From 03b2971ac69a86f10a1566938d1a25afee15746c Mon Sep 17 00:00:00 2001 From: Viet-Hoa Do Date: Wed, 1 Jun 2022 11:47:14 +0100 Subject: Integrate SME2 kernels * Add SME/SME2 detection. * Integrate SME2 implementation for: - Normal convolution - Winograd - Depthwise convolution - Pooling Resolves: COMPMID-5700 Signed-off-by: Viet-Hoa Do Change-Id: I2f1ca1d05f8cfeee9309ed1c0a36096a4a6aad5c Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/8692 Reviewed-by: Gunes Bayir Tested-by: Arm Jenkins Comments-Addressed: Arm Jenkins --- .../sme_interleave4VL_block2_bf16_bf16.hpp | 127 +++++++++++++++++++++ 1 file changed, 127 insertions(+) create mode 100644 src/core/NEON/kernels/arm_gemm/indirect-interleaves/sme_interleave4VL_block2_bf16_bf16.hpp (limited to 'src/core/NEON/kernels/arm_gemm/indirect-interleaves/sme_interleave4VL_block2_bf16_bf16.hpp') diff --git a/src/core/NEON/kernels/arm_gemm/indirect-interleaves/sme_interleave4VL_block2_bf16_bf16.hpp b/src/core/NEON/kernels/arm_gemm/indirect-interleaves/sme_interleave4VL_block2_bf16_bf16.hpp new file mode 100644 index 0000000000..556d1481de --- /dev/null +++ b/src/core/NEON/kernels/arm_gemm/indirect-interleaves/sme_interleave4VL_block2_bf16_bf16.hpp @@ -0,0 +1,127 @@ +/* + * Copyright (c) 2022 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. + */ + +#if defined(__ARM_FEATURE_SVE) + +template <> +void interleave_block<4, 2, VLType::SME, false>( + bfloat16 * &out, const bfloat16 * const *in, + size_t width, size_t height, size_t row_offset, bool first +) +{ + ARM_COMPUTE_UNUSED(first); + + __asm__ __volatile__( + ".inst 0xd503477f // SMSTART ZA\n" + "cntw x15\n" + "cntw x14, ALL, MUL #2\n" + "cntw x13, ALL, MUL #3\n" + "cnth x11\n" + "ptrue p13.s\n" + "cntw x10\n" + "cmp %x[height], x10\n" + "csel x10, %x[height], x10, LT\n" + "sub x10, x10, #0x1\n" + "whilelt p10.h, XZR, %x[height]\n" + "whilelt p9.h, x15, %x[height]\n" + "whilelt p8.h, x14, %x[height]\n" + "zip1 p12.h, p10.h, p8.h\n" + "whilelt p8.h, x13, %x[height]\n" + "zip1 p11.h, p9.h, p8.h\n" + "mov x9, %x[row_offset]\n" + "mov x28, %x[out]\n" + "mov x27, #0x0\n" + "whilelt p10.h, x27, %x[width]\n" + "whilelt p9.h, x27, %x[width]\n" + "whilelt p8.h, x27, %x[width]\n" + "1:" // Width loop + "mov x12, #0x0\n" + "add x26, %x[in], XZR, LSL #3\n" + "add x25, %x[in], x15, LSL #3\n" + "add x24, %x[in], x14, LSL #3\n" + "add x23, %x[in], x13, LSL #3\n" + "ldr x22, [x26], #0x8\n" + "ldr x21, [x25], #0x8\n" + "ldr x20, [x24], #0x8\n" + "ldr x19, [x23], #0x8\n" + "cbz x10, 3f\n" + "2:" // Loads: Loop + ".inst 0x25286580 // dup p0.h, p9.h/Z, p12.h[w12]\n" + ".inst 0xe04902c0 // ld1h { za0h.h[x12] }, p0/Z, [x22, x9, LSL #1]\n" + ".inst 