/* 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. * * Copyright: * 2020 Evan Nemerson * 2020 Christopher Moore */ #if !defined(SIMDE_ARM_NEON_TBL_H) #define SIMDE_ARM_NEON_TBL_H #include "reinterpret.h" #include "types.h" HEDLEY_DIAGNOSTIC_PUSH SIMDE_DISABLE_UNWANTED_DIAGNOSTICS SIMDE_BEGIN_DECLS_ SIMDE_FUNCTION_ATTRIBUTES simde_uint8x8_t simde_vtbl1_u8(simde_uint8x8_t a, simde_uint8x8_t b) { #if defined(SIMDE_ARM_NEON_A32V7_NATIVE) return vtbl1_u8(a, b); #elif defined(SIMDE_X86_SSSE3_NATIVE) && defined(SIMDE_X86_MMX_NATIVE) return _mm_shuffle_pi8(a, _mm_or_si64(b, _mm_cmpgt_pi8(b, _mm_set1_pi8(7)))); #else simde_uint8x8_private r_, a_ = simde_uint8x8_to_private(a), b_ = simde_uint8x8_to_private(b); SIMDE_VECTORIZE for (size_t i = 0 ; i < (sizeof(r_.values) / sizeof(r_.values[0])) ; i++) { r_.values[i] = (b_.values[i] < 8) ? a_.values[b_.values[i]] : 0; } return simde_uint8x8_from_private(r_); #endif } #if defined(SIMDE_ARM_NEON_A32V7_ENABLE_NATIVE_ALIASES) #undef vtbl1_u8 #define vtbl1_u8(a, b) simde_vtbl1_u8((a), (b)) #endif SIMDE_FUNCTION_ATTRIBUTES simde_int8x8_t simde_vtbl1_s8(simde_int8x8_t a, simde_int8x8_t b) { #if defined(SIMDE_ARM_NEON_A32V7_NATIVE) return vtbl1_s8(a, b); #else return simde_vreinterpret_s8_u8(simde_vtbl1_u8(simde_vreinterpret_u8_s8(a), simde_vreinterpret_u8_s8(b))); #endif } #if defined(SIMDE_ARM_NEON_A32V7_ENABLE_NATIVE_ALIASES) #undef vtbl1_s8 #define vtbl1_s8(a, b) simde_vtbl1_s8((a), (b)) #endif #if !defined(SIMDE_BUG_INTEL_857088) SIMDE_FUNCTION_ATTRIBUTES simde_uint8x8_t simde_vtbl2_u8(simde_uint8x8x2_t a, simde_uint8x8_t b) { #if defined(SIMDE_ARM_NEON_A32V7_NATIVE) return vtbl2_u8(a, b); #elif defined(SIMDE_X86_SSSE3_NATIVE) && defined(SIMDE_X86_MMX_NATIVE) __m128i a128 = _mm_set_epi64(a.val[1], a.val[0]); __m128i b128 = _mm_set1_epi64(b); __m128i r128 = _mm_shuffle_epi8(a128, _mm_or_si128(b128, _mm_cmpgt_epi8(b128, _mm_set1_epi8(15)))); return _mm_movepi64_pi64(r128); #else simde_uint8x8_private r_, a_[2] = { simde_uint8x8_to_private(a.val[0]), simde_uint8x8_to_private(a.val[1]) }, b_ = simde_uint8x8_to_private(b); SIMDE_VECTORIZE for (size_t i = 0 ; i < (sizeof(r_.values) / sizeof(r_.values[0])) ; i++) { r_.values[i] = (b_.values[i] < 16) ? a_[b_.values[i] / 8].values[b_.values[i] & 7] : 0; } return simde_uint8x8_from_private(r_); #endif } #if defined(SIMDE_ARM_NEON_A32V7_ENABLE_NATIVE_ALIASES) #undef vtbl2_u8 #define vtbl2_u8(a, b) simde_vtbl2_u8((a), (b)) #endif SIMDE_FUNCTION_ATTRIBUTES simde_int8x8_t simde_vtbl2_s8(simde_int8x8x2_t a, simde_int8x8_t b) { #if defined(SIMDE_ARM_NEON_A32V7_NATIVE) return vtbl2_s8(a, b); #else simde_uint8x8x2_t a_; simde_memcpy(&a_, &a, sizeof(a_)); return simde_vreinterpret_s8_u8(simde_vtbl2_u8(a_, simde_vreinterpret_u8_s8(b))); #endif } #if defined(SIMDE_ARM_NEON_A32V7_ENABLE_NATIVE_ALIASES) #undef vtbl2_s8 #define vtbl2_s8(a, b) simde_vtbl2_s8((a), (b)) #endif SIMDE_FUNCTION_ATTRIBUTES simde_uint8x8_t simde_vtbl3_u8(simde_uint8x8x3_t a, simde_uint8x8_t b) { #if defined(SIMDE_ARM_NEON_A32V7_NATIVE) return vtbl3_u8(a, b); #elif defined(SIMDE_X86_SSE4_1_NATIVE) && defined(SIMDE_X86_MMX_NATIVE) __m128i