src/core/NEON/kernels/arm_gemm/indirect-interleaves/a64_interleave4_block16_s8_s8_summing.hpp

/*
 * Copyright (c) 2019-2021, 2023 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.
 */

#ifdef __aarch64__

template<>
void interleave_block<4, 16, VLType::None, true>(
  int8_t * &out_ptr, const int8_t * const * in, size_t width, size_t height,
  size_t row_offset, bool first
)
{
  __asm__ __volatile__(
      "ldr x24, [%x[in], #0x0]\n"
      "ldr x23, [%x[in], #0x8]\n"
      "cmp %x[height], #0x4\n"
      "mov x22, #0x0\n"
      "ldr x21, [%x[in], #0x10]\n"
      "ldr x20, [%x[in], #0x18]\n"
      "movi v28.8h, #0x0\n"
      "movi v27.8h, #0x0\n"
      "movi v26.8h, #0x0\n"
      "movi v25.8h, #0x0\n"
      "add x24, x24, %x[row_offset]\n"
      "add x23, x23, %x[row_offset]\n"
      "movi v24.4s, #0x0\n"
      "movi v23.4s, #0x0\n"
      "add x21, x21, %x[row_offset]\n"
      "add x20, x20, %x[row_offset]\n"
      "movi v22.4s, #0x0\n"
      "movi v21.4s, #0x0\n"
      "beq 1f\n"
      "cmp %x[height], #0x2\n"
      "mov x20, x24\n"
      "csel x23, x23, x24, GE\n"
      "csel x21, x21, x24, GT\n"
      "1:"  // no_pointer_adj
      "prfm pldl1keep, [x24, #0x0]\n"
      "prfm pldl1keep, [x23, #0x0]\n"
      "movi v20.4s, #0x0\n"
      "prfm pldl1keep, [x21, #0x0]\n"
      "prfm pldl1keep, [x20, #0x0]\n"
      "prfm pldl1keep, [x24, #0x40]\n"
      "prfm pldl1keep, [x23, #0x40]\n"
      "prfm pldl1keep, [x21, #0x40]\n"
      "prfm pldl1keep, [x20, #0x40]\n"
      "cbnz %w[first], 2f\n"
      "sub %x[out_ptr], %x[out_ptr], #0x10\n"
      "ld1 { v20.4s }, [%x[out_ptr]]\n"
      "2:"  // first_pass
      "cmp %x[width], #0x10\n"
      "blt 5f\n"
      "3:"  // Main loop head
      "cmp x22, #0x7e\n"
      "ble 4f\n"
      "sadalp v24.4s, v28.8h\n"
      "movi v28.8h, #0x0\n"
      "mov x22, #0x0\n"
      "sadalp v23.4s, v27.8h\n"
      "movi v27.8h, #0x0\n"
      "sadalp v22.4s, v26.8h\n"
      "movi v26.8h, #0x0\n"
      "sadalp v21.4s, v25.8h\n"
      "movi v25.8h, #0x0\n"
      "4:"  // no_accumulate_16
      "ldr q19, [x24], #0x10\n"
      "ldr q18, [x23], #0x10\n"
      "subs %x[width], %x[width], #0x10\n"
      "cmp %x[width], #0x10\n"
      "ldr q17, [x21], #0x10\n"
      "ldr q16, [x20], #0x10\n"
      "str q19, [%x[out_ptr], #0x0]\n"
      "sadalp v28.8h, v19.16b\n"
      "prfm pldl1keep, [x24, #0x70]\n"
      "prfm pldl1keep, [x23, #0x70]\n"
      "str q18, [%x[out_ptr], #0x10]\n"
      "sadalp v27.8h, v18.16b\n"
      "prfm pldl1keep, [x21, #0x70]\n"
      "prfm pldl1keep, [x20, #0x70]\n"
      "str q17, [%x[out_ptr], #0x20]\n"
      "sadalp v26.8h, v17.16b\n"
      "str q16, [%x[out_ptr], #0x30]\n"
      "sadalp v25.8h, v16.16b\n"
      "add x22, x22, #0x1\n"
      "add %x[out_ptr], %x[out_ptr], #0x40\n"
      "bge 3b\n"
      "5:"  // Main loop skip
      "cbz %x[width], 14f\n"
      "tbz %x[width], #3, 9f\n"
      "ldr d19, [x24], #0x8\n"
      "ldr d18, [x23], #0x8\n"
      "ldr d17, [x21], #0x8\n"
      "ldr d16, [x20], #0x8\n"
      "tbz %x[width], #2, 7f\n"
      "ld1 { v19.s }[2], [x24], #0x4\n"
      "ld1 { v18.s }[2], [x23], #0x4\n"
      "ld1 { v17.s }[2], [x21], #0x4\n"
      "ld1 { v16.s }[2], [x20], #0x4\n"
      "tbz %x[width], #1, 6f\n"
      "ld1 { v19.h }[6], [x24], #0x2\n"
      "ld1 { v18.h }[6], [x23], #0x2\n"
      "ld1 { v17.h }[6], [x21], #0x2\n"
      "ld1 { v16.h }[6], [x20], #0x2\n"
      "tbz %x[width], #0, 13f\n"
