diff options
Diffstat (limited to 'third_party/simde/simde/arm/neon/trn2.h')
| -rw-r--r-- | third_party/simde/simde/arm/neon/trn2.h | 499 |
1 files changed, 499 insertions, 0 deletions
diff --git a/third_party/simde/simde/arm/neon/trn2.h b/third_party/simde/simde/arm/neon/trn2.h new file mode 100644 index 0000000000..31bd7dc4e7 --- /dev/null +++ b/third_party/simde/simde/arm/neon/trn2.h @@ -0,0 +1,499 @@ +/* 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 <evan@nemerson.com> + * 2020 Sean Maher <seanptmaher@gmail.com> (Copyright owned by Google, LLC) + */ + +#if !defined(SIMDE_ARM_NEON_TRN2_H) +#define SIMDE_ARM_NEON_TRN2_H + +#include "types.h" + +HEDLEY_DIAGNOSTIC_PUSH +SIMDE_DISABLE_UNWANTED_DIAGNOSTICS +SIMDE_BEGIN_DECLS_ + +SIMDE_FUNCTION_ATTRIBUTES +simde_float32x2_t +simde_vtrn2_f32(simde_float32x2_t a, simde_float32x2_t b) { + #if defined(SIMDE_ARM_NEON_A64V8_NATIVE) + return vtrn2_f32(a, b); + #else + simde_float32x2_private + r_, + a_ = simde_float32x2_to_private(a), + b_ = simde_float32x2_to_private(b); + + const size_t halfway_point = sizeof(r_.values) / sizeof(r_.values[0]) / 2; + SIMDE_VECTORIZE + for (size_t i = 0 ; i < halfway_point ; i++) { + const size_t idx = i << 1; + r_.values[idx] = a_.values[idx | 1]; + r_.values[idx | 1] = b_.values[idx | 1]; + } + + return simde_float32x2_from_private(r_); + #endif +} +#if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES) + #undef vtrn2_f32 + #define vtrn2_f32(a, b) simde_vtrn2_f32((a), (b)) +#endif + +SIMDE_FUNCTION_ATTRIBUTES +simde_int8x8_t +simde_vtrn2_s8(simde_int8x8_t a, simde_int8x8_t b) { + #if defined(SIMDE_ARM_NEON_A64V8_NATIVE) + return vtrn2_s8(a, b); + #else + simde_int8x8_private + r_, + a_ = simde_int8x8_to_private(a), + b_ = simde_int8x8_to_private(b); + + const size_t halfway_point = sizeof(r_.values) / sizeof(r_.values[0]) / 2; + SIMDE_VECTORIZE + for (size_t i = 0 ; i < halfway_point ; i++) { + const size_t idx = i << 1; + r_.values[idx] = a_.values[idx | 1]; + r_.values[idx | 1] = b_.values[idx | 1]; + } + + return simde_int8x8_from_private(r_); + #endif +} +#if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES) + #undef vtrn2_s8 + #define vtrn2_s8(a, b) simde_vtrn2_s8((a), (b)) +#endif + +SIMDE_FUNCTION_ATTRIBUTES +simde_int16x4_t +simde_vtrn2_s16(simde_int16x4_t a, simde_int16x4_t b) { + #if defined(SIMDE_ARM_NEON_A64V8_NATIVE) + return vtrn2_s16(a, b); + #else + simde_int16x4_private + r_, + a_ = simde_int16x4_to_private(a), + b_ = simde_int16x4_to_private(b); + + const size_t halfway_point = sizeof(r_.values) / sizeof(r_.values[0]) / 2; + SIMDE_VECTORIZE + for (size_t i = 0 ; i < halfway_point ; i++) { + const size_t idx = i << 1; + r_.values[idx] = a_.values[idx | 1]; + r_.values[idx | 1] = b_.values[idx | 1]; + } + + return simde_int16x4_from_private(r_); + #endif +} +#if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES) + #undef vtrn2_s16 + #define vtrn2_s16(a, b) simde_vtrn2_s16((a), (b)) +#endif + +SIMDE_FUNCTION_ATTRIBUTES +simde_int32x2_t +simde_vtrn2_s32(simde_int32x2_t a, simde_int32x2_t b) { + #if defined(SIMDE_ARM_NEON_A64V8_NATIVE) + return vtrn2_s32(a, b); + #else + simde_int32x2_private + r_, + a_ = simde_int32x2_to_private(a), + b_ = simde_int32x2_to_private(b); + + const size_t halfway_point = sizeof(r_.values) / sizeof(r_.values[0]) / 2; + SIMDE_VECTORIZE + for (size_t i = 0 ; i < halfway_point ; i++) { + const size_t idx = i << 1; + r_.values[idx] = a_.values[idx | 1]; + r_.values[idx | 1] = b_.values[idx | 1]; + } + + return simde_int32x2_from_private(r_); + #endif +} +#if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES) + #undef vtrn2_s32 + #define vtrn2_s32(a, b) simde_vtrn2_s32((a), (b)) +#endif + +SIMDE_FUNCTION_ATTRIBUTES +simde_uint8x8_t +simde_vtrn2_u8(simde_uint8x8_t a, simde_uint8x8_t b) { + #if defined(SIMDE_ARM_NEON_A64V8_NATIVE) + return vtrn2_u8(a, b); + #else + simde_uint8x8_private + r_, + a_ = simde_uint8x8_to_private(a), + b_ = simde_uint8x8_to_private(b); + + const size_t halfway_point = sizeof(r_.values) / sizeof(r_.values[0]) / 2; + SIMDE_VECTORIZE + for (size_t i = 0 ; i < halfway_point ; i++) { + const size_t idx = i << 1; + r_.values[idx] = a_.values[idx | 1]; + r_.values[idx | 1] = b_.values[idx | 1]; + } + + return simde_uint8x8_from_private(r_); + #endif +} +#if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES) + #undef vtrn2_u8 + #define vtrn2_u8(a, b) simde_vtrn2_u8((a), (b)) +#endif + +SIMDE_FUNCTION_ATTRIBUTES +simde_uint16x4_t +simde_vtrn2_u16(simde_uint16x4_t a, simde_uint16x4_t b) { + #if defined(SIMDE_ARM_NEON_A64V8_NATIVE) + return vtrn2_u16(a, b); + #else + simde_uint16x4_private + r_, + a_ = simde_uint16x4_to_private(a), + b_ = simde_uint16x4_to_private(b); + + const size_t halfway_point = sizeof(r_.values) / sizeof(r_.values[0]) / 2; + SIMDE_VECTORIZE + for (size_t i = 0 ; i < halfway_point ; i++) { + const size_t idx = i << 1; + r_.values[idx] = a_.values[idx | 1]; + r_.values[idx | 1] = b_.values[idx | 1]; + } + + return simde_uint16x4_from_private(r_); + #endif +} +#if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES) + #undef vtrn2_u16 + #define vtrn2_u16(a, b) simde_vtrn2_u16((a), (b)) +#endif + +SIMDE_FUNCTION_ATTRIBUTES +simde_uint32x2_t +simde_vtrn2_u32(simde_uint32x2_t a, simde_uint32x2_t b) { + #if defined(SIMDE_ARM_NEON_A64V8_NATIVE) + return vtrn2_u32(a, b); + #else + simde_uint32x2_private + r_, + a_ = simde_uint32x2_to_private(a), + b_ = simde_uint32x2_to_private(b); + + const size_t halfway_point = sizeof(r_.values) / sizeof(r_.values[0]) / 2; + SIMDE_VECTORIZE + for (size_t i = 0 ; i < halfway_point ; i++) { + const size_t idx = i << 1; + r_.values[idx] = a_.values[idx | 1]; + r_.values[idx | 1] = b_.values[idx | 1]; + } + + return simde_uint32x2_from_private(r_); + #endif +} +#if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES) + #undef vtrn2_u32 + #define vtrn2_u32(a, b) simde_vtrn2_u32((a), (b)) +#endif + +SIMDE_FUNCTION_ATTRIBUTES +simde_float32x4_t +simde_vtrn2q_f32(simde_float32x4_t a, simde_float32x4_t b) { + #if defined(SIMDE_ARM_NEON_A64V8_NATIVE) + return vtrn2q_f32(a, b); + #else + simde_float32x4_private + r_, + a_ = simde_float32x4_to_private(a), + b_ = simde_float32x4_to_private(b); + + const size_t halfway_point = sizeof(r_.values) / sizeof(r_.values[0]) / 2; + SIMDE_VECTORIZE + for (size_t i = 0 ; i < halfway_point ; i++) { + const size_t idx = i << 1; + r_.values[idx] = a_.values[idx | 1]; + r_.values[idx | 1] = b_.values[idx | 1]; + } + + return simde_float32x4_from_private(r_); + #endif +} +#if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES) + #undef vtrn2q_f32 + #define vtrn2q_f32(a, b) simde_vtrn2q_f32((a), (b)) +#endif + +SIMDE_FUNCTION_ATTRIBUTES +simde_float64x2_t +simde_vtrn2q_f64(simde_float64x2_t