/* * Copyright (c) 2021, 2023-2024 Arm Limited. * * SPDX-License-Identifier: MIT * * Permission is hereby granted, free of charge, to any person obtaining a copy * of this software and associated documentation files (the "Software"), to * deal in the Software without restriction, including without limitation the * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or * sell copies of the Software, and to permit persons to whom the Software is * furnished to do so, subject to the following conditions: * * The above copyright notice and this permission notice shall be included in all * copies or substantial portions of the Software. * * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE * SOFTWARE. */ #pragma once #if defined(__aarch64__) namespace { void a64_transpose_interleave_16_1x4(uint8_t *out, const uint8_t *in, size_t width, size_t in_stride, size_t height) { uint8_t *pad_row = reinterpret_cast(alloca(width * sizeof(uint8_t))); if (height % 4) { memset(pad_row, 0, width * sizeof(uint8_t)); } size_t out_stride = 16 * roundup(height, 4) * sizeof(uint8_t); __asm__ __volatile__( "cmp %x[height], #0x10\n" "blt 9f\n" "1:" // Main row loop: Head "mov x17, %x[in]\n" "mov x16, %x[width]\n" "mov x15, %x[out]\n" "sub %x[height], %x[height], #0x10\n" "add x14, x17, %x[in_stride]\n" "add x13, x14, %x[in_stride]\n" "add x12, x13, %x[in_stride]\n" "cmp x16, #0x10\n" "add x11, x12, %x[in_stride]\n" "add x10, x11, %x[in_stride]\n" "add x9, x10, %x[in_stride]\n" "add x28, x9, %x[in_stride]\n" "add x27, x28, %x[in_stride]\n" "add x26, x27, %x[in_stride]\n" "add x25, x26, %x[in_stride]\n" "add x24, x25, %x[in_stride]\n" "add x23, x24, %x[in_stride]\n" "add x22, x23, %x[in_stride]\n" "add x21, x22, %x[in_stride]\n" "add x20, x21, %x[in_stride]\n" "add %x[in], x20, %x[in_stride]\n" "blt 3f\n" "2:" // Main row loop: Column loop "ldr q19, [x17], #0x10\n" "ldr q18, [x14], #0x10\n" "sub x16, x16, #0x10\n" "ldr q17, [x13], #0x10\n" "ldr q16, [x12], #0x10\n" "cmp x16, #0x10\n" "ldr q24, [x11], #0x10\n" "ldr q23, [x10], #0x10\n" "ldr q22, [x9], #0x10\n" "ldr q21, [x28], #0x10\n" "ldr q30, [x27], #0x10\n" "ldr q29, [x26], #0x10\n" "zip1 v3.16b, v19.16b, v17.16b\n" "zip1 v2.16b, v18.16b, v16.16b\n" "ldr q28, [x25], #0x10\n" "ldr q20, [x24], #0x10\n" "zip2 v1.16b, v19.16b, v17.16b\n" "zip2 v27.16b, v18.16b, v16.16b\n" "ldr q19, [x23], #0x10\n" "ldr q18, [x22], #0x10\n" "zip1 v26.16b, v24.16b, v22.16b\n" "zip1 v25.16b, v23.16b, v21.16b\n" "ldr q17, [x21], #0x10\n" "ldr q16, [x20], #0x10\n" "zip2 