      "ld1 { v19.b }[14], [x24]\n"
      "ld1 { v18.b }[14], [x23]\n"
      "ld1 { v17.b }[14], [x21]\n"
      "ld1 { v16.b }[14], [x20]\n"
      "b 13f\n"
      "6:"  // odd_loads_1_12
      "tbz %x[width], #0, 13f\n"
      "ld1 { v19.b }[12], [x24]\n"
      "ld1 { v18.b }[12], [x23]\n"
      "ld1 { v17.b }[12], [x21]\n"
      "ld1 { v16.b }[12], [x20]\n"
      "b 13f\n"
      "7:"  // odd_loads_2_8
      "tbz %x[width], #1, 8f\n"
      "ld1 { v19.h }[4], [x24], #0x2\n"
      "ld1 { v18.h }[4], [x23], #0x2\n"
      "ld1 { v17.h }[4], [x21], #0x2\n"
      "ld1 { v16.h }[4], [x20], #0x2\n"
      "tbz %x[width], #0, 13f\n"
      "ld1 { v19.b }[10], [x24]\n"
      "ld1 { v18.b }[10], [x23]\n"
      "ld1 { v17.b }[10], [x21]\n"
      "ld1 { v16.b }[10], [x20]\n"
      "b 13f\n"
      "8:"  // odd_loads_1_8
      "tbz %x[width], #0, 13f\n"
      "ld1 { v19.b }[8], [x24]\n"
      "ld1 { v18.b }[8], [x23]\n"
      "ld1 { v17.b }[8], [x21]\n"
      "ld1 { v16.b }[8], [x20]\n"
      "b 13f\n"
      "9:"  // odd_loads_4_0
      "tbz %x[width], #2, 11f\n"
      "ldr s19, [x24], #0x4\n"
      "ldr s18, [x23], #0x4\n"
      "ldr s17, [x21], #0x4\n"
      "ldr s16, [x20], #0x4\n"
      "tbz %x[width], #1, 10f\n"
      "ld1 { v19.h }[2], [x24], #0x2\n"
      "ld1 { v18.h }[2], [x23], #0x2\n"
      "ld1 { v17.h }[2], [x21], #0x2\n"
      "ld1 { v16.h }[2], [x20], #0x2\n"
      "tbz %x[width], #0, 13f\n"
      "ld1 { v19.b }[6], [x24]\n"
      "ld1 { v18.b }[6], [x23]\n"
      "ld1 { v17.b }[6], [x21]\n"
      "ld1 { v16.b }[6], [x20]\n"
      "b 13f\n"
      "10:"  // odd_loads_1_4
      "tbz %x[width], #0, 13f\n"
      "ld1 { v19.b }[4], [x24]\n"
      "ld1 { v18.b }[4], [x23]\n"
      "ld1 { v17.b }[4], [x21]\n"
      "ld1 { v16.b }[4], [x20]\n"
      "b 13f\n"
      "11:"  // odd_loads_2_0
      "tbz %x[width], #1, 12f\n"
      "ldr h19, [x24], #0x2\n"
      "ldr h18, [x23], #0x2\n"
      "ldr h17, [x21], #0x2\n"
      "ldr h16, [x20], #0x2\n"
      "tbz %x[width], #0, 13f\n"
      "ld1 { v19.b }[2], [x24]\n"
      "ld1 { v18.b }[2], [x23]\n"
      "ld1 { v17.b }[2], [x21]\n"
      "ld1 { v16.b }[2], [x20]\n"
      "b 13f\n"
      "12:"  // odd_loads_1_0
      "ldr b19, [x24, #0x0]\n"
      "ldr b18, [x23, #0x0]\n"
      "ldr b17, [x21, #0x0]\n"
      "ldr b16, [x20, #0x0]\n"
      "13:"  // Odd load end
      "str q19, [%x[out_ptr], #0x0]\n"
      "sadalp v28.8h, v19.16b\n"
      "sadalp v27.8h, v18.16b\n"
      "str q18, [%x[out_ptr], #0x10]\n"
      "sadalp v26.8h, v17.16b\n"
      "sadalp v25.8h, v16.16b\n"
      "str q17, [%x[out_ptr], #0x20]\n"
      "str q16, [%x[out_ptr], #0x30]\n"
      "add %x[out_ptr], %x[out_ptr], #0x40\n"
      "14:"  // Odds skip
      "sadalp v24.4s, v28.8h\n"
      "sadalp v23.4s, v27.8h\n"
      "sadalp v22.4s, v26.8h\n"
      "sadalp v21.4s, v25.8h\n"
      "addp v24.4s, v24.4s, v23.4s\n"
      "addp v23.4s, v22.4s, v21.4s\n"
      "addp v24.4s, v24.4s, v23.4s\n"
      "add v24.4s, v24.4s, v20.4s\n"
      "str q24, [%x[out_ptr], #0x0]\n"
      "add %x[out_ptr], %x[out_ptr], #0x10\n"
      : [out_ptr] "+&r" (out_ptr), [width] "+&r" (width)
      : [first] "r" (first), [height] "r" (height), [in] "r" (in), [row_offset] "r" (row_offset)
      : "cc", "memory", "v16", "v17", "v18", "v19", "v20", "v21", "v22", "v23", "v24", "v25", "v26", "v27", "v28", "x20", "x21", "x22", "x23", "x24"
    );
}


#endif // __aarch64__