a, simde_float64x2_t b) { + #if defined(SIMDE_ARM_NEON_A64V8_NATIVE) + return vtrn2q_f64(a, b); + #else + simde_float64x2_private + r_, + a_ = simde_float64x2_to_private(a), + b_ = simde_float64x2_to_private(b); + + const size_t halfway_point = sizeof(r_.values) / sizeof(r_.values[0]) / 2; + SIMDE_VECTORIZE + for (size_t i = 0 ; i < halfway_point ; i++) { + const size_t idx = i << 1; + r_.values[idx] = a_.values[idx | 1]; + r_.values[idx | 1] = b_.values[idx | 1]; + } + + return simde_float64x2_from_private(r_); + #endif +} +#if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES) + #undef vtrn2q_f64 + #define vtrn2q_f64(a, b) simde_vtrn2q_f64((a), (b)) +#endif + +SIMDE_FUNCTION_ATTRIBUTES +simde_int8x16_t +simde_vtrn2q_s8(simde_int8x16_t a, simde_int8x16_t b) { + #if defined(SIMDE_ARM_NEON_A64V8_NATIVE) + return vtrn2q_s8(a, b); + #else + simde_int8x16_private + r_, + a_ = simde_int8x16_to_private(a), + b_ = simde_int8x16_to_private(b); + + const size_t halfway_point = sizeof(r_.values) / sizeof(r_.values[0]) / 2; + SIMDE_VECTORIZE + for (size_t i = 0 ; i < halfway_point ; i++) { + const size_t idx = i << 1; + r_.values[idx] = a_.values[idx | 1]; + r_.values[idx | 1] = b_.values[idx | 1]; + } + + return simde_int8x16_from_private(r_); + #endif +} +#if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES) + #undef vtrn2q_s8 + #define vtrn2q_s8(a, b) simde_vtrn2q_s8((a), (b)) +#endif + +SIMDE_FUNCTION_ATTRIBUTES +simde_int16x8_t +simde_vtrn2q_s16(simde_int16x8_t a, simde_int16x8_t b) { + #if defined(SIMDE_ARM_NEON_A64V8_NATIVE) + return vtrn2q_s16(a, b); + #else + simde_int16x8_private + r_, + a_ = simde_int16x8_to_private(a), + b_ = simde_int16x8_to_private(b); + + const size_t halfway_point = sizeof(r_.values) / sizeof(r_.values[0]) / 2; + SIMDE_VECTORIZE + for (size_t i = 0 ; i < halfway_point ; i++) { + const size_t idx = i << 1; + r_.values[idx] = a_.values[idx | 1]; + r_.values[idx | 1] = b_.values[idx | 1]; + } + + return simde_int16x8_from_private(r_); + #endif +} +#if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES) + #undef vtrn2q_s16 + #define vtrn2q_s16(a, b) simde_vtrn2q_s16((a), (b)) +#endif + +SIMDE_FUNCTION_ATTRIBUTES +simde_int32x4_t +simde_vtrn2q_s32(simde_int32x4_t a, simde_int32x4_t b) { + #if defined(SIMDE_ARM_NEON_A64V8_NATIVE) + return vtrn2q_s32(a, b); + #else + simde_int32x4_private + r_, + a_ = simde_int32x4_to_private(a), + b_ = simde_int32x4_to_private(b); + + const size_t halfway_point = sizeof(r_.values) / sizeof(r_.values[0]) / 2; + SIMDE_VECTORIZE + for (size_t i = 0 ; i < halfway_point ; i++) { + const size_t idx = i << 1; + r_.values[idx] = a_.values[idx | 1]; + r_.values[idx | 1] = b_.values[idx | 1]; + } + + return simde_int32x4_from_private(r_); + #endif +} +#if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES) + #undef vtrn2q_s32 + #define vtrn2q_s32(a, b) simde_vtrn2q_s32((a), (b)) +#endif + +SIMDE_FUNCTION_ATTRIBUTES +simde_int64x2_t +simde_vtrn2q_s64(simde_int64x2_t a, simde_int64x2_t b) { + #if defined(SIMDE_ARM_NEON_A64V8_NATIVE) + return vtrn2q_s64(a, b); + #else + simde_int64x2_private + r_, + a_ = simde_int64x2_to_private(a), + b_ = simde_int64x2_to_private(b); + + const size_t halfway_point = sizeof(r_.values) / sizeof(r_.values[0]) / 2; + SIMDE_VECTORIZE + for (size_t i = 0 ; i < halfway_point ; i++) { + const size_t idx = i << 