v24.16b, v24.16b, v22.16b\n" "zip2 v23.16b, v23.16b, v21.16b\n" "zip1 v22.16b, v30.16b, v28.16b\n" "zip1 v21.16b, v29.16b, v20.16b\n" "zip2 v0.16b, v30.16b, v28.16b\n" "zip2 v20.16b, v29.16b, v20.16b\n" "zip1 v31.16b, v19.16b, v17.16b\n" "zip1 v30.16b, v18.16b, v16.16b\n" "zip2 v29.16b, v19.16b, v17.16b\n" "zip2 v28.16b, v18.16b, v16.16b\n" "zip1 v19.16b, v3.16b, v2.16b\n" "zip2 v18.16b, v3.16b, v2.16b\n" "zip1 v17.16b, v1.16b, v27.16b\n" "zip2 v16.16b, v1.16b, v27.16b\n" "zip1 v27.16b, v26.16b, v25.16b\n" "zip2 v26.16b, v26.16b, v25.16b\n" "zip1 v25.16b, v24.16b, v23.16b\n" "zip2 v24.16b, v24.16b, v23.16b\n" "str q19, [x15, #0x0]\n" "zip1 v23.16b, v22.16b, v21.16b\n" "zip2 v22.16b, v22.16b, v21.16b\n" "str q18, [x15, #0x10]\n" "zip1 v21.16b, v0.16b, v20.16b\n" "zip2 v20.16b, v0.16b, v20.16b\n" "str q17, [x15, #0x20]\n" "zip1 v19.16b, v31.16b, v30.16b\n" "zip2 v18.16b, v31.16b, v30.16b\n" "str q16, [x15, #0x30]\n" "zip1 v17.16b, v29.16b, v28.16b\n" "zip2 v16.16b, v29.16b, v28.16b\n" "str q27, [x15, #0x40]\n" "str q26, [x15, #0x50]\n" "str q25, [x15, #0x60]\n" "str q24, [x15, #0x70]\n" "str q23, [x15, #0x80]\n" "str q22, [x15, #0x90]\n" "str q21, [x15, #0xa0]\n" "str q20, [x15, #0xb0]\n" "str q19, [x15, #0xc0]\n" "str q18, [x15, #0xd0]\n" "str q17, [x15, #0xe0]\n" "str q16, [x15, #0xf0]\n" "add x15, x15, %x[out_stride]\n" "bge 2b\n" "3:" // Main row loop: Column loop skip "cbz x16, 8f\n" "cmp x16, #0x4\n" "movi v16.16b, #0x0\n" "str q16, [x15, #0x0]\n" "str q16, [x15, #0x10]\n" "str q16, [x15, #0x20]\n" "str q16, [x15, #0x30]\n" "str q16, [x15, #0x40]\n" "str q16, [x15, #0x50]\n" "str q16, [x15, #0x60]\n" "str q16, [x15, #0x70]\n" "str q16, [x15, #0x80]\n" "str q16, [x15, #0x90]\n" "str q16, [x15, #0xa0]\n" "str q16, [x15, #0xb0]\n" "str q16, [x15, #0xc0]\n" "str q16, [x15, #0xd0]\n" "str q16, [x15, #0xe0]\n" "str q16, [x15, #0xf0]\n" "blt 5f\n" "4:" // Main row loop: width 4 loop: loop "ldr s23, [x17], #0x4\n" "ldr s21, [x14], #0x4\n" "sub x16, x16, #0x4\n" "ldr s20, [x13], #0x4\n" "ldr s19, [x12], #0x4\n" "cmp x16, #0x4\n" "ldr s22, [x11], #0x4\n" "ldr s18, [x10], #0x4\n" "ldr s17, [x9], #0x4\n" "ldr s16, [x28], #0x4\n" "ldr s27, [x27], #0x4\n" "ldr s26, [x26], #0x4\n" "zip1 v25.16b, v23.16b, v20.16b\n" "zip1 v21.16b, v21.16b, v19.16b\n" "ldr s20, [x25], #0x4\n" "ldr s19, [x24], #0x4\n" "ldr s24, [x23], #0x4\n" "ldr s23, [x22], #0x4\n" "zip1 v22.16b, v22.16b, v17.16b\n" "zip1 v17.16b, v18.16b, v16.16b\n" "ldr s18, [x21], #0x4\n" "ldr s16, [x20], #0x4\n" "zip1 