__neon_x64, |
| .init.u8_minmax = xnn_init_u8_minmax_neon_params, |
| .element_tile = 64, |
| }; |
| xnn_params.u8.maxpool = (struct maxpool_parameters) { |
| .ukernel = (xnn_maxpool_ukernel_function) xnn_u8_maxpool_minmax_ukernel_9p8x__neon_c16, |
| .init.u8 = xnn_init_u8_minmax_neon_params, |
| .mr = 9, |
| .qr = 8, |
| }; |
| xnn_params.u8.lut32norm = xnn_u8_lut32norm_ukernel__scalar; |
| xnn_params.u8.rmax = xnn_u8_rmax_ukernel__neon; |
| #endif // XNN_NO_U8_OPERATORS |
| |
| /**************************** X8 micro-kernels ****************************/ |
| #ifndef XNN_NO_X8_OPERATORS |
| init_flags |= XNN_INIT_FLAG_X8; |
| |
| xnn_params.x8.lut = xnn_x8_lut_ukernel__neon_tbx128x4_x64; |
| xnn_params.x8.zip = (struct zip_parameters) { |
| .x2 = (xnn_zipc_ukernel_function) xnn_x8_zip_x2_ukernel__neon, |
| .x3 = (xnn_zipc_ukernel_function) xnn_x8_zip_x3_ukernel__neon, |
| .x4 = (xnn_zipc_ukernel_function) xnn_x8_zip_x4_ukernel__neon, |
| .xm = (xnn_zipv_ukernel_function) xnn_x8_zip_xm_ukernel__neon, |
| }; |
| #endif // XNN_NO_X8_OPERATORS |
| |
| /**************************** F16 micro-kernels ****************************/ |
| #ifndef XNN_NO_F16_OPERATORS |
| if (cpuinfo_has_arm_neon_fp16_arith()) { |
| init_flags |= XNN_INIT_FLAG_F16; |
| xnn_params.f16.gemm.mr = 6; |
| xnn_params.f16.gemm.nr = 16; |
| |
| #if XNN_ENABLE_ASSEMBLY |
| switch (cpuinfo_get_core(0)->uarch) { |
| case cpuinfo_uarch_cortex_a55: |
| xnn_params.f16.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f16_gemm_minmax_ukernel_6x16__aarch64_neonfp16arith_cortex_a55); |
| break; |
| |
| case cpuinfo_uarch_cortex_a75: |
| case cpuinfo_uarch_cortex_x1: |
| xnn_params.f16.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f16_gemm_minmax_ukernel_6x16__aarch64_neonfp16arith_cortex_a75); |
| break; |
| |
| default: |
| xnn_params.f16.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f16_gemm_minmax_ukernel_6x16__aarch64_neonfp16arith_ld32); |
| break; |
| } |
| xnn_params.f16.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f16_gemm_minmax_ukernel_1x16__aarch64_neonfp16arith_ld32); |
| |
| #if XNN_MAX_UARCH_TYPES > 1 |
| { |
| /* Choose micro-kernels for little cores according to micro-kernel specification for the big core */ |
| const uint32_t mr = xnn_params.f16.gemm.mr; |
| const uint32_t nr = xnn_params.f16.gemm.nr; |
| for (size_t i = 1; i < XNN_MAX_UARCH_TYPES; i++) { |
| const struct cpuinfo_uarch_info* uarch_info = cpuinfo_get_uarch(i); |
| if (uarch_info == NULL) { |
| /* No more microarchitectures in the system */ |
| break; |
| } |
| |
| switch (uarch_info->uarch) { |
| case cpuinfo_uarch_cortex_a55: |
| if (mr == 6 && nr == 16) { |
| xnn_params.f16.gemm.minmax.gemm.function[i] = (xnn_gemm_ukernel_function) xnn_f16_gemm_minmax_ukernel_6x16__aarch64_neonfp16arith_cortex_a55; |
| } |
| break; |
| |
| case cpuinfo_uarch_cortex_a55r0: |
| if (mr == 6 && nr == 16) { |
| xnn_params.f16.gemm.minmax.gemm.function[i] = (xnn_gemm_ukernel_function) xnn_f16_gemm_minmax_ukernel_6x16__neonfp16arith_ld64; |
| } |
| break; |
| |
| /* Cortex A75 is the medium core Exynos 9820 (M4) */ |
| case cpuinfo_uarch_cortex_a75: |
| if (mr == 6 && nr == 16) { |
| xnn_params.f16.gemm.minmax.gemm.function[i] = (xnn_gemm_ukernel_function) xnn_f16_gemm_minmax_ukernel_6x16__aarch64_neonfp16arith_cortex_a75; |
| } |
| break; |
| |
| default: |
| break; |
| } |
| } |
| } |
| #endif // XNN_MAX_UARCH_TYPES > 1 |
| #else // XNN_ENABLE_ASSEMBLY |
| xnn_params.f16.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f16_gemm_minmax_ukernel_6x16__neonfp16arith_ld64); |
| xnn_params.f16.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f16_gemm_minmax_ukernel_1x16__neonfp16arith_ld64); |
| #endif // XNN_ENABLE_ASSEMBLY |
| xnn_params.f16.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f16_igemm_minmax_ukernel_6x16__neonfp16arith_ld64); |
| xnn_params.f16.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f16_igemm_minmax_ukernel_1x16__neonfp16arith_ld64); |
| xnn_params.f16.gemm.init.f16 = xnn_init_f16_scaleminmax_params; |
| |
| xnn_params.f16.dwconv[0].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f16_dwconv_minmax_ukernel_up16x4__neonfp16arith; |
| xnn_params.f16.dwconv[0].init.f16 = xnn_init_f16_minmax_params; |
| xnn_params.f16.dwconv[0].channel_tile = 16; |
| xnn_params.f16.dwconv[0].primary_tile = 4; |
| |
| xnn_params.f16.dwconv[1].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f16_dwconv_minmax_ukernel_up16x9__neonfp16arith; |
| xnn_params.f16.dwconv[1].init.f16 = xnn_init_f16_minmax_params; |
| xnn_params.f16.dwconv[1].channel_tile = 16; |
| xnn_params.f16.dwconv[1].primary_tile = 9; |
| |
| xnn_params.f16.dwconv[2].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f16_dwconv_minmax_ukernel_up8x25__neonfp16arith_acc2; |
| xnn_params.f16.dwconv[2].init.f16 = xnn_init_f16_minmax_params; |
| xnn_params.f16.dwconv[2].channel_tile = 8; |
| xnn_params.f16.dwconv[2].primary_tile = 25; |
| |
| xnn_params.f16.gavgpool = (struct gavgpool_parameters) { |
| .up = (xnn_gavgpool_unipass_ukernel_function) xnn_f16_gavgpool_minmax_ukernel_7x__neonfp16arith_c8, |
| .mp = (xnn_gavgpool_multipass_ukernel_function) xnn_f16_gavgpool_minmax_ukernel_7p7x__neonfp16arith_c8, |
| .mr = 7, |
| }; |
| xnn_params.f16.vadd = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f16_vadd_minmax_ukernel__neonfp16arith_x16, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f16_vaddc_minmax_ukernel__neonfp16arith_x16, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f16_vaddc_minmax_ukernel__neonfp16arith_x16, |
| .element_tile = 16, |
| }; |
| xnn_params.f16.vmul = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f16_vmul_minmax_ukernel__neonfp16arith_x16, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f16_vmulc_minmax_ukernel__neonfp16arith_x16, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f16_vmulc_minmax_ukernel__neonfp16arith_x16, |
| .element_tile = 16, |
| }; |
| xnn_params.f16.vmulcaddc = (struct vmulcaddc_parameters) { |
| .ukernel = (xnn_vmulcaddc_ukernel_function) xnn_f16_vmulcaddc_minmax_ukernel_c8__neonfp16arith_2x, |
| .init.f16 = xnn_init_f16_minmax_params, |
| .channel_tile = 8, |
| .row_tile = 2, |
| }; |
| xnn_params.f16.hswish = (xnn_univector_ukernel_function) xnn_f16_vhswish_ukernel__neonfp16arith_x16; |
| } |
| #endif // XNN_NO_F16_OPERATORS |
| |
| /**************************** F32 micro-kernels ****************************/ |
| #ifndef XNN_NO_F32_OPERATORS |
| init_flags |= XNN_INIT_FLAG_F32; |
| |
| #if XNN_PLATFORM_IOS || XNN_PLATFORM_MAC |
| #if XNN_ENABLE_ASSEMBLY |
| xnn_params.f32.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_6x8__aarch64_neonfma_prfm_cortex_a75); |
| xnn_params.f32.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_6x8__aarch64_neonfma_prfm_cortex_a75); |
| xnn_params.f32.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_1x8__aarch64_neonfma_prfm_cortex_a75); |
| xnn_params.f32.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_1x8__aarch64_neonfma_prfm_cortex_a75); |
| xnn_params.f32.gemm.init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.gemm.mr = 6; |
| xnn_params.f32.gemm.nr = 8; |
| #else // !XNN_ENABLE_ASSEMBLY |
| xnn_params.f32.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_6x8__neonfma_lane_ld64); |
| xnn_params.f32.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_6x8__neonfma_lane_ld64); |
| xnn_params.f32.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_1x8__neonfma_lane_ld64); |
| xnn_params.f32.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_1x8__neonfma_lane_ld64); |
| xnn_params.f32.gemm.init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.gemm.mr = 6; |
| xnn_params.f32.gemm.nr = 8; |
| #endif // XNN_ENABLE_ASSEMBLY |
| #else // !XNN_PLATFORM_IOS && !XNN_PLATFORM_MAC |
| #if XNN_ENABLE_ASSEMBLY |
| switch (cpuinfo_get_core(0)->uarch) { |
| case cpuinfo_uarch_cortex_a57: |
| xnn_params.f32.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_6x8__aarch64_neonfma_cortex_a75); |
| xnn_params.f32.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_6x8__aarch64_neonfma_cortex_a75); |
| xnn_params.f32.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_1x8__aarch64_neonfma_cortex_a75); |
| xnn_params.f32.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_1x8__aarch64_neonfma_cortex_a75); |
| xnn_params.f32.gemm.init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.gemm.mr = 6; |
| xnn_params.f32.gemm.nr = 8; |
| break; |
| case cpuinfo_uarch_cortex_a72: |
| xnn_params.f32.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_4x8__aarch64_neonfma_prfm_cortex_a75); |
| xnn_params.f32.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_4x8__aarch64_neonfma_prfm_cortex_a75); |
| xnn_params.f32.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_1x8__aarch64_neonfma_prfm_cortex_a75); |
| xnn_params.f32.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_1x8__aarch64_neonfma_prfm_cortex_a75); |
| xnn_params.f32.gemm.init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.gemm.mr = 4; |
| xnn_params.f32.gemm.nr = 8; |
| break; |
| case cpuinfo_uarch_cortex_a75: |
| case cpuinfo_uarch_cortex_a76: |
| case cpuinfo_uarch_exynos_m3: |
| case cpuinfo_uarch_exynos_m4: |
| xnn_params.f32.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_6x8__aarch64_neonfma_prfm_cortex_a75); |
| xnn_params.f32.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_6x8__aarch64_neonfma_prfm_cortex_a75); |
| xnn_params.f32.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_1x8__aarch64_neonfma_prfm_cortex_a75); |
| xnn_params.f32.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_1x8__aarch64_neonfma_prfm_cortex_a75); |
| xnn_params.f32.gemm.init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.gemm.mr = 6; |
| xnn_params.f32.gemm.nr = 8; |
| break; |
| case cpuinfo_uarch_exynos_m1: |
| case cpuinfo_uarch_exynos_m2: |
| xnn_params.f32.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_6x8s4__neonfma); |
| xnn_params.f32.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_6x8s4__neonfma); |
| xnn_params.f32.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_1x8s4__neonfma); |
| xnn_params.f32.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_1x8s4__neonfma); |
| xnn_params.f32.gemm.init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.gemm.mr = 6; |
| xnn_params.f32.gemm.nr = 8; |
| xnn_params.f32.gemm.log2_sr = 2; |
| break; |
| case cpuinfo_uarch_cortex_a53: |
| case cpuinfo_uarch_cortex_a55r0: |
| xnn_params.f32.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_6x8__aarch64_neonfma_cortex_a53); |
| xnn_params.f32.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_6x8__aarch64_neonfma_cortex_a53); |
| xnn_params.f32.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_1x8__aarch64_neonfma_cortex_a53); |
| xnn_params.f32.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_1x8__aarch64_neonfma_cortex_a53); |
| xnn_params.f32.gemm.init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.gemm.mr = 6; |
| xnn_params.f32.gemm.nr = 8; |
| break; |
| case cpuinfo_uarch_cortex_a55: |
| xnn_params.f32.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_6x8__aarch64_neonfma_cortex_a55); |
| xnn_params.f32.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_6x8__aarch64_neonfma_cortex_a55); |
| xnn_params.f32.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_1x8__aarch64_neonfma_cortex_a53); |
| xnn_params.f32.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_1x8__aarch64_neonfma_cortex_a53); |
| xnn_params.f32.gemm.init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.gemm.mr = 6; |
| xnn_params.f32.gemm.nr = 8; |
| break; |
| case cpuinfo_uarch_cortex_a73: |
| xnn_params.f32.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_6x8__aarch64_neonfma_cortex_a73); |
| xnn_params.f32.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_6x8__aarch64_neonfma_cortex_a73); |
| xnn_params.f32.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_1x8__aarch64_neonfma_prfm_cortex_a75); |
| xnn_params.f32.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_1x8__aarch64_neonfma_prfm_cortex_a75); |
| xnn_params.f32.gemm.init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.gemm.mr = 6; |
| xnn_params.f32.gemm.nr = 8; |
| break; |
| case cpuinfo_uarch_cortex_a77: |
| case cpuinfo_uarch_exynos_m5: |
| case cpuinfo_uarch_kryo: |
| xnn_params.f32.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_4x8__aarch64_neonfma_cortex_a75); |
| xnn_params.f32.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_4x8__aarch64_neonfma_cortex_a75); |
| xnn_params.f32.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_1x8__aarch64_neonfma_cortex_a75); |
| xnn_params.f32.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_1x8__aarch64_neonfma_cortex_a75); |
| xnn_params.f32.gemm.init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.gemm.mr = 4; |
| xnn_params.f32.gemm.nr = 8; |
| break; |
| case cpuinfo_uarch_cortex_a78: |
| case cpuinfo_uarch_cortex_x1: |
| default: |
| xnn_params.f32.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_6x8__aarch64_neonfma_ld128); |
| xnn_params.f32.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_6x8__aarch64_neonfma_ld128); |
| xnn_params.f32.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_1x8__neonfma_lane_ld64); |
| xnn_params.f32.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_1x8__neonfma_lane_ld64); |
| xnn_params.f32.gemm.init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.gemm.mr = 6; |
| xnn_params.f32.gemm.nr = 8; |
| break; |
| } |
| #if XNN_MAX_UARCH_TYPES > 1 |
| { |
| /* Choose micro-kernels for little cores according to micro-kernel specification for the big core */ |
| const uint32_t mr = xnn_params.f32.gemm.mr; |
| const uint32_t nr = xnn_params.f32.gemm.nr; |
| const uint32_t log2_sr = xnn_params.f32.gemm.log2_sr; |
| for (size_t i = 1; i < XNN_MAX_UARCH_TYPES; i++) { |
| const struct cpuinfo_uarch_info* uarch_info = cpuinfo_get_uarch(i); |
| if (uarch_info == NULL) { |
| /* No more microarchitectures in the system */ |
| break; |
| } |
| |
| switch (uarch_info->uarch) { |
| case cpuinfo_uarch_cortex_a53: |
| case cpuinfo_uarch_cortex_a55r0: |
| if (mr == 6 && nr == 8 && log2_sr == 0) { |
| xnn_params.f32.gemm.minmax.gemm.function[i] = (xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_6x8__aarch64_neonfma_cortex_a53; |
| xnn_params.f32.gemm.minmax.igemm.function[i] = (xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_6x8__aarch64_neonfma_cortex_a53; |
| xnn_params.f32.gemm.minmax.gemm1.function[i] = (xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_1x8__aarch64_neonfma_cortex_a53; |
| xnn_params.f32.gemm.minmax.igemm1.function[i] = (xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_1x8__aarch64_neonfma_cortex_a53; |
| } else if (mr == 4 && nr == 8 && log2_sr == 0) { |
| xnn_params.f32.gemm.minmax.gemm.function[i] = (xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_4x8__aarch64_neonfma_cortex_a53; |
| xnn_params.f32.gemm.minmax.igemm.function[i] = (xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_4x8__aarch64_neonfma_cortex_a53; |
| xnn_params.f32.gemm.minmax.gemm1.function[i] = (xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_1x8__aarch64_neonfma_cortex_a53; |
| xnn_params.f32.gemm.minmax.igemm1.function[i] = (xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_1x8__aarch64_neonfma_cortex_a53; |
| } |
| break; |
| case cpuinfo_uarch_cortex_a55: |
| if (mr == 6 && nr == 8 && log2_sr == 0) { |
| xnn_params.f32.gemm.minmax.gemm.function[i] = (xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_6x8__aarch64_neonfma_cortex_a55; |
| xnn_params.f32.gemm.minmax.igemm.function[i] = (xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_6x8__aarch64_neonfma_cortex_a55; |
| xnn_params.f32.gemm.minmax.gemm1.function[i] = (xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_1x8__aarch64_neonfma_cortex_a53; |
| xnn_params.f32.gemm.minmax.igemm1.function[i] = (xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_1x8__aarch64_neonfma_cortex_a53; |
| } else if (mr == 4 && nr == 8 && log2_sr == 0) { |
| xnn_params.f32.gemm.minmax.gemm.function[i] = (xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_4x8__aarch64_neonfma_cortex_a55; |
| xnn_params.f32.gemm.minmax.igemm.function[i] = (xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_4x8__aarch64_neonfma_cortex_a55; |
| xnn_params.f32.gemm.minmax.gemm1.function[i] = (xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_1x8__aarch64_neonfma_cortex_a53; |
| xnn_params.f32.gemm.minmax.igemm1.function[i] = (xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_1x8__aarch64_neonfma_cortex_a53; |
| } |
| break; |
| default: |
| break; |
| } |
| } |
| } |
| #endif // XNN_MAX_UARCH_TYPES > 1 |
| #else // !XNN_ENABLE_ASSEMBLY |
| xnn_params.f32.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_6x8__neonfma_lane_ld64); |
| xnn_params.f32.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_6x8__neonfma_lane_ld64); |
| xnn_params.f32.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_1x8__neonfma_lane_ld64); |
| xnn_params.f32.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_1x8__neonfma_lane_ld64); |
| xnn_params.f32.gemm.init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.gemm.mr = 6; |
| xnn_params.f32.gemm.nr = 8; |
| #endif // XNN_ENABLE_ASSEMBLY |
| #endif // XNN_PLATFORM_IOS || XNN_PLATFORM_MAC |
| xnn_params.f32.gemm2.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_4x2__neonfma_lane_ld64); |
| xnn_params.f32.gemm2.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_4x2__neonfma_lane_ld64); |
| xnn_params.f32.gemm2.init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.gemm2.mr = 4; |
| xnn_params.f32.gemm2.nr = 2; |
| |
| xnn_params.f32.dwconv[0].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_minmax_ukernel_up8x3__neonfma; |
| xnn_params.f32.dwconv[0].init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.dwconv[0].channel_tile = 8; |
| xnn_params.f32.dwconv[0].primary_tile = 3; |
| |
| xnn_params.f32.dwconv[1].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_minmax_ukernel_up8x4__neonfma; |
| xnn_params.f32.dwconv[1].init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.dwconv[1].channel_tile = 8; |
| xnn_params.f32.dwconv[1].primary_tile = 4; |
| |
| #if XNN_PLATFORM_IOS || XNN_PLATFORM_MAC |
| xnn_params.f32.dwconv[2].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_minmax_ukernel_up8x9__neonfma; |
| xnn_params.f32.dwconv[2].init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.dwconv[2].channel_tile = 8; |
| xnn_params.f32.dwconv[2].primary_tile = 9; |
| #else // !XNN_PLATFORM_IOS && !XNN_PLATFORM_MAC |
| switch (cpuinfo_get_core(0)->uarch) { |
| case cpuinfo_uarch_kryo: |
| xnn_params.f32.dwconv[2].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_minmax_ukernel_up8x9__neonfma; |
| xnn_params.f32.dwconv[2].init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.dwconv[2].channel_tile = 8; |
| xnn_params.f32.dwconv[2].primary_tile = 9; |
| break; |
| #if XNN_ENABLE_ASSEMBLY |
| case cpuinfo_uarch_cortex_a53: |
| case cpuinfo_uarch_cortex_a55r0: |
| case cpuinfo_uarch_cortex_a55: |
| xnn_params.f32.dwconv[2].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_minmax_ukernel_up4x9__aarch64_neonfma_cortex_a55; |
| xnn_params.f32.dwconv[2].init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.dwconv[2].channel_tile = 4; |
| xnn_params.f32.dwconv[2].primary_tile = 9; |
| break; |
| #endif // XNN_ENABLE_ASSEMBLY |
| default: |
| xnn_params.f32.dwconv[2].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_minmax_ukernel_up8x9__neonfma; |
| xnn_params.f32.dwconv[2].init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.dwconv[2].channel_tile = 8; |
| xnn_params.f32.dwconv[2].primary_tile = 9; |
| break; |
| } |
| #endif // XNN_PLATFORM_IOS && XNN_PLATFORM_MAC |
| |
| xnn_params.f32.dwconv[3].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_minmax_ukernel_up8x25__neonfma_acc2; |
| xnn_params.f32.dwconv[3].init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.dwconv[3].channel_tile = 8; |
| xnn_params.f32.dwconv[3].primary_tile = 25; |
| |
| xnn_params.f32.avgpool = (struct avgpool_parameters) { |
| .up = (xnn_avgpool_unipass_ukernel_function) xnn_f32_avgpool_minmax_ukernel_9x__neon_c4, |
| .mp = (xnn_avgpool_multipass_ukernel_function) xnn_f32_avgpool_minmax_ukernel_9p8x__neon_c4, |
| .mr = 9, |
| .qr = 8, |
| }; |
| xnn_params.f32.pavgpool = (struct pavgpool_parameters) { |
| .up = (xnn_pavgpool_unipass_ukernel_function) xnn_f32_pavgpool_minmax_ukernel_9x__neon_c4, |
| .mp = (xnn_pavgpool_multipass_ukernel_function) xnn_f32_pavgpool_minmax_ukernel_9p8x__neon_c4, |
| .mr = 9, |
| .qr = 8, |
| }; |
| xnn_params.f32.gavgpool = (struct gavgpool_parameters) { |
| .up = (xnn_gavgpool_unipass_ukernel_function) xnn_f32_gavgpool_minmax_ukernel_7x__neon_c4, |
| .mp = (xnn_gavgpool_multipass_ukernel_function) xnn_f32_gavgpool_minmax_ukernel_7p7x__neon_c4, |
| .mr = 7, |
| }; |
| xnn_params.f32.maxpool = (struct maxpool_parameters) { |
| .ukernel = (xnn_maxpool_ukernel_function) xnn_f32_maxpool_minmax_ukernel_9p8x__neon_c4, |
| .init.f32 = xnn_init_f32_minmax_scalar_params, |
| .mr = 9, |
| .qr = 8, |
| }; |
| xnn_params.f32.argmaxpool[0] = (struct argmaxpool_parameters) { |
| .up = (xnn_argmaxpool_unipass_ukernel_function) xnn_f32_argmaxpool_ukernel_4x__neon_c4, |
| .mr = 4, |
| }; |
| xnn_params.f32.argmaxpool[1] = (struct argmaxpool_parameters) { |
| .up = (xnn_argmaxpool_unipass_ukernel_function) xnn_f32_argmaxpool_ukernel_9x__neon_c4, |
| .mr = 9, |
| }; |
| xnn_params.f32.argmaxpool[2] = (struct argmaxpool_parameters) { |
| .mp = (xnn_argmaxpool_multipass_ukernel_function) xnn_f32_argmaxpool_ukernel_9p8x__neon_c4, |
| .mr = 9, |
| .qr = 8, |
| }; |
| xnn_params.f32.ibilinear = (struct ibilinear_parameters) { |
| .ukernel = (xnn_ibilinear_ukernel_function) xnn_f32_ibilinear_ukernel__neonfma_c8, |
| .pixel_tile = 1, |
| .channel_tile = 8, |
| }; |
| xnn_params.f32.abs = (xnn_univector_ukernel_function) xnn_f32_vabs_ukernel__neon_x8; |
| xnn_params.f32.clamp = (struct vunary_parameters) { |
| .ukernel = (xnn_univector_ukernel_function) xnn_f32_vclamp_ukernel__neon_x8, |
| .init.f32_minmax = xnn_init_f32_minmax_scalar_params, |
| .element_tile = 8, |
| }; |
| xnn_params.f32.elu = (xnn_univector_ukernel_function) xnn_f32_velu_ukernel__neonfma_rr1_lut16_p3_x16; |
| xnn_params.f32.hswish = (xnn_univector_ukernel_function) xnn_f32_vhswish_ukernel__neon_x16; |