0x25286160 // dup p0.h, p8.h/Z, p11.h[w12]\n" + ".inst 0xe04902a8 // ld1h { za1h.h[x12] }, p0/Z, [x21, x9, LSL #1]\n" + ".inst 0x25386580 // dup p0.h, p9.h/Z, p12.h[w12, #1]\n" + ".inst 0xe0490281 // ld1h { za0h.h[x12, #1] }, p0/Z, [x20, x9, LSL #1]\n" + ".inst 0x25386160 // dup p0.h, p8.h/Z, p11.h[w12, #1]\n" + ".inst 0xe0490269 // ld1h { za1h.h[x12, #1] }, p0/Z, [x19, x9, LSL #1]\n" + "ldr x22, [x26], #0x8\n" + "ldr x21, [x25], #0x8\n" + "ldr x20, [x24], #0x8\n" + "ldr x19, [x23], #0x8\n" + "add x12, x12, #0x2\n" + "cmp x12, x10, LSL #1\n" + "blt 2b\n" + "3:" // Loads: Tail + ".inst 0x25286580 // dup p0.h, p9.h/Z, p12.h[w12]\n" + ".inst 0xe04902c0 // ld1h { za0h.h[x12] }, p0/Z, [x22, x9, LSL #1]\n" + ".inst 0x25286160 // dup p0.h, p8.h/Z, p11.h[w12]\n" + ".inst 0xe04902a8 // ld1h { za1h.h[x12] }, p0/Z, [x21, x9, LSL #1]\n" + ".inst 0x25386580 // dup p0.h, p9.h/Z, p12.h[w12, #1]\n" + ".inst 0xe0490281 // ld1h { za0h.h[x12, #1] }, p0/Z, [x20, x9, LSL #1]\n" + ".inst 0x25386160 // dup p0.h, p8.h/Z, p11.h[w12, #1]\n" + ".inst 0xe0490269 // ld1h { za1h.h[x12, #1] }, p0/Z, [x19, x9, LSL #1]\n" + "mov x12, #0x0\n" + "sub x19, %x[width], x27\n" + "cmp x19, x11\n" + "csel x19, x19, x11, LT\n" + "add x19, x19, #0x1\n" + "lsr x19, x19, #0x1\n" + "4:" // Stores: Loop + ".inst 0x25307540 // dup p0.s, p13.s/Z, p10.s[w12]\n" + ".inst 0xe0bf8380 // st1w { za0v.s[x12] }, p0/Z, [x28, XZR, LSL #2]\n" + ".inst 0x25307540 // dup p0.s, p13.s/Z, p10.s[w12]\n" + ".inst 0x25307541 // dup p1.s, p13.s/Z, p10.s[w12]\n" + ".inst 0xe0af8384 // st1w { za1v.s[x12] }, p0/Z, [x28, x15, LSL #2]\n" + ".inst 0x25307540 // dup p0.s, p13.s/Z, p10.s[w12]\n" + ".inst 0xe0ae8788 // st1w { za2v.s[x12] }, p1/Z, [x28, x14, LSL #2]\n" + ".inst 0xe0ad838c // st1w { za3v.s[x12] }, p0/Z, [x28, x13, LSL #2]\n" + "addvl x28, x28, #4\n" + "add x12, x12, #0x1\n" + "cmp x12, x19\n" + "blt 4b\n" + "inch x9\n" + "inch x27\n" + "whilelt p10.h, x27, %x[width]\n" + "whilelt p9.h, x27, %x[width]\n" + "whilelt p8.h, x27, %x[width]\n" + "b.any 1b\n" + "mov %x[out], x28\n" + ".inst 0xd503467f // SMSTOP\n" + : [out] "+&r" (out) + : [height] "r" (height), [in] "r" (in), [row_offset] "r" (row_offset), [width] "r" (width) + : "cc", "memory", "p0", "p1", "p8", "p9", "p10", "p11", "p12", "p13", "x9", "x10", "x11", "x12", "x13", "x14", "x15", "x19", "x20", "x21", "x22", "x23", "x24", "x25", "x26", "x27", "x28", "z0", "z1", "z2", "z3", "z4", "z5", "z6", "z7", "z8", "z9", "z10", "z11", "z12", "z13", "z14", "z15", "z16", "z17", "z18", "z19", "z20", "z21", "z22", "z23", "z24", "z25", "z26", "z27", "z28", "z29", "z30", "z31" + ); +} + +#endif // defined(__ARM_FEATURE_SVE) -- cgit v1.2.1