v21.16b, v25.16b, v21.16b\n" "zip1 v20.16b, v27.16b, v20.16b\n" "zip1 v19.16b, v26.16b, v19.16b\n" "zip1 v17.16b, v22.16b, v17.16b\n" "zip1 v18.16b, v24.16b, v18.16b\n" "zip1 v16.16b, v23.16b, v16.16b\n" "str q21, [x15, #0x0]\n" "str q17, [x15, #0x40]\n" "zip1 v17.16b, v20.16b, v19.16b\n" "zip1 v16.16b, v18.16b, v16.16b\n" "str q17, [x15, #0x80]\n" "str q16, [x15, #0xc0]\n" "add x15, x15, #0x10\n" "bge 4b\n" "5:" // Main row loop: width 4 loop: skip "cmp x16, #0x1\n" "blt 7f\n" "6:" // Main row loop: width 1 loop: loop "ldr b23, [x17], #0x1\n" "ldr b21, [x14], #0x1\n" "sub x16, x16, #0x1\n" "ldr b20, [x13], #0x1\n" "ldr b19, [x12], #0x1\n" "cmp x16, #0x1\n" "ldr b22, [x11], #0x1\n" "ldr b18, [x10], #0x1\n" "ldr b17, [x9], #0x1\n" "ldr b16, [x28], #0x1\n" "ldr b27, [x27], #0x1\n" "ldr b26, [x26], #0x1\n" "zip1 v25.16b, v23.16b, v20.16b\n" "zip1 v21.16b, v21.16b, v19.16b\n" "ldr b20, [x25], #0x1\n" "ldr b19, [x24], #0x1\n" "ldr b24, [x23], #0x1\n" "ldr b23, [x22], #0x1\n" "zip1 v22.16b, v22.16b, v17.16b\n" "zip1 v17.16b, v18.16b, v16.16b\n" "ldr b18, [x21], #0x1\n" "ldr b16, [x20], #0x1\n" "zip1 v21.16b, v25.16b, v21.16b\n" "zip1 v20.16b, v27.16b, v20.16b\n" "zip1 v19.16b, v26.16b, v19.16b\n" "zip1 v17.16b, v22.16b, v17.16b\n" "zip1 v18.16b, v24.16b, v18.16b\n" "zip1 v16.16b, v23.16b, v16.16b\n" "str s21, [x15, #0x0]\n" "str s17, [x15, #0x40]\n" "zip1 v17.16b, v20.16b, v19.16b\n" "zip1 v16.16b, v18.16b, v16.16b\n" "str s17, [x15, #0x80]\n" "str s16, [x15, #0xc0]\n" "add x15, x15, #0x4\n" "bge 6b\n" "7:" // Main row loop: width 1 loop: skip "8:" // Main row loop: odd col skip "cmp %x[height], #0x10\n" "add %x[out], %x[out], #0x100\n" "bge 1b\n" "cbz %x[height], 18f\n" "9:" // Main loop skip "10:" // Tail row loop: Head "mov x17, %x[in]\n" "mov x20, %x[width]\n" "cmp %x[height], #0x3\n" "mov x15, %x[out]\n" "add x14, x17, %x[in_stride]\n" "add x13, x14, %x[in_stride]\n" "add x12, x13, %x[in_stride]\n" "csel x13, x13, %x[pad_row], GE\n" "add %x[in], x12, %x[in_stride]\n" "csel x12, x12, %x[pad_row], GT\n" "cmp %x[height], #0x1\n" "sub %x[height], %x[height], #0x4\n" "csel x14, x14, %x[pad_row], GT\n" "cmp x20, #0x10\n" "blt 12f\n" "11:" // Tail row loop: Column loop "ldr q20, [x17], #0x10\n" "ldr q21, [x14], #0x10\n" "sub x20, x20, #0x10\n" "ldr q19, [x13], #0x10\n" "ldr q16, [x12], #0x10\n" "cmp x20, #0x10\n" "zip1 v18.16b, v20.16b, v19.16b\n" "zip1 v17.16b, v21.16b, v16.16b\n" "zip2 v20.16b, v20.16b, v19.16b\n" "zip2 v16.16b, v21.16b, v16.16b\n" "zip1 v19.16b, v18.16b, v17.16b\n" "zip2 