| xnn_params.f32.lrelu = (xnn_univector_ukernel_function) xnn_f32_vlrelu_ukernel__neon_x8; |
| xnn_params.f32.neg = (xnn_univector_ukernel_function) xnn_f32_vneg_ukernel__neon_x8; |
| xnn_params.f32.rndne = (xnn_univector_ukernel_function) xnn_f32_vrndne_ukernel__neonv8_x8; |
| xnn_params.f32.rndz = (xnn_univector_ukernel_function) xnn_f32_vrndz_ukernel__neonv8_x8; |
| xnn_params.f32.rndu = (xnn_univector_ukernel_function) xnn_f32_vrndu_ukernel__neonv8_x8; |
| xnn_params.f32.rndd = (xnn_univector_ukernel_function) xnn_f32_vrndd_ukernel__neonv8_x8; |
| xnn_params.f32.sigmoid = (xnn_univector_ukernel_function) xnn_f32_vsigmoid_ukernel__neonfma_rr1_lut64_p2_nr2recps_x16; |
| xnn_params.f32.sqr = (xnn_univector_ukernel_function) xnn_f32_vsqr_ukernel__neon_x8; |
| xnn_params.f32.sqrt = (xnn_univector_ukernel_function) xnn_f32_vsqrt_ukernel__neon_sqrt_x4; |
| xnn_params.f32.prelu = (struct prelu_parameters) { |
| .ukernel = (xnn_prelu_ukernel_function) xnn_f32_prelu_ukernel__neon_2x8, |
| .row_tile = 2, |
| .channel_tile = 8, |
| }; |
| xnn_params.f32.raddstoreexpminusmax = xnn_f32_raddstoreexpminusmax_ukernel__neonfma_lut64_p2_x16; |
| xnn_params.f32.rmax = xnn_f32_rmax_ukernel__neon; |
| xnn_params.f32.vadd = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vadd_minmax_ukernel__neon_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vaddc_minmax_ukernel__neon_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vaddc_minmax_ukernel__neon_x8, |
| .element_tile = 8, |
| }; |
| xnn_params.f32.vdiv = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vdiv_minmax_ukernel__neon_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vdivc_minmax_ukernel__neon_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vrdivc_minmax_ukernel__neon_x8, |
| .element_tile = 8, |
| }; |
| xnn_params.f32.vmax = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmax_ukernel__neon_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmaxc_ukernel__neon_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmaxc_ukernel__neon_x8, |
| .element_tile = 8, |
| }; |
| xnn_params.f32.vmin = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmin_ukernel__neon_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vminc_ukernel__neon_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vminc_ukernel__neon_x8, |
| .element_tile = 8, |
| }; |
| xnn_params.f32.vmul = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmul_minmax_ukernel__neon_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmulc_minmax_ukernel__neon_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmulc_minmax_ukernel__neon_x8, |
| .element_tile = 8, |
| }; |
| xnn_params.f32.vsub = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsub_minmax_ukernel__neon_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsubc_minmax_ukernel__neon_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vrsubc_minmax_ukernel__neon_x8, |
| .element_tile = 8, |
| }; |
| xnn_params.f32.vsqrdiff = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsqrdiff_ukernel__neon_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsqrdiffc_ukernel__neon_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsqrdiffc_ukernel__neon_x8, |
| .element_tile = 8, |
| }; |
| xnn_params.f32.vmulcaddc = (struct vmulcaddc_parameters) { |
| .ukernel = (xnn_vmulcaddc_ukernel_function) xnn_f32_vmulcaddc_minmax_ukernel_c4__neonfma_2x, |
| .init.f32 = xnn_init_f32_minmax_scalar_params, |
| .channel_tile = 4, |
| .row_tile = 2, |
| }; |
| #ifndef XNN_NO_NCHW_OPERATORS |
| init_flags |= XNN_INIT_FLAG_CHW_OPT; |
| |
| xnn_params.f32.spmm = (struct spmm_parameters) { |
| .ukernel = (xnn_spmm_ukernel_function) xnn_f32_spmm_minmax_ukernel_32x1__neonfma_pipelined, |
| .mr = 32, |
| .nr = 1, |
| }; |
| xnn_params.f32.spmm2 = (struct spmm_parameters) { |
| .ukernel = (xnn_spmm_ukernel_function) xnn_f32_spmm_minmax_ukernel_32x2__neonfma, |
| .mr = 32, |
| .nr = 2, |
| }; |
| xnn_params.f32.spmm4 = (struct spmm_parameters) { |
| .ukernel = (xnn_spmm_ukernel_function) xnn_f32_spmm_minmax_ukernel_32x4__neonfma, |
| .mr = 32, |
| .nr = 4, |
| }; |
| xnn_params.f32.conv_hwc2chw_3x3c3s2 = (struct conv_hwc2chw_parameters) { |
| .ukernel_with_symm_padding = |
| (xnn_conv_hwc2chw_ukernel_function) xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__neonfma_2x2, |
| .output_channel_tile = 4, |
| .output_height_tile = 2, |
| .output_width_tile = 2, |
| }; |
| xnn_params.f32.dwconv2d_chw_3x3 = (struct dwconv2d_chw_parameters) { |
| .ukernel = (xnn_dwconv2d_chw_ukernel_function) xnn_f32_dwconv2d_chw_ukernel_3x3p1__neonfma_3x4, |
| .output_width_tile = 4, |
| .output_height_tile = 3, |
| }; |
| xnn_params.f32.dwconv2d_chw_3x3s2 = (struct dwconv2d_chw_parameters) { |
| .ukernel = (xnn_dwconv2d_chw_ukernel_function) xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__neonfma_2x4_acc2, |
| .output_width_tile = 4, |
| .output_height_tile = 2, |
| }; |
| xnn_params.f32.dwconv2d_chw_5x5 = (struct dwconv2d_chw_parameters) { |
| .ukernel = (xnn_dwconv2d_chw_ukernel_function) xnn_f32_dwconv2d_chw_ukernel_5x5p2__neonfma_4x4, |
| .output_width_tile = 4, |
| .output_height_tile = 4, |
| }; |
| xnn_params.f32.dwconv2d_chw_5x5s2 = (struct dwconv2d_chw_parameters) { |
| .ukernel = (xnn_dwconv2d_chw_ukernel_function) xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__neonfma_1x4_acc2, |
| .output_width_tile = 4, |
| .output_height_tile = 1, |
| }; |
| xnn_params.f32.gavgpool_cw = (struct gavgpool_cw_parameters) { |
| .ukernel = (xnn_gavgpool_cw_ukernel_function) xnn_f32_gavgpool_cw_ukernel__neon_x4, |
| .channel_tile = 4, |
| }; |
| xnn_params.f32.ibilinear_chw = (struct ibilinear_chw_parameters) { |
| .ukernel = (xnn_ibilinear_chw_ukernel_function) xnn_f32_ibilinear_chw_ukernel__neonfma_p8, |
| .channel_tile = 1, |
| .pixel_tile = 8, |
| }; |
| #endif // XNN_NO_NCHW_OPERATORS |
| #endif // XNN_NO_F32_OPERATORS |
| |
| /*************************** VCVT micro-kernels ***************************/ |
| #ifndef XNN_NO_VCVT_OPERATORS |
| init_flags |= XNN_INIT_FLAG_VCVT; |
| |
| xnn_params.vcvt.f16_to_f32 = (xnn_univector_ukernel_function) xnn_f16_f32_vcvt_ukernel__neonfp16_x16; |
| #endif // XNN_NO_VCVT_OPERATORS |
| |
| /**************************** X32 micro-kernels ****************************/ |
| #ifndef XNN_NO_X32_OPERATORS |
| init_flags |= XNN_INIT_FLAG_X32; |
| |
| xnn_params.x32.unpool = (xnn_unpool_ukernel_function) xnn_x32_unpool_ukernel__neon; |
| xnn_params.x32.zip = (struct zip_parameters) { |
| .x2 = (xnn_zipc_ukernel_function) xnn_x32_zip_x2_ukernel__neon, |
| .x3 = (xnn_zipc_ukernel_function) xnn_x32_zip_x3_ukernel__neon, |
| .x4 = (xnn_zipc_ukernel_function) xnn_x32_zip_x4_ukernel__neon, |
| .xm = (xnn_zipv_ukernel_function) xnn_x32_zip_xm_ukernel__neon, |
| }; |
| #ifndef XNN_NO_NCHW_OPERATORS |
| xnn_params.x32.depthtospace2d_chw2hwc = (struct depthtospace2d_chw2hwc_parameters) { |
| .ukernel = (xnn_depthtospace2d_chw2hwc_ukernel_function) xnn_x32_depthtospace2d_chw2hwc_ukernel__scalar, |
| .channel_tile = 1, |
| .pixel_tile = 1, |
| }; |
| #endif // XNN_NO_NCHW_OPERATORS |
| #endif // XNN_NO_X32_OPERATORS |
| |
| /**************************** XX micro-kernels ****************************/ |
| #ifndef XNN_NO_XX_OPERATORS |
| init_flags |= XNN_INIT_FLAG_XX; |
| |
| xnn_params.xx.copy = (xnn_univector_ukernel_function) xnn_xx_copy_ukernel__memcpy; |
| xnn_params.xx.fill = (struct fill_parameters) { |
| .ukernel = (xnn_fill_ukernel_function) xnn_xx_fill_ukernel__neon_x64, |
| .row_tile = 1, |
| }; |
| xnn_params.xx.pad = (struct pad_parameters) { |
| .ukernel = (xnn_pad_ukernel_function) xnn_xx_pad_ukernel__neon, |
| .row_tile = 1, |
| }; |
| #endif |
| |
| #elif XNN_ARCH_X86 || XNN_ARCH_X86_64 |
| if (!cpuinfo_has_x86_sse2()) { |
| xnn_log_error("XNNPACK initialization failed: SSE2 is not supported"); |
| return; |
| } |
| |
| /**************************** QC8 micro-kernels ****************************/ |
| #ifndef XNN_NO_QC8_OPERATORS |
| init_flags |= XNN_INIT_FLAG_QC8; |
| |
| if (cpuinfo_has_x86_avx512f() && cpuinfo_has_x86_avx512bw() && cpuinfo_has_x86_avx512dq() && cpuinfo_has_x86_avx512vl()) { |
| xnn_params.qc8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_4x16c8__avx512skx); |
| xnn_params.qc8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_4x16c8__avx512skx); |
| xnn_params.qc8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_1x16c8__avx512skx); |
| xnn_params.qc8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_1x16c8__avx512skx); |
| xnn_params.qc8.gemm.init.qc8 = xnn_init_qs8_minmax_avx512_params; |
| xnn_params.qc8.gemm.mr = 4; |
| xnn_params.qc8.gemm.nr = 16; |
| xnn_params.qc8.gemm.log2_kr = 3; |
| } else if (cpuinfo_has_x86_xop()) { |
| // XOP should be checked before AVX2: AMD Excavator supports both, but performs better with XOP microkernels |
| xnn_params.qc8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_2x4c8__xop_ld64); |
| xnn_params.qc8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_2x4c8__xop_ld64); |
| xnn_params.qc8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_1x4c8__xop_ld64); |
| xnn_params.qc8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_1x4c8__xop_ld64); |
| xnn_params.qc8.gemm.init.qc8 = xnn_init_qs8_minmax_sse4_params; |
| xnn_params.qc8.gemm.mr = 2; |
| xnn_params.qc8.gemm.nr = 4; |
| xnn_params.qc8.gemm.log2_kr = 3; |
| } else if (cpuinfo_has_x86_avx2()) { |
| xnn_params.qc8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_3x8c8__avx2); |
| xnn_params.qc8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_3x8c8__avx2); |
| xnn_params.qc8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_1x8c8__avx2); |
| xnn_params.qc8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_1x8c8__avx2); |
| xnn_params.qc8.gemm.init.qc8 = xnn_init_qs8_minmax_avx2_params; |
| xnn_params.qc8.gemm.mr = 3; |
| xnn_params.qc8.gemm.nr = 8; |
| xnn_params.qc8.gemm.log2_kr = 3; |
| } else if (cpuinfo_has_x86_avx()) { |
| xnn_params.qc8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_2x4c8__avx_ld128); |
| xnn_params.qc8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_2x4c8__avx_ld128); |
| xnn_params.qc8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_1x4c8__avx_ld128); |
| xnn_params.qc8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_1x4c8__avx_ld128); |
| xnn_params.qc8.gemm.init.qc8 = xnn_init_qs8_minmax_sse4_params; |
| xnn_params.qc8.gemm.mr = 2; |
| xnn_params.qc8.gemm.nr = 4; |
| xnn_params.qc8.gemm.log2_kr = 3; |
| } else if (cpuinfo_has_x86_sse4_1()) { |
| xnn_params.qc8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_3x4c8__sse41_ld64); |
| xnn_params.qc8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_3x4c8__sse41_ld64); |
| xnn_params.qc8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_1x4c8__sse41_ld64); |
| xnn_params.qc8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_1x4c8__sse41_ld64); |
| xnn_params.qc8.gemm.init.qc8 = xnn_init_qs8_minmax_sse4_params; |
| xnn_params.qc8.gemm.mr = 3; |
| xnn_params.qc8.gemm.nr = 4; |
| xnn_params.qc8.gemm.log2_kr = 3; |
| } else { |
| xnn_params.qc8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_3x4c8__sse2_ld64); |
| xnn_params.qc8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_3x4c8__sse2_ld64); |
| xnn_params.qc8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_1x4c8__sse2_ld64); |
| xnn_params.qc8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_1x4c8__sse2_ld64); |
| xnn_params.qc8.gemm.init.qc8 = xnn_init_qs8_minmax_sse2_params; |
| xnn_params.qc8.gemm.mr = 3; |
| xnn_params.qc8.gemm.nr = 4; |
| xnn_params.qc8.gemm.log2_kr = 3; |
| } |
| |
| if (cpuinfo_has_x86_avx512f() && cpuinfo_has_x86_avx512bw() && cpuinfo_has_x86_avx512dq() && cpuinfo_has_x86_avx512vl()) { |
| xnn_params.qc8.dwconv[0].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qc8_dwconv_minmax_fp32_ukernel_up32x9__avx512skx_mul32; |
| xnn_params.qc8.dwconv[0].init.qc8 = xnn_init_qs8_minmax_avx512_params; |
| xnn_params.qc8.dwconv[0].channel_tile = 32; |
| xnn_params.qc8.dwconv[1].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qc8_dwconv_minmax_fp32_ukernel_up32x25__avx512skx_mul32; |
| xnn_params.qc8.dwconv[1].init.qc8 = xnn_init_qs8_minmax_avx512_params; |
| xnn_params.qc8.dwconv[1].channel_tile = 32; |
| } else if (cpuinfo_has_x86_xop()) { |
| // XOP should be checked before AVX2: AMD Excavator supports both, but performs better with XOP microkernels |
| xnn_params.qc8.dwconv[0].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qc8_dwconv_minmax_fp32_ukernel_up16x9__xop_mul16_add16; |
| xnn_params.qc8.dwconv[0].init.qc8 = xnn_init_qs8_minmax_sse4_params; |
| xnn_params.qc8.dwconv[0].channel_tile = 16; |
| xnn_params.qc8.dwconv[1].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qc8_dwconv_minmax_fp32_ukernel_up16x25__xop_mul16_add16; |
| xnn_params.qc8.dwconv[1].init.qc8 = xnn_init_qs8_minmax_sse4_params; |
| xnn_params.qc8.dwconv[1].channel_tile = 16; |
| } else if (cpuinfo_has_x86_avx2()) { |
| xnn_params.qc8.dwconv[0].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qc8_dwconv_minmax_fp32_ukernel_up16x9__avx2_mul32; |
| xnn_params.qc8.dwconv[0].init.qc8 = xnn_init_qs8_minmax_avx2_params; |
| xnn_params.qc8.dwconv[0].channel_tile = 16; |
| xnn_params.qc8.dwconv[1].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qc8_dwconv_minmax_fp32_ukernel_up16x25__avx2_mul32; |
| xnn_params.qc8.dwconv[1].init.qc8 = xnn_init_qs8_minmax_avx2_params; |
| xnn_params.qc8.dwconv[1].channel_tile = 16; |
| } else if (cpuinfo_has_x86_avx()) { |
| xnn_params.qc8.dwconv[0].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qc8_dwconv_minmax_fp32_ukernel_up16x9__avx_mul16_add16; |
| xnn_params.qc8.dwconv[0].init.qc8 = xnn_init_qs8_minmax_sse4_params; |
| xnn_params.qc8.dwconv[0].channel_tile = 16; |
| xnn_params.qc8.dwconv[1].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qc8_dwconv_minmax_fp32_ukernel_up16x25__avx_mul16_add16; |
| xnn_params.qc8.dwconv[1].init.qc8 = xnn_init_qs8_minmax_sse4_params; |
| xnn_params.qc8.dwconv[1].channel_tile = 16; |
| } else if (cpuinfo_has_x86_sse4_1()) { |
| xnn_params.qc8.dwconv[0].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qc8_dwconv_minmax_fp32_ukernel_up8x9__sse41_mul16; |
| xnn_params.qc8.dwconv[0].init.qc8 = xnn_init_qs8_minmax_sse4_params; |
| xnn_params.qc8.dwconv[0].channel_tile = 8; |
| xnn_params.qc8.dwconv[1].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qc8_dwconv_minmax_fp32_ukernel_up8x25__sse41_mul16; |
| xnn_params.qc8.dwconv[1].init.qc8 = xnn_init_qs8_minmax_sse4_params; |
| xnn_params.qc8.dwconv[1].channel_tile = 8; |
| } else if (cpuinfo_has_x86_sse2()) { |
| xnn_params.qc8.dwconv[0].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qc8_dwconv_minmax_fp32_ukernel_up8x9__sse2_mul16; |
| xnn_params.qc8.dwconv[0].init.qc8 = xnn_init_qs8_minmax_sse2_params; |
| xnn_params.qc8.dwconv[0].channel_tile = 8; |
| xnn_params.qc8.dwconv[1].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qc8_dwconv_minmax_fp32_ukernel_up8x25__sse2_mul16; |
| xnn_params.qc8.dwconv[1].init.qc8 = xnn_init_qs8_minmax_sse2_params; |
| xnn_params.qc8.dwconv[1].channel_tile = 8; |
| } |
| xnn_params.qc8.dwconv[0].primary_tile = 9; |
| xnn_params.qc8.dwconv[1].primary_tile = 25; |
| #endif // XNN_NO_QC8_OPERATORS |
| |
| /**************************** QS8 micro-kernels ****************************/ |
| #ifndef XNN_NO_QS8_OPERATORS |
| init_flags |= XNN_INIT_FLAG_QS8; |
| |
| if (cpuinfo_has_x86_avx512f() && cpuinfo_has_x86_avx512bw() && cpuinfo_has_x86_avx512dq() && cpuinfo_has_x86_avx512vl()) { |
| xnn_params.qs8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_fp32_ukernel_4x16c8__avx512skx); |
| xnn_params.qs8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_fp32_ukernel_4x16c8__avx512skx); |
| xnn_params.qs8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_fp32_ukernel_1x16c8__avx512skx); |
| xnn_params.qs8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_fp32_ukernel_1x16c8__avx512skx); |
| xnn_params.qs8.gemm.init.qs8 = xnn_init_qs8_conv_minmax_fp32_avx512_params; |
| xnn_params.qs8.gemm.mr = 4; |
| xnn_params.qs8.gemm.nr = 16; |
| xnn_params.qs8.gemm.log2_kr = 3; |
| } else if (cpuinfo_has_x86_xop()) { |
| // XOP should be checked before AVX2: AMD Excavator supports both, but performs better with XOP microkernels |
| xnn_params.qs8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_fp32_ukernel_2x4c8__xop_ld64); |
| xnn_params.qs8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_fp32_ukernel_2x4c8__xop_ld64); |
| xnn_params.qs8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_fp32_ukernel_1x4c8__xop_ld64); |
| xnn_params.qs8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_fp32_ukernel_1x4c8__xop_ld64); |
| xnn_params.qs8.gemm.init.qs8 = xnn_init_qs8_conv_minmax_fp32_sse4_params; |
| xnn_params.qs8.gemm.mr = 2; |
| xnn_params.qs8.gemm.nr = 4; |
| xnn_params.qs8.gemm.log2_kr = 3; |
| } else if (cpuinfo_has_x86_avx2()) { |
| xnn_params.qs8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_fp32_ukernel_3x8c8__avx2); |
| xnn_params.qs8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_fp32_ukernel_3x8c8__avx2); |
| xnn_params.qs8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_fp32_ukernel_1x8c8__avx2); |
| xnn_params.qs8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_fp32_ukernel_1x8c8__avx2); |
| xnn_params.qs8.gemm.init.qs8 = xnn_init_qs8_conv_minmax_fp32_avx2_params; |
| xnn_params.qs8.gemm.mr = 3; |
| xnn_params.qs8.gemm.nr = 8; |
| xnn_params.qs8.gemm.log2_kr = 3; |
| } else if (cpuinfo_has_x86_avx()) { |
| xnn_params.qs8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_fp32_ukernel_2x4c8__avx_ld128); |
| xnn_params.qs8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_fp32_ukernel_2x4c8__avx_ld128); |
| xnn_params.qs8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_fp32_ukernel_1x4c8__avx_ld128); |
| xnn_params.qs8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_fp32_ukernel_1x4c8__avx_ld128); |
| xnn_params.qs8.gemm.init.qs8 = xnn_init_qs8_conv_minmax_fp32_sse4_params; |
| xnn_params.qs8.gemm.mr = 2; |
| xnn_params.qs8.gemm.nr = 4; |
| xnn_params.qs8.gemm.log2_kr = 3; |
| } else if (cpuinfo_has_x86_sse4_1()) { |
| xnn_params.qs8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_fp32_ukernel_3x4c8__sse41_ld64); |
| xnn_params.qs8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_fp32_ukernel_3x4c8__sse41_ld64); |
| xnn_params.qs8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_fp32_ukernel_1x4c8__sse41_ld64); |
| xnn_params.qs8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_fp32_ukernel_1x4c8__sse41_ld64); |
| xnn_params.qs8.gemm.init.qs8 = xnn_init_qs8_conv_minmax_fp32_sse4_params; |
| xnn_params.qs8.gemm.mr = 3; |
| xnn_params.qs8.gemm.nr = 4; |
| xnn_params.qs8.gemm.log2_kr = 3; |
| } else { |
| xnn_params.qs8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_fp32_ukernel_3x4c8__sse2_ld64); |
| xnn_params.qs8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_fp32_ukernel_3x4c8__sse2_ld64); |
| xnn_params.qs8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_fp32_ukernel_1x4c8__sse2_ld64); |
| xnn_params.qs8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_fp32_ukernel_1x4c8__sse2_ld64); |
| xnn_params.qs8.gemm.init.qs8 = xnn_init_qs8_conv_minmax_fp32_sse2_params; |
| xnn_params.qs8.gemm.mr = 3; |
| xnn_params.qs8.gemm.nr = 4; |
| xnn_params.qs8.gemm.log2_kr = 3; |
| } |
| |
| if (cpuinfo_has_x86_avx512f() && cpuinfo_has_x86_avx512bw() && cpuinfo_has_x86_avx512dq() && cpuinfo_has_x86_avx512vl()) { |
| xnn_params.qs8.dwconv[0].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qs8_dwconv_minmax_fp32_ukernel_up32x9__avx512skx_mul32; |
| xnn_params.qs8.dwconv[0].init.qs8 = xnn_init_qs8_conv_minmax_fp32_avx512_params; |
| xnn_params.qs8.dwconv[0].channel_tile = 32; |
| xnn_params.qs8.dwconv[1].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qs8_dwconv_minmax_fp32_ukernel_up32x25__avx512skx_mul32; |
| xnn_params.qs8.dwconv[1].init.qs8 = xnn_init_qs8_conv_minmax_fp32_avx512_params; |
| xnn_params.qs8.dwconv[1].channel_tile = 32; |
| } else if (cpuinfo_has_x86_xop()) { |
| // XOP should be checked before AVX2: AMD Excavator supports both, but performs better with XOP microkernels |
| xnn_params.qs8.dwconv[0].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qs8_dwconv_minmax_fp32_ukernel_up16x9__xop_mul16_add16; |
| xnn_params.qs8.dwconv[0].init.qs8 = xnn_init_qs8_conv_minmax_fp32_sse4_params; |
| xnn_params.qs8.dwconv[0].channel_tile = 16; |
| xnn_params.qs8.dwconv[1].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qs8_dwconv_minmax_fp32_ukernel_up16x25__xop_mul16_add16; |
| xnn_params.qs8.dwconv[1].init.qs8 = xnn_init_qs8_conv_minmax_fp32_sse4_params; |
| xnn_params.qs8.dwconv[1].channel_tile = 16; |
| } else if (cpuinfo_has_x86_avx2()) { |
| xnn_params.qs8.dwconv[0].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qs8_dwconv_minmax_fp32_ukernel_up16x9__avx2_mul32; |
| xnn_params.qs8.dwconv[0].init.qs8 = xnn_init_qs8_conv_minmax_fp32_avx2_params; |
| xnn_params.qs8.dwconv[0].channel_tile = 16; |
| xnn_params.qs8.dwconv[1].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qs8_dwconv_minmax_fp32_ukernel_up16x25__avx2_mul32; |
| xnn_params.qs8.dwconv[1].init.qs8 = xnn_init_qs8_conv_minmax_fp32_avx2_params; |
| xnn_params.qs8.dwconv[1].channel_tile = 16; |
| } else if (cpuinfo_has_x86_avx()) { |
| xnn_params.qs8.dwconv[0].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qs8_dwconv_minmax_fp32_ukernel_up16x9__avx_mul16_add16; |
| xnn_params.qs8.dwconv[0].init.qs8 = xnn_init_qs8_conv_minmax_fp32_sse4_params; |
| xnn_params.qs8.dwconv[0].channel_tile = 16; |
| xnn_params.qs8.dwconv[1].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qs8_dwconv_minmax_fp32_ukernel_up16x25__avx_mul16_add16; |
| xnn_params.qs8.dwconv[1].init.qs8 = xnn_init_qs8_conv_minmax_fp32_sse4_params; |
| xnn_params.qs8.dwconv[1].channel_tile = 16; |
| } else if (cpuinfo_has_x86_sse4_1()) { |
| xnn_params.qs8.dwconv[0].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qs8_dwconv_minmax_fp32_ukernel_up8x9__sse41_mul16_add16; |
| xnn_params.qs8.dwconv[0].init.qs8 = xnn_init_qs8_conv_minmax_fp32_sse4_params; |
| xnn_params.qs8.dwconv[0].channel_tile = 8; |
| xnn_params.qs8.dwconv[1].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qs8_dwconv_minmax_fp32_ukernel_up8x25__sse41_mul16_add16; |
| xnn_params.qs8.dwconv[1].init.qs8 = xnn_init_qs8_conv_minmax_fp32_sse4_params; |