1; + r_.values[idx] = a_.values[idx | 1]; + r_.values[idx | 1] = b_.values[idx | 1]; + } + + return simde_int64x2_from_private(r_); + #endif +} +#if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES) + #undef vtrn2q_s64 + #define vtrn2q_s64(a, b) simde_vtrn2q_s64((a), (b)) +#endif + +SIMDE_FUNCTION_ATTRIBUTES +simde_uint8x16_t +simde_vtrn2q_u8(simde_uint8x16_t a, simde_uint8x16_t b) { + #if defined(SIMDE_ARM_NEON_A64V8_NATIVE) + return vtrn2q_u8(a, b); + #else + simde_uint8x16_private + r_, + a_ = simde_uint8x16_to_private(a), + b_ = simde_uint8x16_to_private(b); + + const size_t halfway_point = sizeof(r_.values) / sizeof(r_.values[0]) / 2; + SIMDE_VECTORIZE + for (size_t i = 0 ; i < halfway_point ; i++) { + const size_t idx = i << 1; + r_.values[idx] = a_.values[idx | 1]; + r_.values[idx | 1] = b_.values[idx | 1]; + } + + return simde_uint8x16_from_private(r_); + #endif +} +#if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES) + #undef vtrn2q_u8 + #define vtrn2q_u8(a, b) simde_vtrn2q_u8((a), (b)) +#endif + +SIMDE_FUNCTION_ATTRIBUTES +simde_uint16x8_t +simde_vtrn2q_u16(simde_uint16x8_t a, simde_uint16x8_t b) { + #if defined(SIMDE_ARM_NEON_A64V8_NATIVE) + return vtrn2q_u16(a, b); + #else + simde_uint16x8_private + r_, + a_ = simde_uint16x8_to_private(a), + b_ = simde_uint16x8_to_private(b); + + const size_t halfway_point = sizeof(r_.values) / sizeof(r_.values[0]) / 2; + SIMDE_VECTORIZE + for (size_t i = 0 ; i < halfway_point ; i++) { + const size_t idx = i << 1; + r_.values[idx] = a_.values[idx | 1]; + r_.values[idx | 1] = b_.values[idx | 1]; + } + + return simde_uint16x8_from_private(r_); + #endif +} +#if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES) + #undef vtrn2q_u16 + #define vtrn2q_u16(a, b) simde_vtrn2q_u16((a), (b)) +#endif + +SIMDE_FUNCTION_ATTRIBUTES +simde_uint32x4_t +simde_vtrn2q_u32(simde_uint32x4_t a, simde_uint32x4_t b) { + #if defined(SIMDE_ARM_NEON_A64V8_NATIVE) + return vtrn2q_u32(a, b); + #else + simde_uint32x4_private + r_, + a_ = simde_uint32x4_to_private(a), + b_ = simde_uint32x4_to_private(b); + + const size_t halfway_point = sizeof(r_.values) / sizeof(r_.values[0]) / 2; + SIMDE_VECTORIZE + for (size_t i = 0 ; i < halfway_point ; i++) { + const size_t idx = i << 1; + r_.values[idx] = a_.values[idx | 1]; + r_.values[idx | 1] = b_.values[idx | 1]; + } + + return simde_uint32x4_from_private(r_); + #endif +} +#if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES) + #undef vtrn2q_u32 + #define vtrn2q_u32(a, b) simde_vtrn2q_u32((a), (b)) +#endif + +SIMDE_FUNCTION_ATTRIBUTES +simde_uint64x2_t +simde_vtrn2q_u64(simde_uint64x2_t a, simde_uint64x2_t b) { + #if defined(SIMDE_ARM_NEON_A64V8_NATIVE) + return vtrn2q_u64(a, b); + #else + simde_uint64x2_private + r_, + a_ = simde_uint64x2_to_private(a), + b_ = simde_uint64x2_to_private(b); + + const size_t halfway_point = sizeof(r_.values) / sizeof(r_.values[0]) / 2; + SIMDE_VECTORIZE + for (size_t i = 0 ; i < halfway_point ; i++) { + const size_t idx = i << 1; + r_.values[idx] = a_.values[idx | 1]; + r_.values[idx | 1] = b_.values[idx | 1]; + } + + return simde_uint64x2_from_private(r_); + #endif +} +#if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES) + #undef vtrn2q_u64 + #define vtrn2q_u64(a, b) simde_vtrn2q_u64((a), (b)) +#endif + +SIMDE_END_DECLS_ +HEDLEY_DIAGNOSTIC_POP + +#endif /* !defined(SIMDE_ARM_NEON_TRN2_H) */ |