b128 = _mm_set1_epi64(b); b128 = _mm_or_si128(b128, _mm_cmpgt_epi8(b128, _mm_set1_epi8(23))); __m128i r128_01 = _mm_shuffle_epi8(_mm_set_epi64(a.val[1], a.val[0]), b128); __m128i r128_2 = _mm_shuffle_epi8(_mm_set1_epi64(a.val[2]), b128); __m128i r128 = _mm_blendv_epi8(r128_01, r128_2, _mm_slli_epi32(b128, 3)); return _mm_movepi64_pi64(r128); #else simde_uint8x8_private r_, a_[3] = { simde_uint8x8_to_private(a.val[0]), simde_uint8x8_to_private(a.val[1]), simde_uint8x8_to_private(a.val[2]) }, b_ = simde_uint8x8_to_private(b); SIMDE_VECTORIZE for (size_t i = 0 ; i < (sizeof(r_.values) / sizeof(r_.values[0])) ; i++) { r_.values[i] = (b_.values[i] < 24) ? a_[b_.values[i] / 8].values[b_.values[i] & 7] : 0; } return simde_uint8x8_from_private(r_); #endif } #if defined(SIMDE_ARM_NEON_A32V7_ENABLE_NATIVE_ALIASES) #undef vtbl3_u8 #define vtbl3_u8(a, b) simde_vtbl3_u8((a), (b)) #endif SIMDE_FUNCTION_ATTRIBUTES simde_int8x8_t simde_vtbl3_s8(simde_int8x8x3_t a, simde_int8x8_t b) { #if defined(SIMDE_ARM_NEON_A32V7_NATIVE) return vtbl3_s8(a, b); #else simde_uint8x8x3_t a_; simde_memcpy(&a_, &a, sizeof(a_)); return simde_vreinterpret_s8_u8(simde_vtbl3_u8(a_, simde_vreinterpret_u8_s8(b))); #endif } #if defined(SIMDE_ARM_NEON_A32V7_ENABLE_NATIVE_ALIASES) #undef vtbl3_s8 #define vtbl3_s8(a, b) simde_vtbl3_s8((a), (b)) #endif SIMDE_FUNCTION_ATTRIBUTES simde_uint8x8_t simde_vtbl4_u8(simde_uint8x8x4_t a, simde_uint8x8_t b) { #if defined(SIMDE_ARM_NEON_A32V7_NATIVE) return vtbl4_u8(a, b); #elif defined(SIMDE_X86_SSE4_1_NATIVE) && defined(SIMDE_X86_MMX_NATIVE) __m128i b128 = _mm_set1_epi64(b); b128 = _mm_or_si128(b128, _mm_cmpgt_epi8(b128, _mm_set1_epi8(31))); __m128i r128_01 = _mm_shuffle_epi8(_mm_set_epi64(a.val[1], a.val[0]), b128); __m128i r128_23 = _mm_shuffle_epi8(_mm_set_epi64(a.val[3], a.val[2]), b128); __m128i r128 = _mm_blendv_epi8(r128_01, r128_23, _mm_slli_epi32(b128, 3)); return _mm_movepi64_pi64(r128); #else simde_uint8x8_private r_, a_[4] = { simde_uint8x8_to_private(a.val[0]), simde_uint8x8_to_private(a.val[1]), simde_uint8x8_to_private(a.val[2]), simde_uint8x8_to_private(a.val[3]) }, b_ = simde_uint8x8_to_private(b); SIMDE_VECTORIZE for (size_t i = 0 ; i < (sizeof(r_.values) / sizeof(r_.values[0])) ; i++) { r_.values[i] = (b_.values[i] < 32) ? a_[b_.values[i] / 8].values[b_.values[i] & 7] : 0; } return simde_uint8x8_from_private(r_); #endif } #if defined(SIMDE_ARM_NEON_A32V7_ENABLE_NATIVE_ALIASES) #undef vtbl4_u8 #define vtbl4_u8(a, b) simde_vtbl4_u8((a), (b)) #endif SIMDE_FUNCTION_ATTRIBUTES simde_int8x8_t simde_vtbl4_s8(simde_int8x8x4_t a, simde_int8x8_t b) { #if defined(SIMDE_ARM_NEON_A32V7_NATIVE) return vtbl4_s8(a, b); #else simde_uint8x8x4_t a_; simde_memcpy(&a_, &a, sizeof(a_)); return simde_vreinterpret_s8_u8(simde_vtbl4_u8(a_, simde_vreinterpret_u8_s8(b))); #endif } #if defined(SIMDE_ARM_NEON_A32V7_ENABLE_NATIVE_ALIASES) #undef vtbl4_s8 #define vtbl4_s8(a, b) simde_vtbl4_s8((a), (b)) #endif #endif /* !defined(SIMDE_BUG_INTEL_857088) */ SIMDE_END_DECLS_ HEDLEY_DIAGNOSTIC_POP #endif /* !defined(SIMDE_ARM_NEON_TBL_H) */