| xnn_params.qs8.dwconv[1].channel_tile = 8; |
| } else if (cpuinfo_has_x86_sse2()) { |
| xnn_params.qs8.dwconv[0].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qs8_dwconv_minmax_fp32_ukernel_up8x9__sse2_mul16_add16; |
| xnn_params.qs8.dwconv[0].init.qs8 = xnn_init_qs8_conv_minmax_fp32_sse2_params; |
| xnn_params.qs8.dwconv[0].channel_tile = 8; |
| xnn_params.qs8.dwconv[1].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qs8_dwconv_minmax_fp32_ukernel_up8x25__sse2_mul16_add16; |
| xnn_params.qs8.dwconv[1].init.qs8 = xnn_init_qs8_conv_minmax_fp32_sse2_params; |
| xnn_params.qs8.dwconv[1].channel_tile = 8; |
| } |
| xnn_params.qs8.dwconv[0].primary_tile = 9; |
| xnn_params.qs8.dwconv[1].primary_tile = 25; |
| |
| if (cpuinfo_has_x86_sse4_1()) { |
| xnn_params.qs8.gavgpool = (struct gavgpool_parameters) { |
| .up = (xnn_gavgpool_unipass_ukernel_function) xnn_qs8_gavgpool_minmax_ukernel_7x__sse41_c8_acc2, |
| .mp = (xnn_gavgpool_multipass_ukernel_function) xnn_qs8_gavgpool_minmax_ukernel_7p7x__sse41_c8_acc2, |
| .mr = 7, |
| }; |
| } else if (cpuinfo_has_x86_ssse3()) { |
| xnn_params.qs8.gavgpool = (struct gavgpool_parameters) { |
| .up = (xnn_gavgpool_unipass_ukernel_function) xnn_qs8_gavgpool_minmax_ukernel_7x__ssse3_c8_acc2, |
| .mp = (xnn_gavgpool_multipass_ukernel_function) xnn_qs8_gavgpool_minmax_ukernel_7p7x__ssse3_c8_acc2, |
| .mr = 7, |
| }; |
| } else if (cpuinfo_has_x86_sse2()) { |
| xnn_params.qs8.gavgpool = (struct gavgpool_parameters) { |
| .up = (xnn_gavgpool_unipass_ukernel_function) xnn_qs8_gavgpool_minmax_ukernel_7x__sse2_c8_acc2, |
| .mp = (xnn_gavgpool_multipass_ukernel_function) xnn_qs8_gavgpool_minmax_ukernel_7p7x__sse2_c8_acc2, |
| .mr = 7, |
| }; |
| } |
| |
| if (cpuinfo_has_x86_avx512f() && cpuinfo_has_x86_avx512bw() && cpuinfo_has_x86_avx512dq() && cpuinfo_has_x86_avx512vl()) { |
| xnn_params.qs8.vadd = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vadd_minmax_ukernel__avx512skx_mul32_ld128_x16, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vaddc_minmax_ukernel__avx512skx_mul32_ld128_x16, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vaddc_minmax_ukernel__avx512skx_mul32_ld128_x16, |
| .init.qs8_addsub = xnn_init_qs8_add_minmax_avx512_params, |
| .element_tile = 16, |
| }; |
| } else if (cpuinfo_has_x86_xop()) { |
| xnn_params.qs8.vadd = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vadd_minmax_ukernel__xop_mul32_ld32_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vaddc_minmax_ukernel__xop_mul32_ld32_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vaddc_minmax_ukernel__xop_mul32_ld32_x8, |
| .init.qs8_addsub = xnn_init_qs8_add_minmax_sse4_mul32_params, |
| .element_tile = 8, |
| }; |
| } else if (cpuinfo_has_x86_avx2()) { |
| xnn_params.qs8.vadd = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vadd_minmax_ukernel__avx2_mul32_ld64_x16, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vaddc_minmax_ukernel__avx2_mul32_ld64_x16, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vaddc_minmax_ukernel__avx2_mul32_ld64_x16, |
| .init.qs8_addsub = xnn_init_qs8_add_minmax_avx2_params, |
| .element_tile = 16, |
| }; |
| } else if (cpuinfo_has_x86_avx()) { |
| xnn_params.qs8.vadd = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vadd_minmax_ukernel__avx_mul32_ld32_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vaddc_minmax_ukernel__avx_mul32_ld32_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vaddc_minmax_ukernel__avx_mul32_ld32_x8, |
| .init.qs8_addsub = xnn_init_qs8_add_minmax_sse4_mul32_params, |
| .element_tile = 8, |
| }; |
| } else if (cpuinfo_has_x86_sse4_1()) { |
| xnn_params.qs8.vadd = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vadd_minmax_ukernel__sse41_mul16_ld64_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vaddc_minmax_ukernel__sse41_mul16_ld64_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vaddc_minmax_ukernel__sse41_mul16_ld64_x8, |
| .init.qs8_addsub = xnn_init_qs8_add_minmax_sse4_mul16_params, |
| .element_tile = 8, |
| }; |
| } else { |
| xnn_params.qs8.vadd = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vadd_minmax_ukernel__sse2_mul16_ld64_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vaddc_minmax_ukernel__sse2_mul16_ld64_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vaddc_minmax_ukernel__sse2_mul16_ld64_x8, |
| .init.qs8_addsub = xnn_init_qs8_add_minmax_sse2_params, |
| .element_tile = 8, |
| }; |
| } |
| if (cpuinfo_has_x86_avx()) { |
| xnn_params.qs8.vmul = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vmul_minmax_fp32_ukernel__avx_mul16_ld64_x16, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vmulc_minmax_fp32_ukernel__avx_mul16_ld64_x16, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vmulc_minmax_fp32_ukernel__avx_mul16_ld64_x16, |
| .init.qs8_mul = xnn_init_qs8_mul_minmax_fp32_sse4_params, |
| .element_tile = 16, |
| }; |
| } else if (cpuinfo_has_x86_sse4_1()) { |
| xnn_params.qs8.vmul = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vmul_minmax_fp32_ukernel__sse41_mul16_ld64_x16, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vmulc_minmax_fp32_ukernel__sse41_mul16_ld64_x16, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vmulc_minmax_fp32_ukernel__sse41_mul16_ld64_x16, |
| .init.qs8_mul = xnn_init_qs8_mul_minmax_fp32_sse4_params, |
| .element_tile = 16, |
| }; |
| } else { |
| xnn_params.qs8.vmul = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vmul_minmax_fp32_ukernel__sse2_mul16_ld64_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vmulc_minmax_fp32_ukernel__sse2_mul16_ld64_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vmulc_minmax_fp32_ukernel__sse2_mul16_ld64_x8, |
| .init.qs8_mul = xnn_init_qs8_mul_minmax_fp32_sse2_params, |
| .element_tile = 8, |
| }; |
| } |
| #endif // XNN_NO_QS8_OPERATORS |
| |
| /**************************** QU8 micro-kernels ****************************/ |
| #ifndef XNN_NO_QU8_OPERATORS |
| init_flags |= XNN_INIT_FLAG_QU8; |
| |
| if (cpuinfo_has_x86_avx512f() && cpuinfo_has_x86_avx512bw() && cpuinfo_has_x86_avx512dq() && cpuinfo_has_x86_avx512vl()) { |
| xnn_params.qu8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qu8_gemm_minmax_fp32_ukernel_4x16c8__avx512skx); |
| xnn_params.qu8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qu8_igemm_minmax_fp32_ukernel_4x16c8__avx512skx); |
| xnn_params.qu8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qu8_gemm_minmax_fp32_ukernel_1x16c8__avx512skx); |
| xnn_params.qu8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qu8_igemm_minmax_fp32_ukernel_1x16c8__avx512skx); |
| xnn_params.qu8.gemm.init.qu8 = xnn_init_qu8_conv_minmax_fp32_avx512_params; |
| xnn_params.qu8.gemm.mr = 4; |
| xnn_params.qu8.gemm.nr = 16; |
| xnn_params.qu8.gemm.log2_kr = 3; |
| } else if (cpuinfo_has_x86_xop()) { |
| // XOP should be checked before AVX2: AMD Excavator supports both, but performs better with XOP microkernels |
| xnn_params.qu8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qu8_gemm_minmax_fp32_ukernel_2x4c8__xop_ld64); |
| xnn_params.qu8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qu8_igemm_minmax_fp32_ukernel_2x4c8__xop_ld64); |
| xnn_params.qu8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qu8_gemm_minmax_fp32_ukernel_1x4c8__xop_ld64); |
| xnn_params.qu8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qu8_igemm_minmax_fp32_ukernel_1x4c8__xop_ld64); |
| xnn_params.qu8.gemm.init.qu8 = xnn_init_qu8_conv_minmax_fp32_sse2_params; |
| xnn_params.qu8.gemm.mr = 2; |
| xnn_params.qu8.gemm.nr = 4; |
| xnn_params.qu8.gemm.log2_kr = 3; |
| } else if (cpuinfo_has_x86_avx2()) { |
| xnn_params.qu8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qu8_gemm_minmax_fp32_ukernel_3x8c8__avx2); |
| xnn_params.qu8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qu8_igemm_minmax_fp32_ukernel_3x8c8__avx2); |
| xnn_params.qu8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qu8_gemm_minmax_fp32_ukernel_1x8c8__avx2); |
| xnn_params.qu8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qu8_igemm_minmax_fp32_ukernel_1x8c8__avx2); |
| xnn_params.qu8.gemm.init.qu8 = xnn_init_qu8_conv_minmax_fp32_avx2_params; |
| xnn_params.qu8.gemm.mr = 3; |
| xnn_params.qu8.gemm.nr = 8; |
| xnn_params.qu8.gemm.log2_kr = 3; |
| } else if (cpuinfo_has_x86_avx()) { |
| xnn_params.qu8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qu8_gemm_minmax_fp32_ukernel_2x4c8__avx_ld128); |
| xnn_params.qu8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qu8_igemm_minmax_fp32_ukernel_2x4c8__avx_ld128); |
| xnn_params.qu8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qu8_gemm_minmax_fp32_ukernel_1x4c8__avx_ld128); |
| xnn_params.qu8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qu8_igemm_minmax_fp32_ukernel_1x4c8__avx_ld128); |
| xnn_params.qu8.gemm.init.qu8 = xnn_init_qu8_conv_minmax_fp32_sse2_params; |
| xnn_params.qu8.gemm.mr = 2; |
| xnn_params.qu8.gemm.nr = 4; |
| xnn_params.qu8.gemm.log2_kr = 3; |
| } else if (cpuinfo_has_x86_sse4_1()) { |
| xnn_params.qu8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qu8_gemm_minmax_fp32_ukernel_3x4c8__sse41_ld64); |
| xnn_params.qu8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qu8_igemm_minmax_fp32_ukernel_3x4c8__sse41_ld64); |
| xnn_params.qu8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qu8_gemm_minmax_fp32_ukernel_1x4c8__sse41_ld64); |
| xnn_params.qu8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qu8_igemm_minmax_fp32_ukernel_1x4c8__sse41_ld64); |
| xnn_params.qu8.gemm.init.qu8 = xnn_init_qu8_conv_minmax_fp32_sse2_params; |
| xnn_params.qu8.gemm.mr = 3; |
| xnn_params.qu8.gemm.nr = 4; |
| xnn_params.qu8.gemm.log2_kr = 3; |
| } else { |
| xnn_params.qu8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qu8_gemm_minmax_fp32_ukernel_3x4c8__sse2_ld64); |
| xnn_params.qu8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qu8_igemm_minmax_fp32_ukernel_3x4c8__sse2_ld64); |
| xnn_params.qu8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qu8_gemm_minmax_fp32_ukernel_1x4c8__sse2_ld64); |
| xnn_params.qu8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qu8_igemm_minmax_fp32_ukernel_1x4c8__sse2_ld64); |
| xnn_params.qu8.gemm.init.qu8 = xnn_init_qu8_conv_minmax_fp32_sse2_params; |
| xnn_params.qu8.gemm.mr = 3; |
| xnn_params.qu8.gemm.nr = 4; |
| xnn_params.qu8.gemm.log2_kr = 3; |
| } |
| |
| if (cpuinfo_has_x86_avx512f() && cpuinfo_has_x86_avx512bw() && cpuinfo_has_x86_avx512dq() && cpuinfo_has_x86_avx512vl()) { |
| xnn_params.qu8.dwconv[0].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qu8_dwconv_minmax_fp32_ukernel_up32x9__avx512skx_mul32; |
| xnn_params.qu8.dwconv[0].init.qu8 = xnn_init_qu8_conv_minmax_fp32_avx512_params; |
| xnn_params.qu8.dwconv[0].channel_tile = 32; |
| xnn_params.qu8.dwconv[1].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qu8_dwconv_minmax_fp32_ukernel_up32x25__avx512skx_mul32; |
| xnn_params.qu8.dwconv[1].init.qu8 = xnn_init_qu8_conv_minmax_fp32_avx512_params; |
| xnn_params.qu8.dwconv[1].channel_tile = 32; |
| } else if (cpuinfo_has_x86_xop()) { |
| // XOP should be checked before AVX2: AMD Excavator supports both, but performs better with XOP microkernels |
| xnn_params.qu8.dwconv[0].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qu8_dwconv_minmax_fp32_ukernel_up16x9__xop_mul32; |
| xnn_params.qu8.dwconv[0].init.qu8 = xnn_init_qu8_conv_minmax_fp32_sse2_params; |
| xnn_params.qu8.dwconv[0].channel_tile = 16; |
| xnn_params.qu8.dwconv[1].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qu8_dwconv_minmax_fp32_ukernel_up16x25__xop_mul32; |
| xnn_params.qu8.dwconv[1].init.qu8 = xnn_init_qu8_conv_minmax_fp32_sse2_params; |
| xnn_params.qu8.dwconv[1].channel_tile = 16; |
| } else if (cpuinfo_has_x86_avx2()) { |
| xnn_params.qu8.dwconv[0].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qu8_dwconv_minmax_fp32_ukernel_up16x9__avx2_mul32; |
| xnn_params.qu8.dwconv[0].init.qu8 = xnn_init_qu8_conv_minmax_fp32_avx2_params; |
| xnn_params.qu8.dwconv[0].channel_tile = 16; |
| xnn_params.qu8.dwconv[1].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qu8_dwconv_minmax_fp32_ukernel_up16x25__avx2_mul32; |
| xnn_params.qu8.dwconv[1].init.qu8 = xnn_init_qu8_conv_minmax_fp32_avx2_params; |
| xnn_params.qu8.dwconv[1].channel_tile = 16; |
| } else if (cpuinfo_has_x86_avx()) { |
| xnn_params.qu8.dwconv[0].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qu8_dwconv_minmax_fp32_ukernel_up16x9__avx_mul16; |
| xnn_params.qu8.dwconv[0].init.qu8 = xnn_init_qu8_conv_minmax_fp32_sse2_params; |
| xnn_params.qu8.dwconv[0].channel_tile = 16; |
| xnn_params.qu8.dwconv[1].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qu8_dwconv_minmax_fp32_ukernel_up16x25__avx_mul16; |
| xnn_params.qu8.dwconv[1].init.qu8 = xnn_init_qu8_conv_minmax_fp32_sse2_params; |
| xnn_params.qu8.dwconv[1].channel_tile = 16; |
| } else if (cpuinfo_has_x86_sse4_1()) { |
| xnn_params.qu8.dwconv[0].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qu8_dwconv_minmax_fp32_ukernel_up8x9__sse41_mul16; |
| xnn_params.qu8.dwconv[0].init.qu8 = xnn_init_qu8_conv_minmax_fp32_sse2_params; |
| xnn_params.qu8.dwconv[0].channel_tile = 8; |
| xnn_params.qu8.dwconv[1].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qu8_dwconv_minmax_fp32_ukernel_up8x25__sse41_mul16; |
| xnn_params.qu8.dwconv[1].init.qu8 = xnn_init_qu8_conv_minmax_fp32_sse2_params; |
| xnn_params.qu8.dwconv[1].channel_tile = 8; |
| } else if (cpuinfo_has_x86_sse2()) { |
| xnn_params.qu8.dwconv[0].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qu8_dwconv_minmax_fp32_ukernel_up8x9__sse2_mul16; |
| xnn_params.qu8.dwconv[0].init.qu8 = xnn_init_qu8_conv_minmax_fp32_sse2_params; |
| xnn_params.qu8.dwconv[0].channel_tile = 8; |
| xnn_params.qu8.dwconv[1].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qu8_dwconv_minmax_fp32_ukernel_up8x25__sse2_mul16; |
| xnn_params.qu8.dwconv[1].init.qu8 = xnn_init_qu8_conv_minmax_fp32_sse2_params; |
| xnn_params.qu8.dwconv[1].channel_tile = 8; |
| } |
| xnn_params.qu8.dwconv[0].primary_tile = 9; |
| xnn_params.qu8.dwconv[1].primary_tile = 25; |
| |
| xnn_params.qu8.avgpool = (struct avgpool_parameters) { |
| .up = (xnn_avgpool_unipass_ukernel_function) xnn_qu8_avgpool_minmax_ukernel_9x__sse2_c8, |
| .mp = (xnn_avgpool_multipass_ukernel_function) xnn_qu8_avgpool_minmax_ukernel_9p8x__sse2_c8, |
| .mr = 9, |
| .qr = 8, |
| }; |
| xnn_params.qu8.gavgpool = (struct gavgpool_parameters) { |
| .up = (xnn_gavgpool_unipass_ukernel_function) xnn_qu8_gavgpool_minmax_ukernel_7x__sse2_c8, |
| .mp = (xnn_gavgpool_multipass_ukernel_function) xnn_qu8_gavgpool_minmax_ukernel_7p7x__sse2_c8, |
| .mr = 7, |
| }; |
| |
| if (cpuinfo_has_x86_avx512f() && cpuinfo_has_x86_avx512bw() && cpuinfo_has_x86_avx512dq() && cpuinfo_has_x86_avx512vl()) { |
| xnn_params.qu8.vadd = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vadd_minmax_ukernel__avx512skx_mul32_ld128_x16, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vaddc_minmax_ukernel__avx512skx_mul32_ld128_x16, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vaddc_minmax_ukernel__avx512skx_mul32_ld128_x16, |
| .init.qu8_addsub = xnn_init_qu8_add_minmax_avx512_params, |
| .element_tile = 16, |
| }; |
| } else if (cpuinfo_has_x86_xop()) { |
| xnn_params.qu8.vadd = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vadd_minmax_ukernel__xop_mul32_ld32_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vaddc_minmax_ukernel__xop_mul32_ld32_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vaddc_minmax_ukernel__xop_mul32_ld32_x8, |
| .init.qu8_addsub = xnn_init_qu8_add_minmax_sse4_params, |
| .element_tile = 8, |
| }; |
| } else if (cpuinfo_has_x86_avx2()) { |
| xnn_params.qu8.vadd = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vadd_minmax_ukernel__avx2_mul32_ld64_x16, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vaddc_minmax_ukernel__avx2_mul32_ld64_x16, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vaddc_minmax_ukernel__avx2_mul32_ld64_x16, |
| .init.qu8_addsub = xnn_init_qu8_add_minmax_avx2_params, |
| .element_tile = 16, |
| }; |
| } else if (cpuinfo_has_x86_avx()) { |
| xnn_params.qu8.vadd = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vadd_minmax_ukernel__avx_mul32_ld32_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vaddc_minmax_ukernel__avx_mul32_ld32_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vaddc_minmax_ukernel__avx_mul32_ld32_x8, |
| .init.qu8_addsub = xnn_init_qu8_add_minmax_sse4_params, |
| .element_tile = 8, |
| }; |
| } else if (cpuinfo_has_x86_sse4_1()) { |
| xnn_params.qu8.vadd = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vadd_minmax_ukernel__sse41_mul16_ld64_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vaddc_minmax_ukernel__sse41_mul16_ld64_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vaddc_minmax_ukernel__sse41_mul16_ld64_x8, |
| .init.qu8_addsub = xnn_init_qu8_add_minmax_sse2_params, |
| .element_tile = 8, |
| }; |
| } else { |
| xnn_params.qu8.vadd = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vadd_minmax_ukernel__sse2_mul16_ld64_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vaddc_minmax_ukernel__sse2_mul16_ld64_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vaddc_minmax_ukernel__sse2_mul16_ld64_x8, |
| .init.qu8_addsub = xnn_init_qu8_add_minmax_sse2_params, |
| .element_tile = 8, |
| }; |
| } |
| if (cpuinfo_has_x86_avx()) { |
| xnn_params.qu8.vmul = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vmul_minmax_fp32_ukernel__avx_mul16_ld64_x16, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vmulc_minmax_fp32_ukernel__avx_mul16_ld64_x16, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vmulc_minmax_fp32_ukernel__avx_mul16_ld64_x16, |
| .init.qu8_mul = xnn_init_qu8_mul_minmax_fp32_sse2_params, |
| .element_tile = 16, |
| }; |
| } else if (cpuinfo_has_x86_sse4_1()) { |
| xnn_params.qu8.vmul = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vmul_minmax_fp32_ukernel__sse41_mul16_ld64_x16, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vmulc_minmax_fp32_ukernel__sse41_mul16_ld64_x16, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vmulc_minmax_fp32_ukernel__sse41_mul16_ld64_x16, |
| .init.qu8_mul = xnn_init_qu8_mul_minmax_fp32_sse2_params, |
| .element_tile = 16, |
| }; |
| } else { |
| xnn_params.qu8.vmul = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vmul_minmax_fp32_ukernel__sse2_mul16_ld64_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vmulc_minmax_fp32_ukernel__sse2_mul16_ld64_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vmulc_minmax_fp32_ukernel__sse2_mul16_ld64_x8, |
| .init.qu8_mul = xnn_init_qu8_mul_minmax_fp32_sse2_params, |
| .element_tile = 8, |
| }; |
| } |
| #endif // XNN_NO_QU8_OPERATORS |
| |
| /**************************** U8 micro-kernels ****************************/ |
| #ifndef XNN_NO_S8_OPERATORS |
| init_flags |= XNN_INIT_FLAG_S8; |
| |
| if (cpuinfo_has_x86_sse4_1()) { |
| xnn_params.s8.clamp = (struct vunary_parameters) { |
| .ukernel = (xnn_univector_ukernel_function) xnn_s8_vclamp_ukernel__sse41_x64, |
| .init.s8_minmax = xnn_init_s8_minmax_sse4_params, |
| .element_tile = 64, |
| }; |
| xnn_params.s8.maxpool = (struct maxpool_parameters) { |
| .ukernel = (xnn_maxpool_ukernel_function) xnn_s8_maxpool_minmax_ukernel_9p8x__sse41_c16, |
| .init.s8 = xnn_init_s8_minmax_sse4_params, |
| .mr = 9, |
| .qr = 8, |
| }; |
| } else { |
| xnn_params.s8.clamp = (struct vunary_parameters) { |
| .ukernel = (xnn_univector_ukernel_function) xnn_s8_vclamp_ukernel__sse2_x64, |
| .init.s8_minmax = xnn_init_s8_minmax_sse2_params, |
| .element_tile = 64, |
| }; |
| xnn_params.s8.maxpool = (struct maxpool_parameters) { |
| .ukernel = (xnn_maxpool_ukernel_function) xnn_s8_maxpool_minmax_ukernel_9p8x__sse2_c16, |
| .init.s8 = xnn_init_s8_minmax_sse2_params, |
| .mr = 9, |
| .qr = 8, |
| }; |
| } |
| #endif // XNN_NO_S8_OPERATORS |
| |
| /**************************** U8 micro-kernels ****************************/ |
| #ifndef XNN_NO_U8_OPERATORS |
| init_flags |= XNN_INIT_FLAG_U8; |
| |
| xnn_params.u8.clamp = (struct vunary_parameters) { |
| .ukernel = (xnn_univector_ukernel_function) xnn_u8_vclamp_ukernel__sse2_x64, |
| .init.u8_minmax = xnn_init_u8_minmax_sse2_params, |
| .element_tile = 64, |
| }; |
| xnn_params.u8.maxpool = (struct maxpool_parameters) { |
| .ukernel = (xnn_maxpool_ukernel_function) xnn_u8_maxpool_minmax_ukernel_9p8x__sse2_c16, |
| .init.u8 = xnn_init_u8_minmax_sse2_params, |
| .mr = 9, |
| .qr = 8, |
| }; |
| xnn_params.u8.lut32norm = xnn_u8_lut32norm_ukernel__scalar; |
| xnn_params.u8.rmax = xnn_u8_rmax_ukernel__sse2; |
| #endif // XNN_NO_U8_OPERATORS |
| |
| /**************************** X8 micro-kernels ****************************/ |
| #ifndef XNN_NO_X8_OPERATORS |
| init_flags |= XNN_INIT_FLAG_X8; |
| |
| if (cpuinfo_has_x86_avx512f() && cpuinfo_has_x86_avx512bw() && cpuinfo_has_x86_avx512dq() && cpuinfo_has_x86_avx512vl()) { |
| xnn_params.x8.lut = xnn_x8_lut_ukernel__avx512skx_vpshufb_x64; |
| } else if (cpuinfo_has_x86_avx2()) { |
| xnn_params.x8.lut = xnn_x8_lut_ukernel__avx2_x128; |
| } else if (cpuinfo_has_x86_avx()) { |
| xnn_params.x8.lut = xnn_x8_lut_ukernel__avx_x64; |
| } else { |
| // Note: SSSE3 version is usually slower than scalar |
| xnn_params.x8.lut = xnn_x8_lut_ukernel__scalar_x4; |
| } |
| xnn_params.x8.zip = (struct zip_parameters) { |
| .x2 = (xnn_zipc_ukernel_function) xnn_x8_zip_x2_ukernel__sse2, |
| .x3 = (xnn_zipc_ukernel_function) xnn_x8_zip_x3_ukernel__sse2, |
| .x4 = (xnn_zipc_ukernel_function) xnn_x8_zip_x4_ukernel__sse2, |
| .xm = (xnn_zipv_ukernel_function) xnn_x8_zip_xm_ukernel__sse2, |
| }; |
| #endif // XNN_NO_X8_OPERATORS |
| |
| /**************************** F32 micro-kernels ****************************/ |
| #ifndef XNN_NO_F32_OPERATORS |
| init_flags |= XNN_INIT_FLAG_F32; |
| |
| if (!XNN_PLATFORM_MOBILE && cpuinfo_has_x86_avx512f()) { |
| xnn_params.f32.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_7x16__avx512f_broadcast); |
| xnn_params.f32.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_7x16__avx512f_broadcast); |
| xnn_params.f32.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_1x16__avx512f_broadcast); |
| xnn_params.f32.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_1x16__avx512f_broadcast); |
| xnn_params.f32.gemm.init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.gemm.mr = 7; |
| xnn_params.f32.gemm.nr = 16; |
| } else if (!XNN_PLATFORM_MOBILE && cpuinfo_has_x86_fma3()) { |
| switch (cpuinfo_get_core(0)->uarch) { |
| case cpuinfo_uarch_zen: |
| case cpuinfo_uarch_dhyana: |