v18.16b, v18.16b, v17.16b\n" "zip1 v17.16b, v20.16b, v16.16b\n" "zip2 v16.16b, v20.16b, v16.16b\n" "str q19, [x15, #0x0]\n" "str q18, [x15, #0x10]\n" "str q17, [x15, #0x20]\n" "str q16, [x15, #0x30]\n" "add x15, x15, %x[out_stride]\n" "bge 11b\n" "12:" // Tail row loop: Column loop skip "cbz x20, 17f\n" "cmp x20, #0x4\n" "movi v16.16b, #0x0\n" "str q16, [x15, #0x0]\n" "str q16, [x15, #0x10]\n" "str q16, [x15, #0x20]\n" "str q16, [x15, #0x30]\n" "blt 14f\n" "13:" // Tail row loop: width 4 loop: loop "ldr s19, [x17], #0x4\n" "ldr s18, [x14], #0x4\n" "sub x20, x20, #0x4\n" "ldr s17, [x13], #0x4\n" "ldr s16, [x12], #0x4\n" "cmp x20, #0x4\n" "zip1 v17.16b, v19.16b, v17.16b\n" "zip1 v16.16b, v18.16b, v16.16b\n" "zip1 v16.16b, v17.16b, v16.16b\n" "str q16, [x15, #0x0]\n" "add x15, x15, #0x10\n" "bge 13b\n" "14:" // Tail row loop: width 4 loop: skip "cmp x20, #0x1\n" "blt 16f\n" "15:" // Tail row loop: width 1 loop: loop "ldr b19, [x17], #0x1\n" "ldr b18, [x14], #0x1\n" "sub x20, x20, #0x1\n" "ldr b17, [x13], #0x1\n" "ldr b16, [x12], #0x1\n" "cmp x20, #0x1\n" "zip1 v17.16b, v19.16b, v17.16b\n" "zip1 v16.16b, v18.16b, v16.16b\n" "zip1 v16.16b, v17.16b, v16.16b\n" "str s16, [x15, #0x0]\n" "add x15, x15, #0x4\n" "bge 15b\n" "16:" // Tail row loop: width 1 loop: skip "17:" // Tail row loop: odd col skip "cmp %x[height], #0x1\n" "add %x[out], %x[out], #0x40\n" "bge 10b\n" "18:" // Done : [height] "+&r" (height), [in] "+&r" (in), [out] "+&r" (out) : [in_stride] "r" (in_stride), [out_stride] "r" (out_stride), [pad_row] "r" (pad_row), [width] "r" (width) : "cc", "memory", "v0", "v1", "v2", "v3", "v16", "v17", "v18", "v19", "v20", "v21", "v22", "v23", "v24", "v25", "v26", "v27", "v28", "v29", "v30", "v31", "x9", "x10", "x11", "x12", "x13", "x14", "x15", "x16", "x17", "x20", "x21", "x22", "x23", "x24", "x25", "x26", "x27", "x28" ); } } // anonymous namespace template<> void Transform<16, 4, true, VLType::None>( uint8_t *out, const uint8_t *in, int stride, int x0, int xmax, int k0, int kmax) { a64_transpose_interleave_16_1x4( reinterpret_cast(out), reinterpret_cast(in + k0 * stride + x0), (xmax-x0) * sizeof(uint8_t) / 1, stride * sizeof(uint8_t), (kmax-k0) ); } template<> void Transform<16, 4, true, VLType::None>( int8_t *out, const int8_t *in, int stride, int x0, int xmax, int k0, int kmax) { a64_transpose_interleave_16_1x4( reinterpret_cast(out), reinterpret_cast(in + k0 * stride + x0), (xmax-x0) * sizeof(int8_t) / 1, stride * sizeof(int8_t), (kmax-k0) ); } #endif // defined(__aarch64__)