| xnn_params.f32.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_4x16s4__fma3_broadcast); |
| xnn_params.f32.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_4x16s4__fma3_broadcast); |
| xnn_params.f32.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_1x16s4__fma3_broadcast); |
| xnn_params.f32.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_1x16s4__fma3_broadcast); |
| xnn_params.f32.gemm.init.f32 = xnn_init_f32_minmax_avx_params; |
| xnn_params.f32.gemm.mr = 4; |
| xnn_params.f32.gemm.nr = 16; |
| xnn_params.f32.gemm.log2_sr = 2; |
| break; |
| default: |
| xnn_params.f32.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_5x16__fma3_broadcast); |
| xnn_params.f32.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_5x16__fma3_broadcast); |
| xnn_params.f32.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_1x16__fma3_broadcast); |
| xnn_params.f32.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_1x16__fma3_broadcast); |
| xnn_params.f32.gemm.init.f32 = xnn_init_f32_minmax_avx_params; |
| xnn_params.f32.gemm.mr = 5; |
| xnn_params.f32.gemm.nr = 16; |
| break; |
| } |
| } else if (!XNN_PLATFORM_MOBILE && cpuinfo_has_x86_avx()) { |
| xnn_params.f32.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_5x16__avx_broadcast); |
| xnn_params.f32.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_5x16__avx_broadcast); |
| xnn_params.f32.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_1x16__avx_broadcast); |
| xnn_params.f32.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_1x16__avx_broadcast); |
| xnn_params.f32.gemm.init.f32 = xnn_init_f32_minmax_avx_params; |
| xnn_params.f32.gemm.mr = 5; |
| xnn_params.f32.gemm.nr = 16; |
| } else { |
| xnn_params.f32.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_4x8__sse_load1); |
| xnn_params.f32.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_4x8__sse_load1); |
| xnn_params.f32.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_1x8__sse_load1); |
| xnn_params.f32.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_1x8__sse_load1); |
| xnn_params.f32.gemm.init.f32 = xnn_init_f32_minmax_sse_params; |
| xnn_params.f32.gemm.mr = 4; |
| xnn_params.f32.gemm.nr = 8; |
| } |
| xnn_params.f32.gemm2.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_4x2c4__sse); |
| xnn_params.f32.gemm2.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_4x2c4__sse); |
| xnn_params.f32.gemm2.init.f32 = xnn_init_f32_minmax_sse_params; |
| xnn_params.f32.gemm2.mr = 4; |
| xnn_params.f32.gemm2.nr = 2; |
| xnn_params.f32.gemm2.log2_kr = 2; |
| |
| if (!XNN_PLATFORM_MOBILE && cpuinfo_has_x86_avx512f()) { |
| xnn_params.f32.dwconv[0].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_minmax_ukernel_up16x3__avx512f; |
| xnn_params.f32.dwconv[0].init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.dwconv[0].channel_tile = 16; |
| xnn_params.f32.dwconv[0].primary_tile = 3; |
| |
| xnn_params.f32.dwconv[1].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_minmax_ukernel_up16x4__avx512f; |
| xnn_params.f32.dwconv[1].init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.dwconv[1].channel_tile = 16; |
| xnn_params.f32.dwconv[1].primary_tile = 4; |
| |
| xnn_params.f32.dwconv[2].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_minmax_ukernel_up16x9__avx512f; |
| xnn_params.f32.dwconv[2].init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.dwconv[2].channel_tile = 16; |
| xnn_params.f32.dwconv[2].primary_tile = 9; |
| |
| xnn_params.f32.dwconv[3].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_minmax_ukernel_up16x25__avx512f; |
| xnn_params.f32.dwconv[3].init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.dwconv[3].channel_tile = 16; |
| xnn_params.f32.dwconv[3].primary_tile = 25; |
| } else if (!XNN_PLATFORM_MOBILE && cpuinfo_has_x86_fma3()) { |
| xnn_params.f32.dwconv[0].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_minmax_ukernel_up16x3__fma3; |
| xnn_params.f32.dwconv[0].init.f32 = xnn_init_f32_minmax_avx_params; |
| xnn_params.f32.dwconv[0].channel_tile = 16; |
| xnn_params.f32.dwconv[0].primary_tile = 3; |
| |
| xnn_params.f32.dwconv[1].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_minmax_ukernel_up16x4__fma3; |
| xnn_params.f32.dwconv[1].init.f32 = xnn_init_f32_minmax_avx_params; |
| xnn_params.f32.dwconv[1].channel_tile = 16; |
| xnn_params.f32.dwconv[1].primary_tile = 4; |
| |
| xnn_params.f32.dwconv[2].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_minmax_ukernel_up16x9__fma3; |
| xnn_params.f32.dwconv[2].init.f32 = xnn_init_f32_minmax_avx_params; |
| xnn_params.f32.dwconv[2].channel_tile = 16; |
| xnn_params.f32.dwconv[2].primary_tile = 9; |
| |
| xnn_params.f32.dwconv[3].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_minmax_ukernel_up8x25__fma3; |
| xnn_params.f32.dwconv[3].init.f32 = xnn_init_f32_minmax_avx_params; |
| xnn_params.f32.dwconv[3].channel_tile = 8; |
| xnn_params.f32.dwconv[3].primary_tile = 25; |
| } else if (!XNN_PLATFORM_MOBILE && cpuinfo_has_x86_avx()) { |
| xnn_params.f32.dwconv[0].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_minmax_ukernel_up16x3__avx; |
| xnn_params.f32.dwconv[0].init.f32 = xnn_init_f32_minmax_avx_params; |
| xnn_params.f32.dwconv[0].channel_tile = 16; |
| xnn_params.f32.dwconv[0].primary_tile = 3; |
| |
| xnn_params.f32.dwconv[1].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_minmax_ukernel_up16x4__avx; |
| xnn_params.f32.dwconv[1].init.f32 = xnn_init_f32_minmax_avx_params; |
| xnn_params.f32.dwconv[1].channel_tile = 16; |
| xnn_params.f32.dwconv[1].primary_tile = 4; |
| |
| xnn_params.f32.dwconv[2].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_minmax_ukernel_up16x9__avx; |
| xnn_params.f32.dwconv[2].init.f32 = xnn_init_f32_minmax_avx_params; |
| xnn_params.f32.dwconv[2].channel_tile = 16; |
| xnn_params.f32.dwconv[2].primary_tile = 9; |
| |
| xnn_params.f32.dwconv[3].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_minmax_ukernel_up8x25__avx; |
| xnn_params.f32.dwconv[3].init.f32 = xnn_init_f32_minmax_avx_params; |
| xnn_params.f32.dwconv[3].channel_tile = 8; |
| xnn_params.f32.dwconv[3].primary_tile = 25; |
| } else { |
| xnn_params.f32.dwconv[0].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_minmax_ukernel_up8x3__sse; |
| xnn_params.f32.dwconv[0].init.f32 = xnn_init_f32_minmax_sse_params; |
| xnn_params.f32.dwconv[0].channel_tile = 8; |
| xnn_params.f32.dwconv[0].primary_tile = 3; |
| |
| xnn_params.f32.dwconv[1].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_minmax_ukernel_up8x4__sse; |
| xnn_params.f32.dwconv[1].init.f32 = xnn_init_f32_minmax_sse_params; |
| xnn_params.f32.dwconv[1].channel_tile = 8; |
| xnn_params.f32.dwconv[1].primary_tile = 4; |
| |
| xnn_params.f32.dwconv[2].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_minmax_ukernel_up8x9__sse; |
| xnn_params.f32.dwconv[2].init.f32 = xnn_init_f32_minmax_sse_params; |
| xnn_params.f32.dwconv[2].channel_tile = 8; |
| xnn_params.f32.dwconv[2].primary_tile = 9; |
| |
| xnn_params.f32.dwconv[3].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_minmax_ukernel_up8x25__sse; |
| xnn_params.f32.dwconv[3].init.f32 = xnn_init_f32_minmax_sse_params; |
| xnn_params.f32.dwconv[3].channel_tile = 8; |
| xnn_params.f32.dwconv[3].primary_tile = 25; |
| } |
| xnn_params.f32.avgpool = (struct avgpool_parameters) { |
| .up = (xnn_avgpool_unipass_ukernel_function) xnn_f32_avgpool_minmax_ukernel_9x__sse_c4, |
| .mp = (xnn_avgpool_multipass_ukernel_function) xnn_f32_avgpool_minmax_ukernel_9p8x__sse_c4, |
| .mr = 9, |
| .qr = 8, |
| }; |
| xnn_params.f32.pavgpool = (struct pavgpool_parameters) { |
| .up = (xnn_pavgpool_unipass_ukernel_function) xnn_f32_pavgpool_minmax_ukernel_9x__sse_c4, |
| .mp = (xnn_pavgpool_multipass_ukernel_function) xnn_f32_pavgpool_minmax_ukernel_9p8x__sse_c4, |
| .mr = 9, |
| .qr = 8, |
| }; |
| xnn_params.f32.gavgpool = (struct gavgpool_parameters) { |
| .up = (xnn_gavgpool_unipass_ukernel_function) xnn_f32_gavgpool_minmax_ukernel_7x__sse_c4, |
| .mp = (xnn_gavgpool_multipass_ukernel_function) xnn_f32_gavgpool_minmax_ukernel_7p7x__sse_c4, |
| .mr = 7, |
| }; |
| xnn_params.f32.maxpool = (struct maxpool_parameters) { |
| .ukernel = (xnn_maxpool_ukernel_function) xnn_f32_maxpool_minmax_ukernel_9p8x__sse_c4, |
| .init.f32 = xnn_init_f32_minmax_sse_params, |
| .mr = 9, |
| .qr = 8, |
| }; |
| xnn_params.f32.argmaxpool[0] = (struct argmaxpool_parameters) { |
| .up = (xnn_argmaxpool_unipass_ukernel_function) xnn_f32_argmaxpool_ukernel_4x__sse2_c4, |
| .mr = 4, |
| }; |
| xnn_params.f32.argmaxpool[1] = (struct argmaxpool_parameters) { |
| .up = (xnn_argmaxpool_unipass_ukernel_function) xnn_f32_argmaxpool_ukernel_9x__sse2_c4, |
| .mr = 9, |
| }; |
| xnn_params.f32.argmaxpool[2] = (struct argmaxpool_parameters) { |
| .mp = (xnn_argmaxpool_multipass_ukernel_function) xnn_f32_argmaxpool_ukernel_9p8x__sse2_c4, |
| .mr = 9, |
| .qr = 8, |
| }; |
| xnn_params.f32.ibilinear = (struct ibilinear_parameters) { |
| .ukernel = (xnn_ibilinear_ukernel_function) xnn_f32_ibilinear_ukernel__sse_c8, |
| .pixel_tile = 1, |
| .channel_tile = 8, |
| }; |
| if (!XNN_PLATFORM_MOBILE && cpuinfo_has_x86_avx512f()) { |
| xnn_params.f32.abs = (xnn_univector_ukernel_function) xnn_f32_vabs_ukernel__avx512f_x16; |
| } else if (!XNN_PLATFORM_MOBILE && cpuinfo_has_x86_avx()) { |
| xnn_params.f32.abs = (xnn_univector_ukernel_function) xnn_f32_vabs_ukernel__avx_x16; |
| } else { |
| xnn_params.f32.abs = (xnn_univector_ukernel_function) xnn_f32_vabs_ukernel__sse_x8; |
| } |
| if (!XNN_PLATFORM_MOBILE && cpuinfo_has_x86_avx512f()) { |
| xnn_params.f32.clamp = (struct vunary_parameters) { |
| .ukernel = (xnn_univector_ukernel_function) xnn_f32_vclamp_ukernel__avx512f_x16, |
| .init.f32_minmax = xnn_init_f32_minmax_scalar_params, |
| .element_tile = 16, |
| }; |
| } else if (!XNN_PLATFORM_MOBILE && cpuinfo_has_x86_avx()) { |
| xnn_params.f32.clamp = (struct vunary_parameters) { |
| .ukernel = (xnn_univector_ukernel_function) xnn_f32_vclamp_ukernel__avx_x16, |
| .init.f32_minmax = xnn_init_f32_minmax_avx_params, |
| .element_tile = 16, |
| }; |
| } else { |
| xnn_params.f32.clamp = (struct vunary_parameters) { |
| .ukernel = (xnn_univector_ukernel_function) xnn_f32_vclamp_ukernel__sse_x8, |
| .init.f32_minmax = xnn_init_f32_minmax_sse_params, |
| .element_tile = 8, |
| }; |
| } |
| if (!XNN_PLATFORM_MOBILE && cpuinfo_has_x86_avx512f()) { |
| xnn_params.f32.elu = (xnn_univector_ukernel_function) xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x64; |
| } else if (!XNN_PLATFORM_MOBILE && cpuinfo_has_x86_avx2()) { |
| xnn_params.f32.elu = (xnn_univector_ukernel_function) xnn_f32_velu_ukernel__avx2_rr1_lut4_p4_perm_x56; |
| } else if (!XNN_PLATFORM_MOBILE && cpuinfo_has_x86_avx()) { |
| xnn_params.f32.elu = (xnn_univector_ukernel_function) xnn_f32_velu_ukernel__avx_rr2_lut4_p4_perm_x32; |
| } else { |
| xnn_params.f32.elu = (xnn_univector_ukernel_function) xnn_f32_velu_ukernel__sse2_rr2_lut16_p3_x12; |
| } |
| if (!XNN_PLATFORM_MOBILE && cpuinfo_has_x86_avx512f()) { |
| xnn_params.f32.hswish = (xnn_univector_ukernel_function) xnn_f32_vhswish_ukernel__avx512f_x16; |
| } else if (!XNN_PLATFORM_MOBILE && cpuinfo_has_x86_fma3()) { |
| xnn_params.f32.hswish = (xnn_univector_ukernel_function) xnn_f32_vhswish_ukernel__fma3_x16; |
| } else if (!XNN_PLATFORM_MOBILE && cpuinfo_has_x86_avx()) { |
| xnn_params.f32.hswish = (xnn_univector_ukernel_function) xnn_f32_vhswish_ukernel__avx_x16; |
| } else { |
| xnn_params.f32.hswish = (xnn_univector_ukernel_function) xnn_f32_vhswish_ukernel__sse_x8; |
| } |
| if (!XNN_PLATFORM_MOBILE && cpuinfo_has_x86_avx512f()) { |
| xnn_params.f32.lrelu = (xnn_univector_ukernel_function) xnn_f32_vlrelu_ukernel__avx512f_x16; |
| } else if (!XNN_PLATFORM_MOBILE && cpuinfo_has_x86_avx()) { |
| xnn_params.f32.lrelu = (xnn_univector_ukernel_function) xnn_f32_vlrelu_ukernel__avx_x16; |
| } else if (cpuinfo_has_x86_sse4_1()) { |
| xnn_params.f32.lrelu = (xnn_univector_ukernel_function) xnn_f32_vlrelu_ukernel__sse41_x8; |
| } else { |
| xnn_params.f32.lrelu = (xnn_univector_ukernel_function) xnn_f32_vlrelu_ukernel__sse_x8; |
| } |
| if (!XNN_PLATFORM_MOBILE && cpuinfo_has_x86_avx512f()) { |
| xnn_params.f32.neg = (xnn_univector_ukernel_function) xnn_f32_vneg_ukernel__avx512f_x16; |
| } else if (!XNN_PLATFORM_MOBILE && cpuinfo_has_x86_avx()) { |
| xnn_params.f32.neg = (xnn_univector_ukernel_function) xnn_f32_vneg_ukernel__avx_x16; |
| } else { |
| xnn_params.f32.neg = (xnn_univector_ukernel_function) xnn_f32_vneg_ukernel__sse_x8; |
| } |
| if (!XNN_PLATFORM_MOBILE && cpuinfo_has_x86_avx512f()) { |
| xnn_params.f32.rndne = (xnn_univector_ukernel_function) xnn_f32_vrndne_ukernel__avx512f_x16; |
| xnn_params.f32.rndz = (xnn_univector_ukernel_function) xnn_f32_vrndz_ukernel__avx512f_x16; |
| xnn_params.f32.rndu = (xnn_univector_ukernel_function) xnn_f32_vrndu_ukernel__avx512f_x16; |
| xnn_params.f32.rndd = (xnn_univector_ukernel_function) xnn_f32_vrndd_ukernel__avx512f_x16; |
| } else if (!XNN_PLATFORM_MOBILE && cpuinfo_has_x86_avx()) { |
| xnn_params.f32.rndne = (xnn_univector_ukernel_function) xnn_f32_vrndne_ukernel__avx_x16; |
| xnn_params.f32.rndz = (xnn_univector_ukernel_function) xnn_f32_vrndz_ukernel__avx_x16; |
| xnn_params.f32.rndu = (xnn_univector_ukernel_function) xnn_f32_vrndu_ukernel__avx_x16; |
| xnn_params.f32.rndd = (xnn_univector_ukernel_function) xnn_f32_vrndd_ukernel__avx_x16; |
| } else if (cpuinfo_has_x86_sse4_1()) { |
| xnn_params.f32.rndne = (xnn_univector_ukernel_function) xnn_f32_vrndne_ukernel__sse41_x8; |
| xnn_params.f32.rndz = (xnn_univector_ukernel_function) xnn_f32_vrndz_ukernel__sse41_x8; |
| xnn_params.f32.rndu = (xnn_univector_ukernel_function) xnn_f32_vrndu_ukernel__sse41_x8; |
| xnn_params.f32.rndd = (xnn_univector_ukernel_function) xnn_f32_vrndd_ukernel__sse41_x8; |
| } else { |
| xnn_params.f32.rndne = (xnn_univector_ukernel_function) xnn_f32_vrndne_ukernel__sse2_x8; |
| xnn_params.f32.rndz = (xnn_univector_ukernel_function) xnn_f32_vrndz_ukernel__sse2_x8; |
| xnn_params.f32.rndu = (xnn_univector_ukernel_function) xnn_f32_vrndu_ukernel__sse2_x8; |
| xnn_params.f32.rndd = (xnn_univector_ukernel_function) xnn_f32_vrndd_ukernel__sse2_x8; |
| } |
| if (!XNN_PLATFORM_MOBILE && cpuinfo_has_x86_avx512f()) { |
| xnn_params.f32.sigmoid = (xnn_univector_ukernel_function) xnn_f32_vsigmoid_ukernel__avx512f_rr2_lut32_p2_perm2_scalef_div_x64; |
| } else if (!XNN_PLATFORM_MOBILE && cpuinfo_has_x86_avx2()) { |
| xnn_params.f32.sigmoid = (xnn_univector_ukernel_function) xnn_f32_vsigmoid_ukernel__avx2_rr1_p5_div_x40; |
| } else if (!XNN_PLATFORM_MOBILE && cpuinfo_has_x86_avx()) { |
| xnn_params.f32.sigmoid = (xnn_univector_ukernel_function) xnn_f32_vsigmoid_ukernel__avx_rr2_p5_nr2_x40; |
| } else if (cpuinfo_has_x86_sse4_1()) { |
| xnn_params.f32.sigmoid = (xnn_univector_ukernel_function) xnn_f32_vsigmoid_ukernel__sse41_lut64_p2_div_x8; |
| } else { |
| xnn_params.f32.sigmoid = (xnn_univector_ukernel_function) xnn_f32_vsigmoid_ukernel__sse2_lut64_p2_div_x8; |
| } |
| if (!XNN_PLATFORM_MOBILE && cpuinfo_has_x86_avx512f()) { |
| xnn_params.f32.sqr = (xnn_univector_ukernel_function) xnn_f32_vsqr_ukernel__avx512f_x16; |
| } else if (!XNN_PLATFORM_MOBILE && cpuinfo_has_x86_avx()) { |
| xnn_params.f32.sqr = (xnn_univector_ukernel_function) xnn_f32_vsqr_ukernel__avx_x16; |
| } else { |
| xnn_params.f32.sqr = (xnn_univector_ukernel_function) xnn_f32_vsqr_ukernel__sse_x8; |
| } |
| if (!XNN_PLATFORM_MOBILE && cpuinfo_has_x86_avx()) { |
| xnn_params.f32.sqrt = (xnn_univector_ukernel_function) xnn_f32_vsqrt_ukernel__avx_sqrt_x8; |
| } else { |
| xnn_params.f32.sqrt = (xnn_univector_ukernel_function) xnn_f32_vsqrt_ukernel__sse_sqrt_x4; |
| } |
| if (!XNN_PLATFORM_MOBILE && cpuinfo_has_x86_avx512f()) { |
| xnn_params.f32.prelu = (struct prelu_parameters) { |
| .ukernel = (xnn_prelu_ukernel_function) xnn_f32_prelu_ukernel__avx512f_2x16, |
| .row_tile = 2, |
| .channel_tile = 16, |
| }; |
| } else if (!XNN_PLATFORM_MOBILE && cpuinfo_has_x86_avx()) { |
| xnn_params.f32.prelu = (struct prelu_parameters) { |
| .ukernel = (xnn_prelu_ukernel_function) xnn_f32_prelu_ukernel__avx_2x16, |
| .row_tile = 2, |
| .channel_tile = 16, |
| }; |
| } else if (cpuinfo_has_x86_sse4_1()) { |
| xnn_params.f32.prelu = (struct prelu_parameters) { |
| .ukernel = (xnn_prelu_ukernel_function) xnn_f32_prelu_ukernel__sse41_2x8, |
| .row_tile = 2, |
| .channel_tile = 8, |
| }; |
| } else { |
| xnn_params.f32.prelu = (struct prelu_parameters) { |
| .ukernel = (xnn_prelu_ukernel_function) xnn_f32_prelu_ukernel__sse2_2x8, |
| .row_tile = 2, |
| .channel_tile = 8, |
| }; |
| } |
| xnn_params.f32.raddstoreexpminusmax = xnn_f32_raddstoreexpminusmax_ukernel__sse2_p5_x20_acc2; |
| xnn_params.f32.rmax = xnn_f32_rmax_ukernel__sse; |
| if (!XNN_PLATFORM_MOBILE && cpuinfo_has_x86_avx512f()) { |
| xnn_params.f32.vadd = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vadd_minmax_ukernel__avx512f_x32, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vaddc_minmax_ukernel__avx512f_x32, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vaddc_minmax_ukernel__avx512f_x32, |
| .element_tile = 32, |
| }; |
| xnn_params.f32.vdiv = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vdiv_minmax_ukernel__avx512f_x32, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vdivc_minmax_ukernel__avx512f_x32, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vrdivc_minmax_ukernel__avx512f_x32, |
| .element_tile = 32, |
| }; |
| xnn_params.f32.vmax = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmax_ukernel__avx512f_x32, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmaxc_ukernel__avx512f_x32, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmaxc_ukernel__avx512f_x32, |
| .element_tile = 32, |
| }; |
| xnn_params.f32.vmin = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmin_ukernel__avx512f_x32, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vminc_ukernel__avx512f_x32, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vminc_ukernel__avx512f_x32, |
| .element_tile = 32, |
| }; |
| xnn_params.f32.vmul = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmul_minmax_ukernel__avx512f_x32, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmulc_minmax_ukernel__avx512f_x32, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmulc_minmax_ukernel__avx512f_x32, |
| .element_tile = 32, |
| }; |
| xnn_params.f32.vsub = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsub_minmax_ukernel__avx512f_x32, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsubc_minmax_ukernel__avx512f_x32, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vrsubc_minmax_ukernel__avx512f_x32, |
| .element_tile = 32, |
| }; |
| xnn_params.f32.vsqrdiff = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsqrdiff_ukernel__avx512f_x32, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsqrdiffc_ukernel__avx512f_x32, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsqrdiffc_ukernel__avx512f_x32, |
| .element_tile = 32, |
| }; |
| } else if (!XNN_PLATFORM_MOBILE && cpuinfo_has_x86_avx()) { |
| xnn_params.f32.vadd = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vadd_minmax_ukernel__avx_x16, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vaddc_minmax_ukernel__avx_x16, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vaddc_minmax_ukernel__avx_x16, |
| .element_tile = 16, |
| }; |
| xnn_params.f32.vdiv = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vdiv_minmax_ukernel__avx_x16, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vdivc_minmax_ukernel__avx_x16, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vrdivc_minmax_ukernel__avx_x16, |
| .element_tile = 16, |
| }; |
| xnn_params.f32.vmax = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmax_ukernel__avx_x16, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmaxc_ukernel__avx_x16, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmaxc_ukernel__avx_x16, |
| .element_tile = 16, |
| }; |
| xnn_params.f32.vmin = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmin_ukernel__avx_x16, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vminc_ukernel__avx_x16, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vminc_ukernel__avx_x16, |
| .element_tile = 16, |
| }; |
| xnn_params.f32.vmul = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmul_minmax_ukernel__avx_x16, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmulc_minmax_ukernel__avx_x16, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmulc_minmax_ukernel__avx_x16, |
| .element_tile = 16, |
| }; |
| xnn_params.f32.vsub = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsub_minmax_ukernel__avx_x16, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsubc_minmax_ukernel__avx_x16, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vrsubc_minmax_ukernel__avx_x16, |
| .element_tile = 16, |
| }; |
| xnn_params.f32.vsqrdiff = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsqrdiff_ukernel__avx_x16, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsqrdiffc_ukernel__avx_x16, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsqrdiffc_ukernel__avx_x16, |
| .element_tile = 16, |
| }; |
| } else { |
| xnn_params.f32.vadd = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vadd_minmax_ukernel__sse_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vaddc_minmax_ukernel__sse_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vaddc_minmax_ukernel__sse_x8, |
| .element_tile = 8, |
| }; |
| xnn_params.f32.vdiv = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vdiv_minmax_ukernel__sse_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vdivc_minmax_ukernel__sse_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vrdivc_minmax_ukernel__sse_x8, |
| .element_tile = 8, |
| }; |
| xnn_params.f32.vmax = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmax_ukernel__sse_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmaxc_ukernel__sse_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmaxc_ukernel__sse_x8, |
| .element_tile = 8, |
| }; |
| xnn_params.f32.vmin = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmin_ukernel__sse_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vminc_ukernel__sse_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vminc_ukernel__sse_x8, |
| .element_tile = 8, |
| }; |
| xnn_params.f32.vmul = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmul_minmax_ukernel__sse_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmulc_minmax_ukernel__sse_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmulc_minmax_ukernel__sse_x8, |
| .element_tile = 8, |
| }; |
| xnn_params.f32.vsub = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsub_minmax_ukernel__sse_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsubc_minmax_ukernel__sse_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vrsubc_minmax_ukernel__sse_x8, |
| .element_tile = 8, |
| }; |
| xnn_params.f32.vsqrdiff = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsqrdiff_ukernel__sse_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsqrdiffc_ukernel__sse_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsqrdiffc_ukernel__sse_x8, |
| .element_tile = 8, |
| }; |
| } |
| xnn_params.f32.vmulcaddc = (struct vmulcaddc_parameters) { |
| .ukernel = (xnn_vmulcaddc_ukernel_function) xnn_f32_vmulcaddc_minmax_ukernel_c4__sse_2x, |
| .init.f32 = xnn_init_f32_minmax_sse_params, |
| .channel_tile = 4, |
| .row_tile = 2, |
| }; |
| #ifndef XNN_NO_NCHW_OPERATORS |
| // Sparse microkernels on x86 currently target only SSE, and on processors |
| // with AVX ISA dense inference is expected to be faster than sparse. |
| if (!cpuinfo_has_x86_avx()) { |
| init_flags |= XNN_INIT_FLAG_CHW_OPT; |
| } |
| |
| xnn_params.f32.spmm = (struct spmm_parameters) { |
| .ukernel = (xnn_spmm_ukernel_function) xnn_f32_spmm_minmax_ukernel_32x1__sse, |
| .mr = 32, |
| .nr = 1, |
| }; |
| xnn_params.f32.conv_hwc2chw_3x3c3s2 = (struct conv_hwc2chw_parameters) { |
| .ukernel_with_symm_padding = |
| (xnn_conv_hwc2chw_ukernel_function) xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__sse_2x2, |
| .output_channel_tile = 4, |
| .output_height_tile = 2, |
| .output_width_tile = 2, |
| }; |
| if (!XNN_PLATFORM_MOBILE && cpuinfo_has_x86_ssse3()) { |
| xnn_params.f32.dwconv2d_chw_3x3 = (struct dwconv2d_chw_parameters) { |
| .ukernel = (xnn_dwconv2d_chw_ukernel_function) xnn_f32_dwconv2d_chw_ukernel_3x3p1__ssse3_2x4_acc2, |
| .output_width_tile = 4, |
| .output_height_tile = 2, |
| }; |
| } else { |
| xnn_params.f32.dwconv2d_chw_3x3 = (struct dwconv2d_chw_parameters) { |
| .ukernel = (xnn_dwconv2d_chw_ukernel_function) xnn_f32_dwconv2d_chw_ukernel_3x3p1__sse_2x4_acc2, |
| .output_width_tile = 4, |
| .output_height_tile = 2, |
| }; |
| } |
| xnn_params.f32.dwconv2d_chw_3x3s2 = (struct dwconv2d_chw_parameters) { |
| .ukernel = (xnn_dwconv2d_chw_ukernel_function) xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__sse_1x4_acc3, |
| .output_width_tile = 4, |
| .output_height_tile = 1, |
| }; |
| xnn_params.f32.dwconv2d_chw_5x5 = (struct dwconv2d_chw_parameters) { |
| .ukernel = (xnn_dwconv2d_chw_ukernel_function) xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_4x4, |
| .output_width_tile = 4, |
| .output_height_tile = 4, |
| }; |
| xnn_params.f32.dwconv2d_chw_5x5s2 = (struct dwconv2d_chw_parameters) { |
| .ukernel = (xnn_dwconv2d_chw_ukernel_function) xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__sse_2x4, |
| .output_width_tile = 4, |
| .output_height_tile = 2, |
| }; |
| xnn_params.f32.gavgpool_cw = (struct gavgpool_cw_parameters) { |
| .ukernel = (xnn_gavgpool_cw_ukernel_function) xnn_f32_gavgpool_cw_ukernel__sse_x4, |
| .channel_tile = 4, |
| }; |
| xnn_params.f32.ibilinear_chw = (struct ibilinear_chw_parameters) { |
| .ukernel = (xnn_ibilinear_chw_ukernel_function) xnn_f32_ibilinear_chw_ukernel__sse_p8, |
| .channel_tile = 1, |
| .pixel_tile = 8, |
| }; |
| #endif // XNN_NO_NCHW_OPERATORS |
| #endif // XNN_NO_F32_OPERATORS |
| |
| /*************************** VCVT micro-kernels ***************************/ |
| #ifndef XNN_NO_VCVT_OPERATORS |
| init_flags |= XNN_INIT_FLAG_VCVT; |
| |
| if (cpuinfo_has_x86_avx512f() && cpuinfo_has_x86_avx512bw() && cpuinfo_has_x86_avx512dq() && cpuinfo_has_x86_avx512vl()) { |
| xnn_params.vcvt.f16_to_f32 = (xnn_univector_ukernel_function) xnn_f16_f32_vcvt_ukernel__avx512skx_x16; |
| } else if (cpuinfo_has_x86_f16c()) { |
| xnn_params.vcvt.f16_to_f32 = (xnn_univector_ukernel_function) xnn_f16_f32_vcvt_ukernel__f16c_x16; |
| } else if (cpuinfo_has_x86_avx()) { |
| xnn_params.vcvt.f16_to_f32 = (xnn_univector_ukernel_function) xnn_f16_f32_vcvt_ukernel__avx_int16_x16; |
| } else if (cpuinfo_has_x86_sse4_1()) { |
| xnn_params.vcvt.f16_to_f32 = (xnn_univector_ukernel_function) xnn_f16_f32_vcvt_ukernel__sse41_int16_x16; |
| } else { |
| xnn_params.vcvt.f16_to_f32 = (xnn_univector_ukernel_function) xnn_f16_f32_vcvt_ukernel__sse2_int16_x32; |
| } |
| #endif // XNN_NO_VCVT_OPERATORS |
| |
| /**************************** X32 micro-kernels ****************************/ |
| #ifndef XNN_NO_X32_OPERATORS |
| init_flags |= XNN_INIT_FLAG_X32; |
| |
| xnn_params.x32.unpool = (xnn_unpool_ukernel_function) xnn_x32_unpool_ukernel__sse2; |
| xnn_params.x32.zip = (struct zip_parameters) { |
| .x2 = (xnn_zipc_ukernel_function) xnn_x32_zip_x2_ukernel__sse2, |
| .x3 = (xnn_zipc_ukernel_function) xnn_x32_zip_x3_ukernel__sse2, |
| .x4 = (xnn_zipc_ukernel_function) xnn_x32_zip_x4_ukernel__sse2, |
| .xm = (xnn_zipv_ukernel_function) xnn_x32_zip_xm_ukernel__sse2, |
| }; |
| #ifndef XNN_NO_NCHW_OPERATORS |
| xnn_params.x32.depthtospace2d_chw2hwc = (struct depthtospace2d_chw2hwc_parameters) { |
| .ukernel = (xnn_depthtospace2d_chw2hwc_ukernel_function) xnn_x32_depthtospace2d_chw2hwc_ukernel__scalar, |
| .channel_tile = 1, |
| .pixel_tile = 1, |
| }; |
| #endif // XNN_NO_NCHW_OPERATORS |
| #endif // XNN_NO_X32_OPERATORS |
| |
| /**************************** XX micro-kernels ****************************/ |
| #ifndef XNN_NO_XX_OPERATORS |
| init_flags |= XNN_INIT_FLAG_XX; |
| |
| xnn_params.xx.copy = (xnn_univector_ukernel_function) xnn_xx_copy_ukernel__memcpy; |
| xnn_params.xx.fill = (struct fill_parameters) { |
| .ukernel = (xnn_fill_ukernel_function) xnn_xx_fill_ukernel__sse2_x64, |
| .row_tile = 1, |
| }; |
| xnn_params.xx.pad = (struct pad_parameters) { |
| .ukernel = (xnn_pad_ukernel_function) xnn_xx_pad_ukernel__sse2, |
| .row_tile = 1, |
| }; |
| #endif |
| |
| #elif XNN_ARCH_WASMSIMD |
| |
| /**************************** QC8 micro-kernels ****************************/ |
| #ifndef XNN_NO_QS8_OPERATORS |
| init_flags |= XNN_INIT_FLAG_QC8; |
| |
| #if defined(XNN_WASMSIMD_VERSION) && (XNN_WASMSIMD_VERSION >= 88) |
| xnn_params.qc8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_3x4c2__wasmsimd_dot16x2_ld128); |
| xnn_params.qc8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_3x4c2__wasmsimd_dot16x2_ld128); |
| xnn_params.qc8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_1x4c2__wasmsimd_dot16x2_ld128); |
| xnn_params.qc8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_1x4c2__wasmsimd_dot16x2_ld128); |
| xnn_params.qc8.gemm.init.qc8 = xnn_init_qs8_minmax_wasmsimd_params; |
| xnn_params.qc8.gemm.mr = 3; |
| xnn_params.qc8.gemm.nr = 4; |
| xnn_params.qc8.gemm.log2_kr = 1; |
| #else |
| xnn_params.qc8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_3x4c8__wasmsimd_mul16_ld64); |
| xnn_params.qc8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_3x4c8__wasmsimd_mul16_ld64); |
| xnn_params.qc8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_1x4c8__wasmsimd_mul16_ld64); |
| xnn_params.qc8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_1x4c8__wasmsimd_mul16_ld64); |
| xnn_params.qc8.gemm.init.qc8 = xnn_init_qs8_minmax_wasmsimd_params; |
| xnn_params.qc8.gemm.mr = 3; |
| xnn_params.qc8.gemm.nr = 4; |
| xnn_params.qc8.gemm.log2_kr = 3; |
| #endif |
| |
| xnn_params.qc8.dwconv[0].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qc8_dwconv_minmax_fp32_ukernel_up16x9__wasmsimd_mul16_add16; |
| xnn_params.qc8.dwconv[0].init.qc8 = xnn_init_qs8_minmax_wasmsimd_params; |
| xnn_params.qc8.dwconv[0].channel_tile = 16; |
| xnn_params.qc8.dwconv[0].primary_tile = 9; |
| xnn_params.qc8.dwconv[1].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qc8_dwconv_minmax_fp32_ukernel_up16x25__wasmsimd_mul16_add16; |
| xnn_params.qc8.dwconv[1].init.qc8 = xnn_init_qs8_minmax_wasmsimd_params; |
| xnn_params.qc8.dwconv[1].channel_tile = 16; |
| xnn_params.qc8.dwconv[1].primary_tile = 25; |
| #endif // XNN_NO_QC8_OPERATORS |
| |
| /**************************** QS8 micro-kernels ****************************/ |
| #ifndef XNN_NO_QS8_OPERATORS |
| init_flags |= XNN_INIT_FLAG_QS8; |
| |
| #if defined(XNN_WASMSIMD_VERSION) && (XNN_WASMSIMD_VERSION >= 88) |
| xnn_params.qs8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_fp32_ukernel_3x4c2__wasmsimd_dot16x2_ld128); |
| xnn_params.qs8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_fp32_ukernel_3x4c2__wasmsimd_dot16x2_ld128); |
| xnn_params.qs8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_fp32_ukernel_1x4c2__wasmsimd_dot16x2_ld128); |
| xnn_params.qs8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_fp32_ukernel_1x4c2__wasmsimd_dot16x2_ld128); |
| xnn_params.qs8.gemm.init.qs8 = xnn_init_qs8_conv_minmax_fp32_wasmsimd_params; |
| xnn_params.qs8.gemm.mr = 3; |
| xnn_params.qs8.gemm.nr = 4; |
| xnn_params.qs8.gemm.log2_kr = 1; |
| #else // XNN_WASMSIMD_VERSION >= 88 |
| xnn_params.qs8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_fp32_ukernel_3x4c8__wasmsimd_mul16_ld64); |
| xnn_params.qs8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_fp32_ukernel_3x4c8__wasmsimd_mul16_ld64); |
| xnn_params.qs8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_fp32_ukernel_1x4c8__wasmsimd_mul16_ld64); |
| xnn_params.qs8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_fp32_ukernel_1x4c8__wasmsimd_mul16_ld64); |
| xnn_params.qs8.gemm.init.qs8 = xnn_init_qs8_conv_minmax_fp32_wasmsimd_params; |
| xnn_params.qs8.gemm.mr = 3; |
| xnn_params.qs8.gemm.nr = 4; |
| xnn_params.qs8.gemm.log2_kr = 3; |
| #endif // XNN_WASMSIMD_VERSION >= 88 |
| |
| xnn_params.qs8.dwconv[0].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qs8_dwconv_minmax_fp32_ukernel_up16x9__wasmsimd_mul16_add16; |
| xnn_params.qs8.dwconv[0].init.qs8 = xnn_init_qs8_conv_minmax_fp32_wasmsimd_params; |
| xnn_params.qs8.dwconv[0].channel_tile = 16; |
| xnn_params.qs8.dwconv[0].primary_tile = 9; |
| xnn_params.qs8.dwconv[1].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qs8_dwconv_minmax_fp32_ukernel_up16x25__wasmsimd_mul16_add16; |
| xnn_params.qs8.dwconv[1].init.qs8 = xnn_init_qs8_conv_minmax_fp32_wasmsimd_params; |
| xnn_params.qs8.dwconv[1].channel_tile = 16; |
| xnn_params.qs8.dwconv[1].primary_tile = 25; |
| |
| xnn_params.qs8.gavgpool = (struct gavgpool_parameters) { |
| .up = (xnn_gavgpool_unipass_ukernel_function) xnn_qs8_gavgpool_minmax_ukernel_7x__wasmsimd_c8_acc2, |
| .mp = (xnn_gavgpool_multipass_ukernel_function) xnn_qs8_gavgpool_minmax_ukernel_7p7x__wasmsimd_c8_acc2, |
| .mr = 7, |
| }; |
| |
| xnn_params.qs8.vadd = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vadd_minmax_ukernel__wasmsimd_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vaddc_minmax_ukernel__wasmsimd_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vaddc_minmax_ukernel__wasmsimd_x8, |
| .init.qs8_addsub = xnn_init_qs8_add_minmax_wasmsimd_params, |
| .element_tile = 8, |
| }; |
| xnn_params.qs8.vmul = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vmul_minmax_fp32_ukernel__wasmsimd_mul32_ld64_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vmulc_minmax_fp32_ukernel__wasmsimd_mul32_ld64_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vmulc_minmax_fp32_ukernel__wasmsimd_mul32_ld64_x8, |
| .init.qs8_mul = xnn_init_qs8_mul_minmax_fp32_wasmsimd_params, |
| .element_tile = 8, |
| }; |
| #endif // XNN_NO_QS8_OPERATORS |
| |
| /**************************** QU8 micro-kernels ****************************/ |
| #ifndef XNN_NO_QU8_OPERATORS |
| init_flags |= XNN_INIT_FLAG_QU8; |
| |
| #if defined(XNN_WASMSIMD_VERSION) && (XNN_WASMSIMD_VERSION >= 88) |
| xnn_params.qu8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qu8_gemm_minmax_fp32_ukernel_3x4c2__wasmsimd_dot16x2_ld128); |
| xnn_params.qu8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qu8_igemm_minmax_fp32_ukernel_3x4c2__wasmsimd_dot16x2_ld128); |
| xnn_params.qu8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qu8_gemm_minmax_fp32_ukernel_1x4c2__wasmsimd_dot16x2_ld128); |
| xnn_params.qu8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qu8_igemm_minmax_fp32_ukernel_1x4c2__wasmsimd_dot16x2_ld128); |
| xnn_params.qu8.gemm.init.qu8 = xnn_init_qu8_conv_minmax_fp32_wasmsimd_params; |
| xnn_params.qu8.gemm.mr = 3; |
| xnn_params.qu8.gemm.nr = 4; |
| xnn_params.qu8.gemm.log2_kr = 1; |
| #else // XNN_WASMSIMD_VERSION >= 88 |
| xnn_params.qu8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qu8_gemm_minmax_fp32_ukernel_3x4c8__wasmsimd_mul32_ld64); |
| xnn_params.qu8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qu8_igemm_minmax_fp32_ukernel_3x4c8__wasmsimd_mul32_ld64); |
| xnn_params.qu8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qu8_gemm_minmax_fp32_ukernel_1x4c8__wasmsimd_mul32_ld64); |
| xnn_params.qu8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qu8_igemm_minmax_fp32_ukernel_1x4c8__wasmsimd_mul32_ld64); |
| xnn_params.qu8.gemm.init.qu8 = xnn_init_qu8_conv_minmax_fp32_wasmsimd_params; |
| xnn_params.qu8.gemm.mr = 3; |
| xnn_params.qu8.gemm.nr = 4; |
| xnn_params.qu8.gemm.log2_kr = 3; |
| #endif // XNN_WASMSIMD_VERSION >= 88 |
| |
| xnn_params.qu8.dwconv[0].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qu8_dwconv_minmax_fp32_ukernel_up8x9__wasmsimd_mul16; |
| xnn_params.qu8.dwconv[0].init.qu8 = xnn_init_qu8_conv_minmax_fp32_wasmsimd_params; |
| xnn_params.qu8.dwconv[0].channel_tile = 8; |
| xnn_params.qu8.dwconv[0].primary_tile = 9; |
| xnn_params.qu8.dwconv[1].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qu8_dwconv_minmax_fp32_ukernel_up8x25__wasmsimd_mul16; |
| xnn_params.qu8.dwconv[1].init.qu8 = xnn_init_qu8_conv_minmax_fp32_wasmsimd_params; |
| xnn_params.qu8.dwconv[1].channel_tile = 8; |
| xnn_params.qu8.dwconv[1].primary_tile = 25; |
| |
| xnn_params.qu8.avgpool = (struct avgpool_parameters) { |
| .up = (xnn_avgpool_unipass_ukernel_function) xnn_qu8_avgpool_minmax_ukernel_9x__scalar_c1, |
| .mp = (xnn_avgpool_multipass_ukernel_function) xnn_qu8_avgpool_minmax_ukernel_9p8x__scalar_c1, |
| .mr = 9, |
| .qr = 8, |
| }; |
| xnn_params.qu8.gavgpool = (struct gavgpool_parameters) { |
| .up = (xnn_gavgpool_unipass_ukernel_function) xnn_qu8_gavgpool_minmax_ukernel_7x__scalar_c1, |
| .mp = (xnn_gavgpool_multipass_ukernel_function) xnn_qu8_gavgpool_minmax_ukernel_7p7x__scalar_c1, |
| .mr = 7, |
| }; |
| |
| xnn_params.qu8.vadd = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vadd_minmax_ukernel__wasmsimd_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vaddc_minmax_ukernel__wasmsimd_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vaddc_minmax_ukernel__wasmsimd_x8, |
| .init.qu8_addsub = xnn_init_qu8_add_minmax_wasmsimd_params, |
| .element_tile = 8, |
| }; |
| xnn_params.qu8.vmul = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vmul_minmax_fp32_ukernel__wasmsimd_mul32_ld64_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vmulc_minmax_fp32_ukernel__wasmsimd_mul32_ld64_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vmulc_minmax_fp32_ukernel__wasmsimd_mul32_ld64_x8, |
| .init.qu8_mul = xnn_init_qu8_mul_minmax_fp32_wasmsimd_params, |
| .element_tile = 8, |
| }; |
| #endif // XNN_NO_QU8_OPERATORS |
| |
| /**************************** S8 micro-kernels ****************************/ |
| #ifndef XNN_NO_S8_OPERATORS |
| init_flags |= XNN_INIT_FLAG_S8; |
| |
| xnn_params.s8.clamp = (struct vunary_parameters) { |
| .ukernel = (xnn_univector_ukernel_function) xnn_s8_vclamp_ukernel__wasmsimd_x64, |
| .init.s8_minmax = xnn_init_s8_minmax_wasmsimd_params, |
| .element_tile = 64, |
| }; |
| xnn_params.s8.maxpool = (struct maxpool_parameters) { |
| .ukernel = (xnn_maxpool_ukernel_function) xnn_s8_maxpool_minmax_ukernel_9p8x__wasmsimd_c16, |
| .init.s8 = xnn_init_s8_minmax_wasmsimd_params, |
| .mr = 9, |
| .qr = 8, |
| }; |
| #endif // XNN_NO_S8_OPERATORS |
| |
| /**************************** U8 micro-kernels ****************************/ |
| #ifndef XNN_NO_U8_OPERATORS |
| init_flags |= XNN_INIT_FLAG_U8; |
| |
| xnn_params.u8.clamp = (struct vunary_parameters) { |
| .ukernel = (xnn_univector_ukernel_function) xnn_u8_vclamp_ukernel__wasmsimd_x64, |
| .init.u8_minmax = xnn_init_u8_minmax_wasmsimd_params, |
| .element_tile = 64, |
| }; |
| xnn_params.u8.maxpool = (struct maxpool_parameters) { |
| .ukernel = (xnn_maxpool_ukernel_function) xnn_u8_maxpool_minmax_ukernel_9p8x__wasmsimd_c16, |
| .init.u8 = xnn_init_u8_minmax_wasmsimd_params, |
| .mr = 9, |
| .qr = 8, |
| }; |
| xnn_params.u8.lut32norm = xnn_u8_lut32norm_ukernel__scalar; |
| xnn_params.u8.rmax = xnn_u8_rmax_ukernel__scalar; |
| #endif // XNN_NO_U8_OPERATORS |
| |
| /**************************** X8 micro-kernels ****************************/ |
| #ifndef XNN_NO_X8_OPERATORS |
| init_flags |= XNN_INIT_FLAG_X8; |
| |
| xnn_params.x8.lut = xnn_x8_lut_ukernel__scalar_x4; |
| xnn_params.x8.zip = (struct zip_parameters) { |
| .x2 = (xnn_zipc_ukernel_function) xnn_x8_zip_x2_ukernel__scalar, |
| .x3 = (xnn_zipc_ukernel_function) xnn_x8_zip_x3_ukernel__scalar, |
| .x4 = (xnn_zipc_ukernel_function) xnn_x8_zip_x4_ukernel__scalar, |
| .xm = (xnn_zipv_ukernel_function) xnn_x8_zip_xm_ukernel__scalar, |
| }; |
| #endif // XNN_NO_X8_OPERATORS |
| |
| /**************************** F32 micro-kernels ****************************/ |
| #ifndef XNN_NO_F32_OPERATORS |
| init_flags |= XNN_INIT_FLAG_F32; |
| |
| if (is_wasm_x86) { |
| xnn_params.f32.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_4x8__wasmsimd_x86_splat); |
| xnn_params.f32.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_4x8__wasmsimd_x86_splat); |
| xnn_params.f32.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_1x8__wasmsimd_x86_splat); |
| xnn_params.f32.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_1x8__wasmsimd_x86_splat); |
| xnn_params.f32.gemm.relu.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_relu_ukernel_4x8__wasmsimd_splat); |
| xnn_params.f32.gemm.relu.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_relu_ukernel_4x8__wasmsimd_splat); |
| xnn_params.f32.gemm.relu.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_relu_ukernel_1x8__wasmsimd_splat); |
| xnn_params.f32.gemm.relu.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_relu_ukernel_1x8__wasmsimd_splat); |
| xnn_params.f32.gemm.linear.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_ukernel_4x8__wasmsimd_splat); |
| xnn_params.f32.gemm.linear.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_ukernel_4x8__wasmsimd_splat); |
| xnn_params.f32.gemm.linear.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_ukernel_1x8__wasmsimd_splat); |
| xnn_params.f32.gemm.linear.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_ukernel_1x8__wasmsimd_splat); |
| xnn_params.f32.gemm.init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.gemm.mr = 4; |
| xnn_params.f32.gemm.nr = 8; |
| |
| xnn_params.f32.gemm2.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_4x2c4__wasmsimd_x86); |
| xnn_params.f32.gemm2.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_4x2c4__wasmsimd_x86); |
| xnn_params.f32.gemm2.linear.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_ukernel_4x2c4__wasmsimd); |
| xnn_params.f32.gemm2.linear.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_ukernel_4x2c4__wasmsimd); |
| xnn_params.f32.gemm2.init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.gemm2.mr = 4; |
| xnn_params.f32.gemm2.nr = 2; |
| xnn_params.f32.gemm2.log2_kr = 2; |
| } else { |
| xnn_params.f32.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_5x8__wasmsimd_arm_splat); |
| xnn_params.f32.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_5x8__wasmsimd_arm_splat); |
| xnn_params.f32.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_1x8__wasmsimd_arm_splat); |
| xnn_params.f32.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_1x8__wasmsimd_arm_splat); |
| xnn_params.f32.gemm.relu.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_relu_ukernel_5x8__wasmsimd_splat); |
| xnn_params.f32.gemm.relu.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_relu_ukernel_5x8__wasmsimd_splat); |
| xnn_params.f32.gemm.relu.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_relu_ukernel_1x8__wasmsimd_splat); |
| xnn_params.f32.gemm.relu.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_relu_ukernel_1x8__wasmsimd_splat); |
| xnn_params.f32.gemm.linear.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_ukernel_5x8__wasmsimd_splat); |
| xnn_params.f32.gemm.linear.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_ukernel_5x8__wasmsimd_splat); |
| xnn_params.f32.gemm.linear.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_ukernel_1x8__wasmsimd_splat); |
| xnn_params.f32.gemm.linear.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_ukernel_1x8__wasmsimd_splat); |
| xnn_params.f32.gemm.init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.gemm.mr = 5; |
| xnn_params.f32.gemm.nr = 8; |
| |
| xnn_params.f32.gemm2.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_4x2c4__wasmsimd_arm); |
| xnn_params.f32.gemm2.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_4x2c4__wasmsimd_arm); |
| xnn_params.f32.gemm2.linear.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_ukernel_4x2c4__wasmsimd); |
| xnn_params.f32.gemm2.linear.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_ukernel_4x2c4__wasmsimd); |
| xnn_params.f32.gemm2.init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.gemm2.mr = 4; |
| xnn_params.f32.gemm2.nr = 2; |
| xnn_params.f32.gemm2.log2_kr = 2; |
| } |
| |
| if (is_wasm_x86) { |
| xnn_params.f32.dwconv[0].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_minmax_ukernel_up8x3__wasmsimd_x86; |
| xnn_params.f32.dwconv[0].linear.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_ukernel_up8x3__wasmsimd; |
| xnn_params.f32.dwconv[0].init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.dwconv[0].channel_tile = 8; |
| xnn_params.f32.dwconv[0].primary_tile = 3; |
| |
| xnn_params.f32.dwconv[1].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_minmax_ukernel_up8x4__wasmsimd_x86; |
| xnn_params.f32.dwconv[1].linear.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_ukernel_up8x4__wasmsimd; |
| xnn_params.f32.dwconv[1].init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.dwconv[1].channel_tile = 8; |
| xnn_params.f32.dwconv[1].primary_tile = 4; |
| |
| xnn_params.f32.dwconv[2].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_minmax_ukernel_up8x9__wasmsimd_x86; |
| xnn_params.f32.dwconv[2].linear.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_ukernel_up8x9__wasmsimd; |
| xnn_params.f32.dwconv[2].init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.dwconv[2].channel_tile = 8; |
| xnn_params.f32.dwconv[2].primary_tile = 9; |
| } else { |
| xnn_params.f32.dwconv[0].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_minmax_ukernel_up4x3__wasmsimd_arm; |
| xnn_params.f32.dwconv[0].linear.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_ukernel_up4x3__wasmsimd; |
| xnn_params.f32.dwconv[0].init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.dwconv[0].channel_tile = 4; |
| xnn_params.f32.dwconv[0].primary_tile = 3; |
| |
| xnn_params.f32.dwconv[1].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_minmax_ukernel_up4x4__wasmsimd_arm; |
| xnn_params.f32.dwconv[1].linear.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_ukernel_up4x4__wasmsimd; |
| xnn_params.f32.dwconv[1].init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.dwconv[1].channel_tile = 4; |
| xnn_params.f32.dwconv[1].primary_tile = 4; |
| |
| xnn_params.f32.dwconv[2].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_minmax_ukernel_up4x9__wasmsimd_arm; |
| xnn_params.f32.dwconv[2].linear.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_ukernel_up4x9__wasmsimd; |
| xnn_params.f32.dwconv[2].init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.dwconv[2].channel_tile = 4; |
| xnn_params.f32.dwconv[2].primary_tile = 9; |
| } |
| |
| xnn_params.f32.dwconv[3].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_minmax_ukernel_up4x25__wasmsimd_arm; |
| xnn_params.f32.dwconv[3].linear.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_ukernel_up4x25__wasmsimd; |
| xnn_params.f32.dwconv[3].init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.dwconv[3].channel_tile = 4; |
| xnn_params.f32.dwconv[3].primary_tile = 25; |
| |
| if (is_wasm_x86) { |
| xnn_params.f32.avgpool = (struct avgpool_parameters) { |
| .up = (xnn_avgpool_unipass_ukernel_function) xnn_f32_avgpool_minmax_ukernel_9x__wasmsimd_x86_c4, |
| .mp = (xnn_avgpool_multipass_ukernel_function) xnn_f32_avgpool_minmax_ukernel_9p8x__wasmsimd_x86_c4, |
| .mr = 9, |
| .qr = 8, |
| }; |
| xnn_params.f32.pavgpool = (struct pavgpool_parameters) { |
| .up = (xnn_pavgpool_unipass_ukernel_function) xnn_f32_pavgpool_minmax_ukernel_9x__wasmsimd_x86_c4, |
| .mp = (xnn_pavgpool_multipass_ukernel_function) xnn_f32_pavgpool_minmax_ukernel_9p8x__wasmsimd_x86_c4, |
| .mr = 9, |
| .qr = 8, |
| }; |
| xnn_params.f32.gavgpool = (struct gavgpool_parameters) { |
| .up = (xnn_gavgpool_unipass_ukernel_function) xnn_f32_gavgpool_minmax_ukernel_7x__wasmsimd_x86_c4, |
| .mp = (xnn_gavgpool_multipass_ukernel_function) xnn_f32_gavgpool_minmax_ukernel_7p7x__wasmsimd_x86_c4, |
| .mr = 7, |
| }; |
| } else { |
| xnn_params.f32.avgpool = (struct avgpool_parameters) { |
| .up = (xnn_avgpool_unipass_ukernel_function) xnn_f32_avgpool_minmax_ukernel_9x__wasmsimd_arm_c4, |
| .mp = (xnn_avgpool_multipass_ukernel_function) xnn_f32_avgpool_minmax_ukernel_9p8x__wasmsimd_arm_c4, |
| .mr = 9, |
| .qr = 8, |
| }; |
| xnn_params.f32.pavgpool = (struct pavgpool_parameters) { |
| .up = (xnn_pavgpool_unipass_ukernel_function) xnn_f32_pavgpool_minmax_ukernel_9x__wasmsimd_arm_c4, |
| .mp = (xnn_pavgpool_multipass_ukernel_function) xnn_f32_pavgpool_minmax_ukernel_9p8x__wasmsimd_arm_c4, |
| .mr = 9, |
| .qr = 8, |
| }; |
| xnn_params.f32.gavgpool = (struct gavgpool_parameters) { |
| .up = (xnn_gavgpool_unipass_ukernel_function) xnn_f32_gavgpool_minmax_ukernel_7x__wasmsimd_arm_c4, |
| .mp = (xnn_gavgpool_multipass_ukernel_function) xnn_f32_gavgpool_minmax_ukernel_7p7x__wasmsimd_arm_c4, |
| .mr = 7, |
| }; |
| } |
| if (is_wasm_x86) { |
| xnn_params.f32.maxpool = (struct maxpool_parameters) { |
| .ukernel = (xnn_maxpool_ukernel_function) xnn_f32_maxpool_minmax_ukernel_9p8x__wasmsimd_x86_c4, |
| .init.f32 = xnn_init_f32_minmax_scalar_params, |
| .mr = 9, |
| .qr = 8, |
| }; |
| } else { |
| xnn_params.f32.maxpool = (struct maxpool_parameters) { |
| .ukernel = (xnn_maxpool_ukernel_function) xnn_f32_maxpool_minmax_ukernel_9p8x__wasmsimd_arm_c4, |
| .init.f32 = xnn_init_f32_minmax_scalar_params, |
| .mr = 9, |
| .qr = 8, |
| }; |
| } |
| xnn_params.f32.argmaxpool[0] = (struct argmaxpool_parameters) { |
| .up = (xnn_argmaxpool_unipass_ukernel_function) xnn_f32_argmaxpool_ukernel_4x__wasmsimd_c4, |
| .mr = 4, |
| }; |
| xnn_params.f32.argmaxpool[1] = (struct argmaxpool_parameters) { |
| .up = (xnn_argmaxpool_unipass_ukernel_function) xnn_f32_argmaxpool_ukernel_9x__wasmsimd_c4, |
| .mr = 9, |
| }; |
| xnn_params.f32.argmaxpool[2] = (struct argmaxpool_parameters) { |
| .mp = (xnn_argmaxpool_multipass_ukernel_function) xnn_f32_argmaxpool_ukernel_9p8x__wasmsimd_c4, |
| .mr = 9, |
| .qr = 8, |
| }; |
| xnn_params.f32.ibilinear = (struct ibilinear_parameters) { |
| .ukernel = (xnn_ibilinear_ukernel_function) xnn_f32_ibilinear_ukernel__wasmsimd_c8, |
| .pixel_tile = 1, |
| .channel_tile = 8, |
| }; |
| xnn_params.f32.abs = (xnn_univector_ukernel_function) xnn_f32_vabs_ukernel__wasmsimd_x8; |
| if (is_wasm_x86) { |
| xnn_params.f32.clamp = (struct vunary_parameters) { |
| .ukernel = (xnn_univector_ukernel_function) xnn_f32_vclamp_ukernel__wasmsimd_x86_x8, |
| .init.f32_minmax = xnn_init_f32_minmax_scalar_params, |
| .element_tile = 8, |
| }; |
| } else { |
| xnn_params.f32.clamp = (struct vunary_parameters) { |
| .ukernel = (xnn_univector_ukernel_function) xnn_f32_vclamp_ukernel__wasmsimd_arm_x8, |
| .init.f32_minmax = xnn_init_f32_minmax_scalar_params, |
| .element_tile = 8, |
| }; |
| } |
| if (is_wasm_x86) { |
| xnn_params.f32.elu = (xnn_univector_ukernel_function) xnn_f32_velu_ukernel__wasmsimd_x86_rr2_p6_x20; |
| } else { |
| xnn_params.f32.elu = (xnn_univector_ukernel_function) xnn_f32_velu_ukernel__wasmsimd_arm_rr2_p6_x20; |
| } |
| xnn_params.f32.hswish = (xnn_univector_ukernel_function) xnn_f32_vhswish_ukernel__wasmsimd_x16; |
| if (is_wasm_x86) { |
| xnn_params.f32.lrelu = (xnn_univector_ukernel_function) xnn_f32_vlrelu_ukernel__wasmsimd_minmax_x8; |
| } else { |
| xnn_params.f32.lrelu = (xnn_univector_ukernel_function) xnn_f32_vlrelu_ukernel__wasmsimd_bitselect_x8; |
| } |
| xnn_params.f32.neg = (xnn_univector_ukernel_function) xnn_f32_vneg_ukernel__wasmsimd_x8; |
| xnn_params.f32.relu = (xnn_univector_ukernel_function) xnn_f32_vrelu_ukernel__wasmsimd_x16; |
| #if defined(XNN_WASMSIMD_VERSION) && (XNN_WASMSIMD_VERSION >= 91) |
| xnn_params.f32.rndne = (xnn_univector_ukernel_function) xnn_f32_vrndne_ukernel__wasmsimd_native_x8; |
| xnn_params.f32.rndz = (xnn_univector_ukernel_function) xnn_f32_vrndz_ukernel__wasmsimd_native_x8; |
| xnn_params.f32.rndu = (xnn_univector_ukernel_function) xnn_f32_vrndu_ukernel__wasmsimd_native_x8; |
| xnn_params.f32.rndd = (xnn_univector_ukernel_function) xnn_f32_vrndd_ukernel__wasmsimd_native_x8; |
| #else // XNN_WASMSIMD_VERSION >= 91 |
| xnn_params.f32.rndne = (xnn_univector_ukernel_function) xnn_f32_vrndne_ukernel__wasmsimd_addsub_x8; |
| if (is_wasm_x86) { |
| xnn_params.f32.rndz = (xnn_univector_ukernel_function) xnn_f32_vrndz_ukernel__wasmsimd_addsub_x8; |
| } else { |
| xnn_params.f32.rndz = (xnn_univector_ukernel_function) xnn_f32_vrndz_ukernel__wasmsimd_cvt_x8; |
| } |
| xnn_params.f32.rndu = (xnn_univector_ukernel_function) xnn_f32_vrndu_ukernel__wasmsimd_addsub_x8; |
| xnn_params.f32.rndd = (xnn_univector_ukernel_function) xnn_f32_vrndd_ukernel__wasmsimd_addsub_x8; |
| #endif // XNN_WASMSIMD_VERSION >= 91 |
| xnn_params.f32.sigmoid = (xnn_univector_ukernel_function) xnn_f32_vsigmoid_ukernel__wasmsimd_p5_div_x16; |
| xnn_params.f32.sqr = (xnn_univector_ukernel_function) xnn_f32_vsqr_ukernel__wasmsimd_x8; |
| xnn_params.f32.sqrt = (xnn_univector_ukernel_function) xnn_f32_vsqrt_ukernel__wasmsimd_sqrt_x8; |
| if (is_wasm_x86) { |
| xnn_params.f32.prelu = (struct prelu_parameters) { |
| .ukernel = (xnn_prelu_ukernel_function) xnn_f32_prelu_ukernel__wasmsimd_minmax_2x8, |
| .row_tile = 2, |
| .channel_tile = 8, |
| }; |
| } else { |
| xnn_params.f32.prelu = (struct prelu_parameters) { |
| .ukernel = (xnn_prelu_ukernel_function) xnn_f32_prelu_ukernel__wasmsimd_bitselect_2x8, |
| .row_tile = 2, |
| .channel_tile = 8, |
| }; |
| } |
| xnn_params.f32.raddstoreexpminusmax = xnn_f32_raddstoreexpminusmax_ukernel__wasmsimd_p5_x16_acc2; |
| if (is_wasm_x86) { |
| xnn_params.f32.rmax = xnn_f32_rmax_ukernel__wasmsimd_x86; |
| xnn_params.f32.vadd = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vadd_minmax_ukernel__wasmsimd_x86_x16, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vaddc_minmax_ukernel__wasmsimd_x86_x16, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vaddc_minmax_ukernel__wasmsimd_x86_x16, |
| .linear.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vadd_ukernel__wasmsimd_x16, |
| .linear.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vaddc_ukernel__wasmsimd_x16, |
| .linear.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vaddc_ukernel__wasmsimd_x16, |
| .element_tile = 16, |
| }; |
| xnn_params.f32.vdiv = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vdiv_minmax_ukernel__wasmsimd_x86_x16, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vdivc_minmax_ukernel__wasmsimd_x86_x16, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vrdivc_minmax_ukernel__wasmsimd_x86_x16, |
| .linear.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vdiv_ukernel__wasmsimd_x16, |
| .linear.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vdivc_ukernel__wasmsimd_x16, |
| .linear.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vrdivc_ukernel__wasmsimd_x16, |
| .element_tile = 16, |
| }; |
| xnn_params.f32.vmax = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmax_ukernel__wasmsimd_x86_x16, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmaxc_ukernel__wasmsimd_x86_x16, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmaxc_ukernel__wasmsimd_x86_x16, |
| .element_tile = 16, |
| }; |
| xnn_params.f32.vmin = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmin_ukernel__wasmsimd_x86_x16, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vminc_ukernel__wasmsimd_x86_x16, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vminc_ukernel__wasmsimd_x86_x16, |
| |
| .element_tile = 16, |
| }; |
| xnn_params.f32.vmul = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmul_minmax_ukernel__wasmsimd_x86_x16, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmulc_minmax_ukernel__wasmsimd_x86_x16, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmulc_minmax_ukernel__wasmsimd_x86_x16, |
| .linear.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmul_ukernel__wasmsimd_x16, |
| .linear.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmulc_ukernel__wasmsimd_x16, |
| .linear.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmulc_ukernel__wasmsimd_x16, |
| .element_tile = 16, |
| }; |
| xnn_params.f32.vsub = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsub_minmax_ukernel__wasmsimd_x86_x16, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsubc_minmax_ukernel__wasmsimd_x86_x16, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vrsubc_minmax_ukernel__wasmsimd_x86_x16, |
| .linear.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsub_ukernel__wasmsimd_x16, |
| .linear.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsubc_ukernel__wasmsimd_x16, |
| .linear.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vrsubc_ukernel__wasmsimd_x16, |
| .element_tile = 16, |
| }; |
| } else { |
| xnn_params.f32.rmax = xnn_f32_rmax_ukernel__wasmsimd_arm; |
| xnn_params.f32.vadd = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vadd_minmax_ukernel__wasmsimd_arm_x16, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vaddc_minmax_ukernel__wasmsimd_arm_x16, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vaddc_minmax_ukernel__wasmsimd_arm_x16, |
| .linear.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vadd_ukernel__wasmsimd_x16, |
| .linear.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vaddc_ukernel__wasmsimd_x16, |
| .linear.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vaddc_ukernel__wasmsimd_x16, |
| .element_tile = 16, |
| }; |
| xnn_params.f32.vdiv = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vdiv_minmax_ukernel__wasmsimd_arm_x16, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vdivc_minmax_ukernel__wasmsimd_arm_x16, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vrdivc_minmax_ukernel__wasmsimd_arm_x16, |
| .linear.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vdiv_ukernel__wasmsimd_x16, |
| .linear.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vdivc_ukernel__wasmsimd_x16, |
| .linear.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vrdivc_ukernel__wasmsimd_x16, |
| .element_tile = 16, |
| }; |
| xnn_params.f32.vmax = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmax_ukernel__wasmsimd_arm_x16, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmaxc_ukernel__wasmsimd_arm_x16, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmaxc_ukernel__wasmsimd_arm_x16, |
| .element_tile = 16, |
| }; |
| xnn_params.f32.vmin = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmin_ukernel__wasmsimd_arm_x16, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vminc_ukernel__wasmsimd_arm_x16, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vminc_ukernel__wasmsimd_arm_x16, |
| .element_tile = 16, |
| }; |
| xnn_params.f32.vmul = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmul_minmax_ukernel__wasmsimd_arm_x16, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmulc_minmax_ukernel__wasmsimd_arm_x16, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmulc_minmax_ukernel__wasmsimd_arm_x16, |
| .linear.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmul_ukernel__wasmsimd_x16, |
| .linear.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmulc_ukernel__wasmsimd_x16, |
| .linear.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmulc_ukernel__wasmsimd_x16, |
| .element_tile = 16, |
| }; |
| xnn_params.f32.vsub = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsub_minmax_ukernel__wasmsimd_arm_x16, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsubc_minmax_ukernel__wasmsimd_arm_x16, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vrsubc_minmax_ukernel__wasmsimd_arm_x16, |
| .linear.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsub_ukernel__wasmsimd_x16, |
| .linear.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsubc_ukernel__wasmsimd_x16, |
| .linear.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vrsubc_ukernel__wasmsimd_x16, |
| .element_tile = 16, |
| }; |
| } |
| xnn_params.f32.vsqrdiff = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsqrdiff_ukernel__wasmsimd_x16, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsqrdiffc_ukernel__wasmsimd_x16, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsqrdiffc_ukernel__wasmsimd_x16, |
| .element_tile = 16, |
| }; |
| if (is_wasm_x86) { |
| xnn_params.f32.vmulcaddc = (struct vmulcaddc_parameters) { |
| .ukernel = (xnn_vmulcaddc_ukernel_function) xnn_f32_vmulcaddc_minmax_ukernel_c4__wasmsimd_x86_2x, |
| .init.f32 = xnn_init_f32_minmax_scalar_params, |
| .channel_tile = 4, |
| .row_tile = 2, |
| }; |
| } else { |
| xnn_params.f32.vmulcaddc = (struct vmulcaddc_parameters) { |
| .ukernel = (xnn_vmulcaddc_ukernel_function) xnn_f32_vmulcaddc_minmax_ukernel_c4__wasmsimd_arm_2x, |
| .init.f32 = xnn_init_f32_minmax_scalar_params, |
| .channel_tile = 4, |
| .row_tile = 2, |
| }; |
| } |
| #ifndef XNN_NO_NCHW_OPERATORS |
| init_flags |= XNN_INIT_FLAG_CHW_OPT; |
| |
| if (is_wasm_x86) { |
| xnn_params.f32.spmm = (struct spmm_parameters) { |
| .ukernel = (xnn_spmm_ukernel_function) xnn_f32_spmm_minmax_ukernel_32x1__wasmsimd_x86, |
| .mr = 32, |
| .nr = 1, |
| }; |
| } else { |
| xnn_params.f32.spmm = (struct spmm_parameters) { |
| .ukernel = (xnn_spmm_ukernel_function) xnn_f32_spmm_minmax_ukernel_32x1__wasmsimd_arm, |
| .mr = 32, |
| .nr = 1, |
| }; |
| } |
| xnn_params.f32.conv_hwc2chw_3x3c3s2 = (struct conv_hwc2chw_parameters) { |
| .ukernel_with_symm_padding = |
| (xnn_conv_hwc2chw_ukernel_function) xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__wasmsimd_2x2, |
| .output_channel_tile = 4, |
| .output_height_tile = 2, |
| .output_width_tile = 2, |
| }; |
| if (is_wasm_x86) { |
| xnn_params.f32.dwconv2d_chw_3x3 = (struct dwconv2d_chw_parameters) { |
| .ukernel = (xnn_dwconv2d_chw_ukernel_function) xnn_f32_dwconv2d_chw_ukernel_3x3p1__wasmsimd_x86_loadsplat_2x4, |
| .output_width_tile = 4, |
| .output_height_tile = 2, |
| }; |
| xnn_params.f32.dwconv2d_chw_3x3s2 = (struct dwconv2d_chw_parameters) { |
| .ukernel = (xnn_dwconv2d_chw_ukernel_function) xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_x86_splat_1x4_acc2, |
| .output_width_tile = 4, |
| .output_height_tile = 1, |
| }; |
| xnn_params.f32.dwconv2d_chw_5x5 = (struct dwconv2d_chw_parameters) { |
| .ukernel = (xnn_dwconv2d_chw_ukernel_function) xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_3x4, |
| .output_width_tile = 4, |
| .output_height_tile = 3, |
| }; |
| xnn_params.f32.dwconv2d_chw_5x5s2 = (struct dwconv2d_chw_parameters) { |
| .ukernel = (xnn_dwconv2d_chw_ukernel_function) xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_splat_1x4_acc2, |
| .output_width_tile = 4, |
| .output_height_tile = 1, |
| }; |
| } else { |
| xnn_params.f32.dwconv2d_chw_3x3 = (struct dwconv2d_chw_parameters) { |
| .ukernel = (xnn_dwconv2d_chw_ukernel_function) xnn_f32_dwconv2d_chw_ukernel_3x3p1__wasmsimd_arm_loadsplat_2x4, |
| .output_width_tile = 4, |
| .output_height_tile = 2, |
| }; |
| xnn_params.f32.dwconv2d_chw_3x3s2 = (struct dwconv2d_chw_parameters) { |
| .ukernel = (xnn_dwconv2d_chw_ukernel_function) xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__wasmsimd_arm_splat_1x4_acc4, |
| .output_width_tile = 4, |
| .output_height_tile = 1, |
| }; |
| xnn_params.f32.dwconv2d_chw_5x5 = (struct dwconv2d_chw_parameters) { |
| .ukernel = (xnn_dwconv2d_chw_ukernel_function) xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_arm_splat_3x4, |
| .output_width_tile = 4, |
| .output_height_tile = 3, |
| }; |
| xnn_params.f32.dwconv2d_chw_5x5s2 = (struct dwconv2d_chw_parameters) { |
| .ukernel = (xnn_dwconv2d_chw_ukernel_function) xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_splat_1x4_acc2, |
| .output_width_tile = 4, |
| .output_height_tile = 1, |
| }; |
| } |
| if (is_wasm_x86) { |
| xnn_params.f32.gavgpool_cw = (struct gavgpool_cw_parameters) { |
| .ukernel = (xnn_gavgpool_cw_ukernel_function) xnn_f32_gavgpool_cw_ukernel__wasmsimd_x86_x4, |
| .channel_tile = 4, |
| }; |
| } else { |
| xnn_params.f32.gavgpool_cw = (struct gavgpool_cw_parameters) { |
| .ukernel = (xnn_gavgpool_cw_ukernel_function) xnn_f32_gavgpool_cw_ukernel__wasmsimd_arm_x4, |
| .channel_tile = 4, |
| }; |
| } |
| xnn_params.f32.ibilinear_chw = (struct ibilinear_chw_parameters) { |
| .ukernel = (xnn_ibilinear_chw_ukernel_function) xnn_f32_ibilinear_chw_ukernel__wasmsimd_p8, |
| .channel_tile = 1, |
| .pixel_tile = 8, |
| }; |
| #endif // XNN_NO_NCHW_OPERATORS |
| #endif // XNN_NO_F32_OPERATORS |
| |
| /*************************** VCVT micro-kernels ***************************/ |
| #ifndef XNN_NO_VCVT_OPERATORS |
| init_flags |= XNN_INIT_FLAG_VCVT; |
| |
| xnn_params.vcvt.f16_to_f32 = (xnn_univector_ukernel_function) xnn_f16_f32_vcvt_ukernel__wasmsimd_int16_x16; |
| #endif // XNN_NO_VCVT_OPERATORS |
| |
| /**************************** X32 micro-kernels ****************************/ |
| #ifndef XNN_NO_X32_OPERATORS |
| init_flags |= XNN_INIT_FLAG_X32; |
| |
| xnn_params.x32.unpool = (xnn_unpool_ukernel_function) xnn_x32_unpool_ukernel__wasmsimd; |
| xnn_params.x32.zip = (struct zip_parameters) { |
| .x2 = (xnn_zipc_ukernel_function) xnn_x32_zip_x2_ukernel__wasmsimd, |
| .x3 = (xnn_zipc_ukernel_function) xnn_x32_zip_x3_ukernel__wasmsimd, |
| .x4 = (xnn_zipc_ukernel_function) xnn_x32_zip_x4_ukernel__wasmsimd, |
| .xm = (xnn_zipv_ukernel_function) xnn_x32_zip_xm_ukernel__wasmsimd, |
| }; |
| #ifndef XNN_NO_NCHW_OPERATORS |
| xnn_params.x32.depthtospace2d_chw2hwc = (struct depthtospace2d_chw2hwc_parameters) { |
| .ukernel = (xnn_depthtospace2d_chw2hwc_ukernel_function) xnn_x32_depthtospace2d_chw2hwc_ukernel__scalar, |
| .channel_tile = 1, |
| .pixel_tile = 1, |
| }; |
| #endif // XNN_NO_NCHW_OPERATORS |
| #endif // XNN_NO_X32_OPERATORS |
| |
| /**************************** XX micro-kernels ****************************/ |
| #ifndef XNN_NO_XX_OPERATORS |
| init_flags |= XNN_INIT_FLAG_XX; |
| |
| xnn_params.xx.copy = (xnn_univector_ukernel_function) xnn_xx_copy_ukernel__memcpy; |
| xnn_params.xx.fill = (struct fill_parameters) { |
| .ukernel = (xnn_fill_ukernel_function) xnn_xx_fill_ukernel__wasmsimd_x64, |
| .row_tile = 1, |
| }; |
| xnn_params.xx.pad = (struct pad_parameters) { |
| .ukernel = (xnn_pad_ukernel_function) xnn_xx_pad_ukernel__wasmsimd, |
| .row_tile = 1, |
| }; |
| #endif |
| |
| #elif XNN_ARCH_WASM |
| |
| /**************************** QC8 micro-kernels ****************************/ |
| #ifndef XNN_NO_QC8_OPERATORS |
| init_flags |= XNN_INIT_FLAG_QC8; |
| |
| if (is_wasm_x86) { |
| xnn_params.qc8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_2x2__scalar_magic); |
| xnn_params.qc8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_2x2__scalar_magic); |
| xnn_params.qc8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_1x2__scalar_magic); |
| xnn_params.qc8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_1x2__scalar_magic); |
| xnn_params.qc8.gemm.init.qc8 = xnn_init_qs8_minmax_scalar_magic_params; |
| xnn_params.qc8.gemm.mr = 2; |
| xnn_params.qc8.gemm.nr = 2; |
| } else { |
| xnn_params.qc8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_4x4__scalar_magic); |
| xnn_params.qc8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_4x4__scalar_magic); |
| xnn_params.qc8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qc8_gemm_minmax_fp32_ukernel_1x4__scalar_magic); |
| xnn_params.qc8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qc8_igemm_minmax_fp32_ukernel_1x4__scalar_magic); |
| xnn_params.qc8.gemm.init.qc8 = xnn_init_qs8_minmax_scalar_magic_params; |
| xnn_params.qc8.gemm.mr = 4; |
| xnn_params.qc8.gemm.nr = 4; |
| } |
| |
| xnn_params.qc8.dwconv[0].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qc8_dwconv_minmax_fp32_ukernel_up2x9__scalar_magic; |
| xnn_params.qc8.dwconv[0].init.qc8 = xnn_init_qs8_minmax_scalar_magic_params; |
| xnn_params.qc8.dwconv[0].channel_tile = 2; |
| xnn_params.qc8.dwconv[0].primary_tile = 9; |
| xnn_params.qc8.dwconv[1].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qc8_dwconv_minmax_fp32_ukernel_up2x25__scalar_magic; |
| xnn_params.qc8.dwconv[1].init.qc8 = xnn_init_qs8_minmax_scalar_magic_params; |
| xnn_params.qc8.dwconv[1].channel_tile = 2; |
| xnn_params.qc8.dwconv[1].primary_tile = 25; |
| #endif // XNN_NO_QC8_OPERATORS |
| |
| /**************************** QS8 micro-kernels ****************************/ |
| #ifndef XNN_NO_QS8_OPERATORS |
| init_flags |= XNN_INIT_FLAG_QS8; |
| |
| if (is_wasm_x86) { |
| xnn_params.qs8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_fp32_ukernel_2x2__scalar_magic); |
| xnn_params.qs8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_fp32_ukernel_2x2__scalar_magic); |
| xnn_params.qs8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_fp32_ukernel_1x2__scalar_magic); |
| xnn_params.qs8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_fp32_ukernel_1x2__scalar_magic); |
| xnn_params.qs8.gemm.init.qs8 = xnn_init_qs8_conv_minmax_fp32_scalar_magic_params; |
| xnn_params.qs8.gemm.mr = 2; |
| xnn_params.qs8.gemm.nr = 2; |
| } else { |
| xnn_params.qs8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_fp32_ukernel_4x4__scalar_magic); |
| xnn_params.qs8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_fp32_ukernel_4x4__scalar_magic); |
| xnn_params.qs8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_fp32_ukernel_1x4__scalar_magic); |
| xnn_params.qs8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_fp32_ukernel_1x4__scalar_magic); |
| xnn_params.qs8.gemm.init.qs8 = xnn_init_qs8_conv_minmax_fp32_scalar_magic_params; |
| xnn_params.qs8.gemm.mr = 4; |
| xnn_params.qs8.gemm.nr = 4; |
| } |
| |
| xnn_params.qs8.dwconv[0].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qs8_dwconv_minmax_fp32_ukernel_up2x9__scalar_magic; |
| xnn_params.qs8.dwconv[0].init.qs8 = xnn_init_qs8_conv_minmax_fp32_scalar_magic_params; |
| xnn_params.qs8.dwconv[0].channel_tile = 2; |
| xnn_params.qs8.dwconv[0].primary_tile = 9; |
| xnn_params.qs8.dwconv[1].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qs8_dwconv_minmax_fp32_ukernel_up2x25__scalar_magic; |
| xnn_params.qs8.dwconv[1].init.qs8 = xnn_init_qs8_conv_minmax_fp32_scalar_magic_params; |
| xnn_params.qs8.dwconv[1].channel_tile = 2; |
| xnn_params.qs8.dwconv[1].primary_tile = 25; |
| |
| xnn_params.qs8.gavgpool = (struct gavgpool_parameters) { |
| .up = (xnn_gavgpool_unipass_ukernel_function) xnn_qs8_gavgpool_minmax_ukernel_7x__scalar_c4, |
| .mp = (xnn_gavgpool_multipass_ukernel_function) xnn_qs8_gavgpool_minmax_ukernel_7p7x__scalar_c4, |
| .mr = 7, |
| }; |
| |
| xnn_params.qs8.vadd = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vadd_minmax_ukernel__scalar_x4, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vaddc_minmax_ukernel__scalar_x4, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vaddc_minmax_ukernel__scalar_x4, |
| .init.qs8_addsub = xnn_init_qs8_add_minmax_scalar_params, |
| .element_tile = 4, |
| }; |
| xnn_params.qs8.vmul = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vmul_minmax_fp32_ukernel__scalar_x4, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vmulc_minmax_fp32_ukernel__scalar_x4, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vmulc_minmax_fp32_ukernel__scalar_x4, |
| .init.qs8_mul = xnn_init_qs8_mul_minmax_fp32_scalar_params, |
| .element_tile = 4, |
| }; |
| #endif // XNN_NO_QS8_OPERATORS |
| |
| /**************************** QU8 micro-kernels ****************************/ |
| #ifndef XNN_NO_QU8_OPERATORS |
| init_flags |= XNN_INIT_FLAG_QU8; |
| |
| if (is_wasm_x86) { |
| xnn_params.qu8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qu8_gemm_minmax_fp32_ukernel_2x2__scalar_magic); |
| xnn_params.qu8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qu8_igemm_minmax_fp32_ukernel_2x2__scalar_magic); |
| xnn_params.qu8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qu8_gemm_minmax_fp32_ukernel_1x2__scalar_magic); |
| xnn_params.qu8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qu8_igemm_minmax_fp32_ukernel_1x2__scalar_magic); |
| xnn_params.qu8.gemm.init.qu8 = xnn_init_qu8_conv_minmax_fp32_scalar_magic_params; |
| xnn_params.qu8.gemm.mr = 2; |
| xnn_params.qu8.gemm.nr = 2; |
| } else { |
| xnn_params.qu8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qu8_gemm_minmax_fp32_ukernel_4x4__scalar_magic); |
| xnn_params.qu8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qu8_igemm_minmax_fp32_ukernel_4x4__scalar_magic); |
| xnn_params.qu8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qu8_gemm_minmax_fp32_ukernel_1x4__scalar_magic); |
| xnn_params.qu8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qu8_igemm_minmax_fp32_ukernel_1x4__scalar_magic); |
| xnn_params.qu8.gemm.init.qu8 = xnn_init_qu8_conv_minmax_fp32_scalar_magic_params; |
| xnn_params.qu8.gemm.mr = 4; |
| xnn_params.qu8.gemm.nr = 4; |
| } |
| |
| xnn_params.qu8.dwconv[0].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qu8_dwconv_minmax_fp32_ukernel_up2x9__scalar_magic; |
| xnn_params.qu8.dwconv[0].init.qu8 = xnn_init_qu8_conv_minmax_fp32_scalar_magic_params; |
| xnn_params.qu8.dwconv[0].channel_tile = 2; |
| xnn_params.qu8.dwconv[0].primary_tile = 9; |
| xnn_params.qu8.dwconv[1].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qu8_dwconv_minmax_fp32_ukernel_up2x25__scalar_magic; |
| xnn_params.qu8.dwconv[1].init.qu8 = xnn_init_qu8_conv_minmax_fp32_scalar_magic_params; |
| xnn_params.qu8.dwconv[1].channel_tile = 2; |
| xnn_params.qu8.dwconv[1].primary_tile = 25; |
| |
| xnn_params.qu8.avgpool = (struct avgpool_parameters) { |
| .up = (xnn_avgpool_unipass_ukernel_function) xnn_qu8_avgpool_minmax_ukernel_9x__scalar_c1, |
| .mp = (xnn_avgpool_multipass_ukernel_function) xnn_qu8_avgpool_minmax_ukernel_9p8x__scalar_c1, |
| .mr = 9, |
| .qr = 8, |
| }; |
| xnn_params.qu8.gavgpool = (struct gavgpool_parameters) { |
| .up = (xnn_gavgpool_unipass_ukernel_function) xnn_qu8_gavgpool_minmax_ukernel_7x__scalar_c1, |
| .mp = (xnn_gavgpool_multipass_ukernel_function) xnn_qu8_gavgpool_minmax_ukernel_7p7x__scalar_c1, |
| .mr = 7, |
| }; |
| |
| xnn_params.qu8.vadd = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vadd_minmax_ukernel__scalar_x4, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vaddc_minmax_ukernel__scalar_x4, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vaddc_minmax_ukernel__scalar_x4, |
| .init.qu8_addsub = xnn_init_qu8_add_minmax_scalar_params, |
| .element_tile = 4, |
| }; |
| xnn_params.qu8.vmul = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vmul_minmax_fp32_ukernel__scalar_x4, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vmulc_minmax_fp32_ukernel__scalar_x4, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vmulc_minmax_fp32_ukernel__scalar_x4, |
| .init.qu8_mul = xnn_init_qu8_mul_minmax_fp32_scalar_params, |
| .element_tile = 4, |
| }; |
| #endif // XNN_NO_QU8_OPERATORS |
| |
| /**************************** S8 micro-kernels ****************************/ |
| #ifndef XNN_NO_S8_OPERATORS |
| init_flags |= XNN_INIT_FLAG_S8; |
| |
| xnn_params.s8.clamp = (struct vunary_parameters) { |
| .ukernel = (xnn_univector_ukernel_function) xnn_s8_vclamp_ukernel__scalar_x4, |
| .init.s8_minmax = xnn_init_s8_minmax_scalar_params, |
| .element_tile = 4, |
| }; |
| xnn_params.s8.maxpool = (struct maxpool_parameters) { |
| .ukernel = (xnn_maxpool_ukernel_function) xnn_s8_maxpool_minmax_ukernel_9p8x__scalar_c1, |
| .init.s8 = xnn_init_s8_minmax_scalar_params, |
| .mr = 9, |
| .qr = 8, |
| }; |
| #endif // XNN_NO_S8_OPERATORS |
| |
| /**************************** U8 micro-kernels ****************************/ |
| #ifndef XNN_NO_U8_OPERATORS |
| init_flags |= XNN_INIT_FLAG_U8; |
| |
| xnn_params.u8.clamp = (struct vunary_parameters) { |
| .ukernel = (xnn_univector_ukernel_function) xnn_u8_vclamp_ukernel__scalar_x4, |
| .init.u8_minmax = xnn_init_u8_minmax_scalar_params, |
| .element_tile = 4, |
| }; |
| xnn_params.u8.maxpool = (struct maxpool_parameters) { |
| .ukernel = (xnn_maxpool_ukernel_function) xnn_u8_maxpool_minmax_ukernel_9p8x__scalar_c1, |
| .init.u8 = xnn_init_u8_minmax_scalar_params, |
| .mr = 9, |
| .qr = 8, |
| }; |
| xnn_params.u8.lut32norm = xnn_u8_lut32norm_ukernel__scalar; |
| xnn_params.u8.rmax = xnn_u8_rmax_ukernel__scalar; |
| #endif // XNN_NO_U8_OPERATORS |
| |
| /**************************** X8 micro-kernels ****************************/ |
| #ifndef XNN_NO_X8_OPERATORS |
| init_flags |= XNN_INIT_FLAG_X8; |
| |
| xnn_params.x8.lut = xnn_x8_lut_ukernel__scalar_x4; |
| xnn_params.x8.zip = (struct zip_parameters) { |
| .x2 = (xnn_zipc_ukernel_function) xnn_x8_zip_x2_ukernel__scalar, |
| .x3 = (xnn_zipc_ukernel_function) xnn_x8_zip_x3_ukernel__scalar, |
| .x4 = (xnn_zipc_ukernel_function) xnn_x8_zip_x4_ukernel__scalar, |
| .xm = (xnn_zipv_ukernel_function) xnn_x8_zip_xm_ukernel__scalar, |
| }; |
| #endif // XNN_NO_X8_OPERATORS |
| |
| /**************************** F32 micro-kernels ****************************/ |
| #ifndef XNN_NO_F32_OPERATORS |
| init_flags |= XNN_INIT_FLAG_F32; |
| |
| if (is_wasm_x86) { |
| xnn_params.f32.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_2x4__scalar); |
| xnn_params.f32.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_2x4__scalar); |
| xnn_params.f32.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_1x4__wasm); |
| xnn_params.f32.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_1x4__wasm); |
| xnn_params.f32.gemm.relu.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_relu_ukernel_2x4__scalar); |
| xnn_params.f32.gemm.relu.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_relu_ukernel_2x4__scalar); |
| xnn_params.f32.gemm.relu.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_relu_ukernel_1x4__wasm); |
| xnn_params.f32.gemm.relu.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_relu_ukernel_1x4__wasm); |
| xnn_params.f32.gemm.linear.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_ukernel_2x4__scalar); |
| xnn_params.f32.gemm.linear.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_ukernel_2x4__scalar); |
| xnn_params.f32.gemm.linear.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_ukernel_1x4__wasm); |
| xnn_params.f32.gemm.linear.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_ukernel_1x4__wasm); |
| xnn_params.f32.gemm.init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.gemm.mr = 2; |
| xnn_params.f32.gemm.nr = 4; |
| } else { |
| xnn_params.f32.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_4x4__wasm); |
| xnn_params.f32.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_4x4__wasm); |
| xnn_params.f32.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_1x4__wasm); |
| xnn_params.f32.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_1x4__wasm); |
| xnn_params.f32.gemm.relu.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_relu_ukernel_4x4__wasm); |
| xnn_params.f32.gemm.relu.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_relu_ukernel_4x4__wasm); |
| xnn_params.f32.gemm.relu.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_relu_ukernel_1x4__wasm); |
| xnn_params.f32.gemm.relu.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_relu_ukernel_1x4__wasm); |
| xnn_params.f32.gemm.linear.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_ukernel_4x4__wasm); |
| xnn_params.f32.gemm.linear.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_ukernel_4x4__wasm); |
| xnn_params.f32.gemm.linear.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_ukernel_1x4__wasm); |
| xnn_params.f32.gemm.linear.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_ukernel_1x4__wasm); |
| xnn_params.f32.gemm.init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.gemm.mr = 4; |
| xnn_params.f32.gemm.nr = 4; |
| } |
| xnn_params.f32.gemm2.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_4x2__wasm); |
| xnn_params.f32.gemm2.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_4x2__wasm), |
| xnn_params.f32.gemm2.linear.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_ukernel_4x2__wasm); |
| xnn_params.f32.gemm2.linear.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_ukernel_4x2__wasm), |
| xnn_params.f32.gemm2.init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.gemm2.mr = 4; |
| xnn_params.f32.gemm2.nr = 2; |
| |
| xnn_params.f32.dwconv[0].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_minmax_ukernel_up1x3__wasm_acc2; |
| xnn_params.f32.dwconv[0].linear.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_ukernel_up1x3__wasm_acc2; |
| xnn_params.f32.dwconv[0].init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.dwconv[0].channel_tile = 1; |
| xnn_params.f32.dwconv[0].primary_tile = 3; |
| |
| xnn_params.f32.dwconv[1].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_minmax_ukernel_up1x4__wasm_acc2; |
| xnn_params.f32.dwconv[1].linear.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_ukernel_up1x4__wasm_acc2; |
| xnn_params.f32.dwconv[1].init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.dwconv[1].channel_tile = 1; |
| xnn_params.f32.dwconv[1].primary_tile = 4; |
| |
| xnn_params.f32.dwconv[2].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_minmax_ukernel_up1x9__wasm_acc2; |
| xnn_params.f32.dwconv[2].linear.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_ukernel_up1x9__wasm_acc2; |
| xnn_params.f32.dwconv[2].init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.dwconv[2].channel_tile = 1; |
| xnn_params.f32.dwconv[2].primary_tile = 9; |
| |
| xnn_params.f32.dwconv[3].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_minmax_ukernel_up1x25__wasm_acc2; |
| xnn_params.f32.dwconv[3].linear.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_ukernel_up1x25__wasm_acc2; |
| xnn_params.f32.dwconv[3].init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.dwconv[3].channel_tile = 1; |
| xnn_params.f32.dwconv[3].primary_tile = 25; |
| |
| xnn_params.f32.avgpool = (struct avgpool_parameters) { |
| .up = (xnn_avgpool_unipass_ukernel_function) xnn_f32_avgpool_minmax_ukernel_9x__wasm_c1, |
| .mp = (xnn_avgpool_multipass_ukernel_function) xnn_f32_avgpool_minmax_ukernel_9p8x__wasm_c1, |
| .mr = 9, |
| .qr = 8, |
| }; |
| xnn_params.f32.pavgpool = (struct pavgpool_parameters) { |
| .up = (xnn_pavgpool_unipass_ukernel_function) xnn_f32_pavgpool_minmax_ukernel_9x__wasm_c1, |
| .mp = (xnn_pavgpool_multipass_ukernel_function) xnn_f32_pavgpool_minmax_ukernel_9p8x__wasm_c1, |
| .mr = 9, |
| .qr = 8, |
| }; |
| xnn_params.f32.gavgpool = (struct gavgpool_parameters) { |
| .up = (xnn_gavgpool_unipass_ukernel_function) xnn_f32_gavgpool_minmax_ukernel_7x__wasm_c1, |
| .mp = (xnn_gavgpool_multipass_ukernel_function) xnn_f32_gavgpool_minmax_ukernel_7p7x__wasm_c1, |
| .mr = 7, |
| }; |
| xnn_params.f32.maxpool = (struct maxpool_parameters) { |
| .ukernel = (xnn_maxpool_ukernel_function) xnn_f32_maxpool_minmax_ukernel_9p8x__wasm_c1, |
| .init.f32 = xnn_init_f32_minmax_scalar_params, |
| .mr = 9, |
| .qr = 8, |
| }; |
| xnn_params.f32.argmaxpool[0] = (struct argmaxpool_parameters) { |
| .up = (xnn_argmaxpool_unipass_ukernel_function) xnn_f32_argmaxpool_ukernel_4x__scalar_c1, |
| .mr = 4, |
| }; |
| xnn_params.f32.argmaxpool[1] = (struct argmaxpool_parameters) { |
| .up = (xnn_argmaxpool_unipass_ukernel_function) xnn_f32_argmaxpool_ukernel_9x__scalar_c1, |
| .mr = 9, |
| }; |
| xnn_params.f32.argmaxpool[2] = (struct argmaxpool_parameters) { |
| .mp = (xnn_argmaxpool_multipass_ukernel_function) xnn_f32_argmaxpool_ukernel_9p8x__scalar_c1, |
| .mr = 9, |
| .qr = 8, |
| }; |
| xnn_params.f32.ibilinear = (struct ibilinear_parameters) { |
| .ukernel = (xnn_ibilinear_ukernel_function) xnn_f32_ibilinear_ukernel__scalar_c2, |
| .pixel_tile = 1, |
| .channel_tile = 2, |
| }; |
| xnn_params.f32.abs = (xnn_univector_ukernel_function) xnn_f32_vabs_ukernel__scalar_x4; |
| xnn_params.f32.clamp = (struct vunary_parameters) { |
| .ukernel = (xnn_univector_ukernel_function) xnn_f32_vclamp_ukernel__wasm_x4, |
| .init.f32_minmax = xnn_init_f32_minmax_scalar_params, |
| .element_tile = 4, |
| }; |
| if (is_wasm_x86) { |
| xnn_params.f32.hswish = (xnn_univector_ukernel_function) xnn_f32_vhswish_ukernel__scalar_x4; |
| } else { |
| xnn_params.f32.hswish = (xnn_univector_ukernel_function) xnn_f32_vhswish_ukernel__wasm_x4; |
| } |
| if (is_wasm_x86) { |
| xnn_params.f32.elu = (xnn_univector_ukernel_function) xnn_f32_velu_ukernel__scalar_rr2_lut16_p3_x2; |
| } else { |
| xnn_params.f32.elu = (xnn_univector_ukernel_function) xnn_f32_velu_ukernel__wasm_rr2_p6_x6; |
| } |
| xnn_params.f32.lrelu = (xnn_univector_ukernel_function) xnn_f32_vlrelu_ukernel__scalar_x4; |
| xnn_params.f32.neg = (xnn_univector_ukernel_function) xnn_f32_vneg_ukernel__scalar_x4; |
| if (is_wasm_x86) { |
| xnn_params.f32.relu = (xnn_univector_ukernel_function) xnn_f32_vrelu_ukernel__scalar_x8; |
| } else { |
| xnn_params.f32.relu = (xnn_univector_ukernel_function) xnn_f32_vrelu_ukernel__wasm_x8; |
| } |
| xnn_params.f32.rndne = (xnn_univector_ukernel_function) xnn_f32_vrndne_ukernel__scalar_libm_x4; |
| xnn_params.f32.rndz = (xnn_univector_ukernel_function) xnn_f32_vrndz_ukernel__scalar_libm_x4; |
| xnn_params.f32.rndu = (xnn_univector_ukernel_function) xnn_f32_vrndu_ukernel__scalar_libm_x4; |
| xnn_params.f32.rndd = (xnn_univector_ukernel_function) xnn_f32_vrndd_ukernel__scalar_libm_x4; |
| xnn_params.f32.sigmoid = (xnn_univector_ukernel_function) xnn_f32_vsigmoid_ukernel__scalar_lut64_p2_div_x2; |
| xnn_params.f32.sqr = (xnn_univector_ukernel_function) xnn_f32_vsqr_ukernel__scalar_x4; |
| xnn_params.f32.sqrt = (xnn_univector_ukernel_function) xnn_f32_vsqrt_ukernel__scalar_sqrt_x1; |
| if (is_wasm_x86) { |
| xnn_params.f32.prelu = (struct prelu_parameters) { |
| .ukernel = (xnn_prelu_ukernel_function) xnn_f32_prelu_ukernel__scalar_2x4, |
| .row_tile = 2, |
| .channel_tile = 4, |
| }; |
| } else { |
| xnn_params.f32.prelu = (struct prelu_parameters) { |
| .ukernel = (xnn_prelu_ukernel_function) xnn_f32_prelu_ukernel__wasm_2x4, |
| .row_tile = 2, |
| .channel_tile = 4, |
| }; |
| } |
| xnn_params.f32.raddstoreexpminusmax = xnn_f32_raddstoreexpminusmax_ukernel__scalar_p5_x4_acc2; |
| xnn_params.f32.rmax = xnn_f32_rmax_ukernel__scalar; |
| xnn_params.f32.vadd = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vadd_minmax_ukernel__wasm_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vaddc_minmax_ukernel__wasm_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vaddc_minmax_ukernel__wasm_x8, |
| .element_tile = 8, |
| }; |
| xnn_params.f32.vdiv = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vdiv_minmax_ukernel__wasm_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vdivc_minmax_ukernel__wasm_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vrdivc_minmax_ukernel__wasm_x8, |
| .element_tile = 8, |
| }; |
| xnn_params.f32.vmax = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmax_ukernel__wasm_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmaxc_ukernel__wasm_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmaxc_ukernel__wasm_x8, |
| .element_tile = 8, |
| }; |
| xnn_params.f32.vmin = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmin_ukernel__wasm_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vminc_ukernel__wasm_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vminc_ukernel__wasm_x8, |
| .element_tile = 8, |
| }; |
| xnn_params.f32.vmul = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmul_minmax_ukernel__wasm_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmulc_minmax_ukernel__wasm_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmulc_minmax_ukernel__wasm_x8, |
| .element_tile = 8, |
| }; |
| xnn_params.f32.vsub = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsub_minmax_ukernel__wasm_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsubc_minmax_ukernel__wasm_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vrsubc_minmax_ukernel__wasm_x8, |
| .element_tile = 8, |
| }; |
| xnn_params.f32.vsqrdiff = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsqrdiff_ukernel__scalar_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsqrdiffc_ukernel__scalar_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsqrdiffc_ukernel__scalar_x8, |
| .element_tile = 8, |
| }; |
| xnn_params.f32.vmulcaddc = (struct vmulcaddc_parameters) { |
| .ukernel = (xnn_vmulcaddc_ukernel_function) xnn_f32_vmulcaddc_minmax_ukernel_c1__wasm_2x, |
| .init.f32 = xnn_init_f32_minmax_scalar_params, |
| .channel_tile = 1, |
| .row_tile = 2, |
| }; |
| #ifndef XNN_NO_NCHW_OPERATORS |
| init_flags |= XNN_INIT_FLAG_CHW_OPT; |
| |
| xnn_params.f32.spmm = (struct spmm_parameters) { |
| .ukernel = (xnn_spmm_ukernel_function) xnn_f32_spmm_minmax_ukernel_8x1__scalar, |
| .mr = 8, |
| .nr = 1, |
| }; |
| xnn_params.f32.spmm2 = (struct spmm_parameters) { |
| .ukernel = (xnn_spmm_ukernel_function) xnn_f32_spmm_minmax_ukernel_8x2__scalar, |
| .mr = 8, |
| .nr = 2, |
| }; |
| xnn_params.f32.spmm4 = (struct spmm_parameters) { |
| .ukernel = (xnn_spmm_ukernel_function) xnn_f32_spmm_minmax_ukernel_8x4__scalar, |
| .mr = 8, |
| .nr = 4, |
| }; |
| xnn_params.f32.conv_hwc2chw_3x3c3s2 = (struct conv_hwc2chw_parameters) { |
| .ukernel_with_symm_padding = |
| (xnn_conv_hwc2chw_ukernel_function) xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__scalar_1x1, |
| .output_channel_tile = 4, |
| .output_height_tile = 1, |
| .output_width_tile = 1, |
| }; |
| xnn_params.f32.dwconv2d_chw_3x3 = (struct dwconv2d_chw_parameters) { |
| .ukernel = (xnn_dwconv2d_chw_ukernel_function) xnn_f32_dwconv2d_chw_ukernel_3x3p1__scalar_2x1_acc2, |
| .output_width_tile = 1, |
| .output_height_tile = 2, |
| }; |
| xnn_params.f32.dwconv2d_chw_3x3s2 = (struct dwconv2d_chw_parameters) { |
| .ukernel = (xnn_dwconv2d_chw_ukernel_function) xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__scalar_1x1_acc2, |
| .output_width_tile = 1, |
| .output_height_tile = 1, |
| }; |
| xnn_params.f32.dwconv2d_chw_5x5 = (struct dwconv2d_chw_parameters) { |
| .ukernel = (xnn_dwconv2d_chw_ukernel_function) xnn_f32_dwconv2d_chw_ukernel_5x5p2__scalar_1x1_acc5, |
| .output_width_tile = 1, |
| .output_height_tile = 1, |
| }; |
| xnn_params.f32.dwconv2d_chw_5x5s2 = (struct dwconv2d_chw_parameters) { |
| .ukernel = (xnn_dwconv2d_chw_ukernel_function) xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__scalar_1x1_acc5, |
| .output_width_tile = 1, |
| .output_height_tile = 1, |
| }; |
| xnn_params.f32.gavgpool_cw = (struct gavgpool_cw_parameters) { |
| .ukernel = (xnn_gavgpool_cw_ukernel_function) xnn_f32_gavgpool_cw_ukernel__scalar_x1, |
| .channel_tile = 1, |
| }; |
| xnn_params.f32.ibilinear_chw = (struct ibilinear_chw_parameters) { |
| .ukernel = (xnn_ibilinear_chw_ukernel_function) xnn_f32_ibilinear_chw_ukernel__scalar_p4, |
| .channel_tile = 1, |
| .pixel_tile = 4, |
| }; |
| #endif // XNN_NO_NCHW_OPERATORS |
| #endif // XNN_NO_F32_OPERATORS |
| |
| /*************************** VCVT micro-kernels ***************************/ |
| #ifndef XNN_NO_VCVT_OPERATORS |
| init_flags |= XNN_INIT_FLAG_VCVT; |
| |
| xnn_params.vcvt.f16_to_f32 = (xnn_univector_ukernel_function) xnn_f16_f32_vcvt_ukernel__scalar_float_x1; |
| #endif // XNN_NO_VCVT_OPERATORS |
| |
| /**************************** X32 micro-kernels ****************************/ |
| #ifndef XNN_NO_X32_OPERATORS |
| init_flags |= XNN_INIT_FLAG_X32; |
| |
| xnn_params.x32.unpool = (xnn_unpool_ukernel_function) xnn_x32_unpool_ukernel__scalar; |
| xnn_params.x32.zip = (struct zip_parameters) { |
| .x2 = (xnn_zipc_ukernel_function) xnn_x32_zip_x2_ukernel__scalar, |
| .x3 = (xnn_zipc_ukernel_function) xnn_x32_zip_x3_ukernel__scalar, |
| .x4 = (xnn_zipc_ukernel_function) xnn_x32_zip_x4_ukernel__scalar, |
| .xm = (xnn_zipv_ukernel_function) xnn_x32_zip_xm_ukernel__scalar, |
| }; |
| #ifndef XNN_NO_NCHW_OPERATORS |
| xnn_params.x32.depthtospace2d_chw2hwc = (struct depthtospace2d_chw2hwc_parameters) { |
| .ukernel = (xnn_depthtospace2d_chw2hwc_ukernel_function) xnn_x32_depthtospace2d_chw2hwc_ukernel__scalar, |
| .channel_tile = 1, |
| .pixel_tile = 1, |
| }; |
| #endif // XNN_NO_NCHW_OPERATORS |
| #endif // XNN_NO_X32_OPERATORS |
| |
| /**************************** XX micro-kernels ****************************/ |
| #ifndef XNN_NO_XX_OPERATORS |
| init_flags |= XNN_INIT_FLAG_XX; |
| |
| xnn_params.xx.copy = (xnn_univector_ukernel_function) xnn_xx_copy_ukernel__memcpy; |
| xnn_params.xx.fill = (struct fill_parameters) { |
| .ukernel = (xnn_fill_ukernel_function) xnn_xx_fill_ukernel__scalar_x16, |
| .row_tile = 1, |
| }; |
| xnn_params.xx.pad = (struct pad_parameters) { |
| .ukernel = (xnn_pad_ukernel_function) xnn_xx_pad_ukernel__scalar, |
| .row_tile = 1, |
| }; |
| #endif |
| |
| #elif XNN_ARCH_RISCV |
| |
| /**************************** QS8 micro-kernels ****************************/ |
| #ifndef XNN_NO_QS8_OPERATORS |
| init_flags |= XNN_INIT_FLAG_QS8; |
| |
| xnn_params.qs8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_rndnu_ukernel_3x4__scalar); |
| xnn_params.qs8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_rndnu_ukernel_3x4__scalar); |
| xnn_params.qs8.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qs8_gemm_minmax_rndnu_ukernel_1x4__scalar); |
| xnn_params.qs8.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qs8_igemm_minmax_rndnu_ukernel_1x4__scalar); |
| xnn_params.qs8.gemm.init.qs8 = xnn_init_qs8_conv_minmax_rndnu_scalar_params; |
| xnn_params.qs8.gemm.mr = 3; |
| xnn_params.qs8.gemm.nr = 4; |
| |
| xnn_params.qs8.dwconv[0].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qs8_dwconv_minmax_gemmlowp_ukernel_up2x9__scalar; |
| xnn_params.qs8.dwconv[0].init.qs8 = xnn_init_qs8_conv_minmax_gemmlowp_scalar_params; |
| xnn_params.qs8.dwconv[0].channel_tile = 2; |
| xnn_params.qs8.dwconv[0].primary_tile = 9; |
| xnn_params.qs8.dwconv[1].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qs8_dwconv_minmax_gemmlowp_ukernel_up2x25__scalar; |
| xnn_params.qs8.dwconv[1].init.qs8 = xnn_init_qs8_conv_minmax_gemmlowp_scalar_params; |
| xnn_params.qs8.dwconv[1].channel_tile = 2; |
| xnn_params.qs8.dwconv[1].primary_tile = 25; |
| |
| xnn_params.qs8.gavgpool = (struct gavgpool_parameters) { |
| .up = (xnn_gavgpool_unipass_ukernel_function) xnn_qs8_gavgpool_minmax_ukernel_7x__scalar_c1, |
| .mp = (xnn_gavgpool_multipass_ukernel_function) xnn_qs8_gavgpool_minmax_ukernel_7p7x__scalar_c1, |
| .mr = 7, |
| }; |
| |
| xnn_params.qs8.vadd = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vadd_minmax_ukernel__scalar_x4, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vaddc_minmax_ukernel__scalar_x4, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_qs8_vaddc_minmax_ukernel__scalar_x4, |
| .init.qs8_addsub = xnn_init_qs8_add_minmax_scalar_params, |
| .element_tile = 4, |
| }; |
| #endif // XNN_NO_QS8_OPERATORS |
| |
| /**************************** QU8 micro-kernels ****************************/ |
| #ifndef XNN_NO_QU8_OPERATORS |
| init_flags |= XNN_INIT_FLAG_QU8; |
| |
| xnn_params.qu8.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_qu8_gemm_minmax_gemmlowp_ukernel_2x2__scalar); |
| xnn_params.qu8.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_qu8_igemm_minmax_gemmlowp_ukernel_2x2__scalar); |
| xnn_params.qu8.gemm.init.qu8 = xnn_init_qu8_conv_minmax_gemmlowp_scalar_params; |
| xnn_params.qu8.gemm.mr = 2; |
| xnn_params.qu8.gemm.nr = 2; |
| |
| xnn_params.qu8.dwconv[0].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_qu8_dwconv_minmax_gemmlowp_ukernel_up1x9__scalar; |
| xnn_params.qu8.dwconv[0].init.qu8 = xnn_init_qu8_conv_minmax_gemmlowp_scalar_params; |
| xnn_params.qu8.dwconv[0].channel_tile = 1; |
| xnn_params.qu8.dwconv[0].primary_tile = 9; |
| |
| xnn_params.qu8.avgpool = (struct avgpool_parameters) { |
| .up = (xnn_avgpool_unipass_ukernel_function) xnn_qu8_avgpool_minmax_ukernel_9x__scalar_c1, |
| .mp = (xnn_avgpool_multipass_ukernel_function) xnn_qu8_avgpool_minmax_ukernel_9p8x__scalar_c1, |
| .mr = 9, |
| .qr = 8, |
| }; |
| xnn_params.qu8.gavgpool = (struct gavgpool_parameters) { |
| .up = (xnn_gavgpool_unipass_ukernel_function) xnn_qu8_gavgpool_minmax_ukernel_7x__scalar_c1, |
| .mp = (xnn_gavgpool_multipass_ukernel_function) xnn_qu8_gavgpool_minmax_ukernel_7p7x__scalar_c1, |
| .mr = 7, |
| }; |
| |
| xnn_params.qu8.vadd = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vadd_minmax_ukernel__scalar_x4, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vaddc_minmax_ukernel__scalar_x4, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_qu8_vaddc_minmax_ukernel__scalar_x4, |
| .init.qu8_addsub = xnn_init_qu8_add_minmax_scalar_params, |
| .element_tile = 4, |
| }; |
| #endif // XNN_NO_QU8_OPERATORS |
| |
| /**************************** U8 micro-kernels ****************************/ |
| #ifndef XNN_NO_U8_OPERATORS |
| init_flags |= XNN_INIT_FLAG_U8; |
| |
| xnn_params.u8.clamp = (struct vunary_parameters) { |
| .ukernel = (xnn_univector_ukernel_function) xnn_u8_vclamp_ukernel__scalar_x4, |
| .init.u8_minmax = xnn_init_u8_minmax_scalar_params, |
| .element_tile = 4, |
| }; |
| xnn_params.u8.maxpool = (struct maxpool_parameters) { |
| .ukernel = (xnn_maxpool_ukernel_function) xnn_u8_maxpool_minmax_ukernel_9p8x__scalar_c1, |
| .init.u8 = xnn_init_u8_minmax_scalar_params, |
| .mr = 9, |
| .qr = 8, |
| }; |
| xnn_params.u8.lut32norm = xnn_u8_lut32norm_ukernel__scalar; |
| xnn_params.u8.rmax = xnn_u8_rmax_ukernel__scalar; |
| #endif // XNN_NO_U8_OPERATORS |
| |
| /**************************** X8 micro-kernels ****************************/ |
| #ifndef XNN_NO_X8_OPERATORS |
| init_flags |= XNN_INIT_FLAG_X8; |
| |
| xnn_params.x8.lut = xnn_x8_lut_ukernel__scalar_x4; |
| xnn_params.x8.zip = (struct zip_parameters) { |
| .x2 = (xnn_zipc_ukernel_function) xnn_x8_zip_x2_ukernel__scalar, |
| .x3 = (xnn_zipc_ukernel_function) xnn_x8_zip_x3_ukernel__scalar, |
| .x4 = (xnn_zipc_ukernel_function) xnn_x8_zip_x4_ukernel__scalar, |
| .xm = (xnn_zipv_ukernel_function) xnn_x8_zip_xm_ukernel__scalar, |
| }; |
| #endif // XNN_NO_X8_OPERATORS |
| |
| /**************************** F32 micro-kernels ****************************/ |
| #ifndef XNN_NO_F32_OPERATORS |
| init_flags |= XNN_INIT_FLAG_F32; |
| |
| xnn_params.f32.gemm.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_4x4__scalar); |
| xnn_params.f32.gemm.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_4x4__scalar); |
| xnn_params.f32.gemm.minmax.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_1x4__scalar); |
| xnn_params.f32.gemm.minmax.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_1x4__scalar); |
| xnn_params.f32.gemm.relu.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_relu_ukernel_4x4__scalar); |
| xnn_params.f32.gemm.relu.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_relu_ukernel_4x4__scalar); |
| xnn_params.f32.gemm.relu.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_relu_ukernel_1x4__scalar); |
| xnn_params.f32.gemm.relu.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_relu_ukernel_1x4__scalar); |
| xnn_params.f32.gemm.linear.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_ukernel_4x4__scalar); |
| xnn_params.f32.gemm.linear.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_ukernel_4x4__scalar); |
| xnn_params.f32.gemm.linear.gemm1 = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_ukernel_1x4__scalar); |
| xnn_params.f32.gemm.linear.igemm1 = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_ukernel_1x4__scalar); |
| xnn_params.f32.gemm.init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.gemm.mr = 4; |
| xnn_params.f32.gemm.nr = 4; |
| |
| xnn_params.f32.gemm2.minmax.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_minmax_ukernel_4x2__scalar); |
| xnn_params.f32.gemm2.minmax.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_minmax_ukernel_4x2__scalar), |
| xnn_params.f32.gemm2.linear.gemm = xnn_init_hmp_gemm_ukernel((xnn_gemm_ukernel_function) xnn_f32_gemm_ukernel_4x2__scalar); |
| xnn_params.f32.gemm2.linear.igemm = xnn_init_hmp_igemm_ukernel((xnn_igemm_ukernel_function) xnn_f32_igemm_ukernel_4x2__scalar), |
| xnn_params.f32.gemm2.init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.gemm2.mr = 4; |
| xnn_params.f32.gemm2.nr = 2; |
| |
| xnn_params.f32.dwconv[0].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_minmax_ukernel_up1x3__scalar_acc2; |
| xnn_params.f32.dwconv[0].linear.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_ukernel_up1x3__scalar_acc2; |
| xnn_params.f32.dwconv[0].init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.dwconv[0].channel_tile = 1; |
| xnn_params.f32.dwconv[0].primary_tile = 3; |
| |
| xnn_params.f32.dwconv[1].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_minmax_ukernel_up1x4__scalar_acc2; |
| xnn_params.f32.dwconv[1].linear.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_ukernel_up1x4__scalar_acc2; |
| xnn_params.f32.dwconv[1].init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.dwconv[1].channel_tile = 1; |
| xnn_params.f32.dwconv[1].primary_tile = 4; |
| |
| xnn_params.f32.dwconv[2].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_minmax_ukernel_up1x9__scalar_acc2; |
| xnn_params.f32.dwconv[2].linear.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_ukernel_up1x9__scalar_acc2; |
| xnn_params.f32.dwconv[2].init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.dwconv[2].channel_tile = 1; |
| xnn_params.f32.dwconv[2].primary_tile = 9; |
| |
| xnn_params.f32.dwconv[3].minmax.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_minmax_ukernel_up1x25__scalar_acc2; |
| xnn_params.f32.dwconv[3].linear.unipass = (xnn_dwconv_unipass_ukernel_function) xnn_f32_dwconv_ukernel_up1x25__scalar_acc2; |
| xnn_params.f32.dwconv[3].init.f32 = xnn_init_f32_minmax_scalar_params; |
| xnn_params.f32.dwconv[3].channel_tile = 1; |
| xnn_params.f32.dwconv[3].primary_tile = 25; |
| |
| xnn_params.f32.avgpool = (struct avgpool_parameters) { |
| .up = (xnn_avgpool_unipass_ukernel_function) xnn_f32_avgpool_minmax_ukernel_9x__scalar_c1, |
| .mp = (xnn_avgpool_multipass_ukernel_function) xnn_f32_avgpool_minmax_ukernel_9p8x__scalar_c1, |
| .mr = 9, |
| .qr = 8, |
| }; |
| xnn_params.f32.pavgpool = (struct pavgpool_parameters) { |
| .up = (xnn_pavgpool_unipass_ukernel_function) xnn_f32_pavgpool_minmax_ukernel_9x__scalar_c1, |
| .mp = (xnn_pavgpool_multipass_ukernel_function) xnn_f32_pavgpool_minmax_ukernel_9p8x__scalar_c1, |
| .mr = 9, |
| .qr = 8, |
| }; |
| xnn_params.f32.gavgpool = (struct gavgpool_parameters) { |
| .up = (xnn_gavgpool_unipass_ukernel_function) xnn_f32_gavgpool_minmax_ukernel_7x__scalar_c1, |
| .mp = (xnn_gavgpool_multipass_ukernel_function) xnn_f32_gavgpool_minmax_ukernel_7p7x__scalar_c1, |
| .mr = 7, |
| }; |
| xnn_params.f32.maxpool = (struct maxpool_parameters) { |
| .ukernel = (xnn_maxpool_ukernel_function) xnn_f32_maxpool_minmax_ukernel_9p8x__scalar_c1, |
| .init.f32 = xnn_init_f32_minmax_scalar_params, |
| .mr = 9, |
| .qr = 8, |
| }; |
| xnn_params.f32.argmaxpool[0] = (struct argmaxpool_parameters) { |
| .up = (xnn_argmaxpool_unipass_ukernel_function) xnn_f32_argmaxpool_ukernel_4x__scalar_c1, |
| .mr = 4, |
| }; |
| xnn_params.f32.argmaxpool[1] = (struct argmaxpool_parameters) { |
| .up = (xnn_argmaxpool_unipass_ukernel_function) xnn_f32_argmaxpool_ukernel_9x__scalar_c1, |
| .mr = 9, |
| }; |
| xnn_params.f32.argmaxpool[2] = (struct argmaxpool_parameters) { |
| .mp = (xnn_argmaxpool_multipass_ukernel_function) xnn_f32_argmaxpool_ukernel_9p8x__scalar_c1, |
| .mr = 9, |
| .qr = 8, |
| }; |
| xnn_params.f32.ibilinear = (struct ibilinear_parameters) { |
| .ukernel = (xnn_ibilinear_ukernel_function) xnn_f32_ibilinear_ukernel__scalar_c2, |
| .pixel_tile = 1, |
| .channel_tile = 2, |
| }; |
| xnn_params.f32.abs = (xnn_univector_ukernel_function) xnn_f32_vabs_ukernel__scalar_x4; |
| xnn_params.f32.clamp = (xnn_univector_ukernel_function) xnn_f32_vclamp_ukernel__scalar_x4; |
| xnn_params.f32.hswish = (xnn_univector_ukernel_function) xnn_f32_vhswish_ukernel__scalar_x4; |
| xnn_params.f32.elu = (xnn_univector_ukernel_function) xnn_f32_velu_ukernel__scalar_rr2_lut16_p3_x2; |
| xnn_params.f32.lrelu = (xnn_univector_ukernel_function) xnn_f32_vlrelu_ukernel__scalar_x4; |
| xnn_params.f32.neg = (xnn_univector_ukernel_function) xnn_f32_vneg_ukernel__scalar_x4; |
| xnn_params.f32.relu = (xnn_univector_ukernel_function) xnn_f32_vrelu_ukernel__scalar_x8; |
| xnn_params.f32.rndne = (xnn_univector_ukernel_function) xnn_f32_vrndne_ukernel__scalar_libm_x4; |
| xnn_params.f32.rndz = (xnn_univector_ukernel_function) xnn_f32_vrndz_ukernel__scalar_libm_x4; |
| xnn_params.f32.rndu = (xnn_univector_ukernel_function) xnn_f32_vrndu_ukernel__scalar_libm_x4; |
| xnn_params.f32.rndd = (xnn_univector_ukernel_function) xnn_f32_vrndd_ukernel__scalar_libm_x4; |
| xnn_params.f32.sigmoid = (xnn_univector_ukernel_function) xnn_f32_vsigmoid_ukernel__scalar_lut64_p2_div_x2; |
| xnn_params.f32.sqr = (xnn_univector_ukernel_function) xnn_f32_vsqr_ukernel__scalar_x4; |
| xnn_params.f32.sqrt = (xnn_univector_ukernel_function) xnn_f32_vsqrt_ukernel__scalar_sqrt_x1; |
| xnn_params.f32.prelu = (struct prelu_parameters) { |
| .ukernel = (xnn_prelu_ukernel_function) xnn_f32_prelu_ukernel__scalar_2x4, |
| .row_tile = 2, |
| .channel_tile = 4, |
| }; |
| xnn_params.f32.raddstoreexpminusmax = xnn_f32_raddstoreexpminusmax_ukernel__scalar_p5_x4_acc2; |
| xnn_params.f32.rmax = xnn_f32_rmax_ukernel__scalar; |
| xnn_params.f32.vadd = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vadd_minmax_ukernel__scalar_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vaddc_minmax_ukernel__scalar_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vaddc_minmax_ukernel__scalar_x8, |
| .element_tile = 8, |
| }; |
| xnn_params.f32.vdiv = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vdiv_minmax_ukernel__scalar_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vdivc_minmax_ukernel__scalar_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vrdivc_minmax_ukernel__scalar_x8, |
| .element_tile = 8, |
| }; |
| xnn_params.f32.vmax = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmax_ukernel__scalar_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmaxc_ukernel__scalar_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmaxc_ukernel__scalar_x8, |
| .element_tile = 8, |
| }; |
| xnn_params.f32.vmin = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmin_ukernel__scalar_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vminc_ukernel__scalar_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vminc_ukernel__scalar_x8, |
| .element_tile = 8, |
| }; |
| xnn_params.f32.vmul = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmul_minmax_ukernel__scalar_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmulc_minmax_ukernel__scalar_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vmulc_minmax_ukernel__scalar_x8, |
| .element_tile = 8, |
| }; |
| xnn_params.f32.vsub = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsub_minmax_ukernel__scalar_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsubc_minmax_ukernel__scalar_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vrsubc_minmax_ukernel__scalar_x8, |
| .element_tile = 8, |
| }; |
| xnn_params.f32.vsqrdiff = (struct vbinary_parameters) { |
| .minmax.op_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsqrdiff_ukernel__scalar_x8, |
| .minmax.opc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsqrdiffc_ukernel__scalar_x8, |
| .minmax.ropc_ukernel = (xnn_vbinary_ukernel_function) xnn_f32_vsqrdiffc_ukernel__scalar_x8, |
| .element_tile = 8, |
| }; |
| xnn_params.f32.vmulcaddc = (struct vmulcaddc_parameters) { |
| .ukernel = (xnn_vmulcaddc_ukernel_function) xnn_f32_vmulcaddc_minmax_ukernel_c1__scalar_2x, |
| .init.f32 = xnn_init_f32_minmax_scalar_params, |
| .channel_tile = 1, |
| .row_tile = 2, |
| }; |
| #ifndef XNN_NO_NCHW_OPERATORS |
| init_flags |= XNN_INIT_FLAG_CHW_OPT; |
| |
| xnn_params.f32.spmm = (struct spmm_parameters) { |
| .ukernel = (xnn_spmm_ukernel_function) xnn_f32_spmm_minmax_ukernel_8x1__scalar, |
| .mr = 8, |
| .nr = 1, |
| }; |
| xnn_params.f32.spmm2 = (struct spmm_parameters) { |
| .ukernel = (xnn_spmm_ukernel_function) xnn_f32_spmm_minmax_ukernel_8x2__scalar, |
| .mr = 8, |
| .nr = 2, |
| }; |
| xnn_params.f32.spmm4 = (struct spmm_parameters) { |
| .ukernel = (xnn_spmm_ukernel_function) xnn_f32_spmm_minmax_ukernel_8x4__scalar, |
| .mr = 8, |
| .nr = 4, |
| }; |
| xnn_params.f32.conv_hwc2chw_3x3c3s2 = (struct conv_hwc2chw_parameters) { |
| .ukernel_with_symm_padding = |
| (xnn_conv_hwc2chw_ukernel_function) xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__scalar_1x1, |
| .output_channel_tile = 4, |
| .output_height_tile = 1, |
| .output_width_tile = 1, |
| }; |
| xnn_params.f32.dwconv2d_chw_3x3 = (struct dwconv2d_chw_parameters) { |
| .ukernel = (xnn_dwconv2d_chw_ukernel_function) xnn_f32_dwconv2d_chw_ukernel_3x3p1__scalar_2x1_acc2, |
| .output_width_tile = 1, |
| .output_height_tile = 2, |
| }; |
| xnn_params.f32.dwconv2d_chw_3x3s2 = (struct dwconv2d_chw_parameters) { |
| .ukernel = (xnn_dwconv2d_chw_ukernel_function) xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__scalar_1x1_acc2, |
| .output_width_tile = 1, |
| .output_height_tile = 1, |
| }; |
| xnn_params.f32.dwconv2d_chw_5x5 = (struct dwconv2d_chw_parameters) { |
| .ukernel = (xnn_dwconv2d_chw_ukernel_function) xnn_f32_dwconv2d_chw_ukernel_5x5p2__scalar_1x1_acc5, |
| .output_width_tile = 1, |
| .output_height_tile = 1, |
| }; |
| xnn_params.f32.dwconv2d_chw_5x5s2 = (struct dwconv2d_chw_parameters) { |
| .ukernel = (xnn_dwconv2d_chw_ukernel_function) xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__scalar_1x1_acc5, |
| .output_width_tile = 1, |
| .output_height_tile = 1, |
| }; |
| xnn_params.f32.gavgpool_cw = (struct gavgpool_cw_parameters) { |
| .ukernel = (xnn_gavgpool_cw_ukernel_function) xnn_f32_gavgpool_cw_ukernel__scalar_x1, |
| .channel_tile = 1, |
| }; |
| xnn_params.f32.ibilinear_chw = (struct ibilinear_chw_parameters) { |
| .ukernel = (xnn_ibilinear_chw_ukernel_function) xnn_f32_ibilinear_chw_ukernel__scalar_p4, |
| .channel_tile = 1, |
| .pixel_tile = 4, |
| }; |
| #endif // XNN_NO_NCHW_OPERATORS |
| #endif // XNN_NO_F32_OPERATORS |
| |
| /**************************** X32 micro-kernels ****************************/ |
| #ifndef XNN_NO_X32_OPERATORS |
| init_flags |= XNN_INIT_FLAG_X32; |
| |
| xnn_params.x32.unpool = (xnn_unpool_ukernel_function) xnn_x32_unpool_ukernel__scalar; |
| xnn_params.x32.zip = (struct zip_parameters) { |
| .x2 = (xnn_zipc_ukernel_function) xnn_x32_zip_x2_ukernel__scalar, |
| .x3 = (xnn_zipc_ukernel_function) xnn_x32_zip_x3_ukernel__scalar, |
| .x4 = (xnn_zipc_ukernel_function) xnn_x32_zip_x4_ukernel__scalar, |
| .xm = (xnn_zipv_ukernel_function) xnn_x32_zip_xm_ukernel__scalar, |
| }; |
| #ifndef XNN_NO_NCHW_OPERATORS |
| xnn_params.x32.depthtospace2d_chw2hwc = (struct depthtospace2d_chw2hwc_parameters) { |
| .ukernel = (xnn_depthtospace2d_chw2hwc_ukernel_function) xnn_x32_depthtospace2d_chw2hwc_ukernel__scalar, |
| .channel_tile = 1, |
| .pixel_tile = 1, |
| }; |
| #endif // XNN_NO_NCHW_OPERATORS |
| #endif // XNN_NO_X32_OPERATORS |
| |
| /**************************** XX micro-kernels ****************************/ |
| #ifndef XNN_NO_XX_OPERATORS |
| init_flags |= XNN_INIT_FLAG_XX; |
| |
| xnn_params.xx.copy = (xnn_univector_ukernel_function) xnn_xx_copy_ukernel__memcpy; |
| xnn_params.xx.fill = (struct fill_parameters) { |
| .ukernel = (xnn_fill_ukernel_function) xnn_xx_fill_ukernel__scalar_x16, |
| .row_tile = 1, |
| }; |
| xnn_params.xx.pad = (struct pad_parameters) { |
| .ukernel = (xnn_pad_ukernel_function) xnn_xx_pad_ukernel__scalar, |
| .row_tile = 1, |
| }; |
| #endif |
| |
| #else |
| #error "Unsupported architecture" |
| #endif |
| |
| memcpy(&xnn_params.allocator, init_allocator, sizeof(struct xnn_allocator)); |
| xnn_params.init_flags = init_flags; |
| } |
| |
| #ifdef _WIN32 |
| static BOOL CALLBACK init_windows(PINIT_ONCE init_once, PVOID parameter, PVOID* context) { |
| init(); |
| return TRUE; |
| } |
| #endif |
| |
| enum xnn_status xnn_initialize(const struct xnn_allocator* allocator) { |
| #ifndef __EMSCRIPTEN__ |
| if (!cpuinfo_initialize()) { |
| return xnn_status_out_of_memory; |
| } |
| #endif |
| if (allocator == NULL) { |
| allocator = &xnn_default_allocator; |
| } |
| #ifdef _MSC_VER |
| _InterlockedCompareExchangePointer((PVOID volatile*) &init_allocator, (PVOID) allocator, NULL); |
| #else |
| __sync_bool_compare_and_swap(&init_allocator, NULL, allocator); |
| #endif |
| #ifdef _WIN32 |
| InitOnceExecuteOnce(&init_guard, &init_windows, NULL, NULL); |
| #else |
| pthread_once(&init_guard, &init); |
| #endif |
| if ((xnn_params.init_flags & XNN_INIT_FLAG_XNNPACK) != 0) { |
| return xnn_status_success; |
| } else { |
| return xnn_status_unsupported_hardware; |
| } |
| } |
| |
| enum xnn_status xnn_deinitialize(void) { |
| #ifndef __EMSCRIPTEN__ |
| cpuinfo_deinitialize(); |
| #endif |
| return xnn_status_success; |
| } |