192 lines
5.3 KiB
C
192 lines
5.3 KiB
C
/* 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:
|
|
* 2023 Yi-Yen Chung <eric681@andestech.com> (Copyright owned by Andes Technology)
|
|
*/
|
|
|
|
#if !defined(SIMDE_ARM_NEON_RNDA_H)
|
|
#define SIMDE_ARM_NEON_RNDA_H
|
|
|
|
#include "types.h"
|
|
|
|
HEDLEY_DIAGNOSTIC_PUSH
|
|
SIMDE_DISABLE_UNWANTED_DIAGNOSTICS
|
|
SIMDE_BEGIN_DECLS_
|
|
|
|
SIMDE_FUNCTION_ATTRIBUTES
|
|
simde_float16_t
|
|
simde_vrndah_f16(simde_float16_t a) {
|
|
#if defined(SIMDE_ARM_NEON_A32V8_NATIVE) && defined(SIMDE_ARM_NEON_FP16)
|
|
return vrndah_f16(a);
|
|
#else
|
|
return simde_float16_from_float32(simde_math_roundf(simde_float16_to_float32(a)));
|
|
#endif
|
|
}
|
|
#if defined(SIMDE_ARM_NEON_A32V8_ENABLE_NATIVE_ALIASES)
|
|
#undef vrndah_f16
|
|
#define vrndah_f16(a) simde_vrndah_f16(a)
|
|
#endif
|
|
|
|
SIMDE_FUNCTION_ATTRIBUTES
|
|
simde_float16x4_t
|
|
simde_vrnda_f16(simde_float16x4_t a) {
|
|
#if defined(SIMDE_ARM_NEON_A32V8_NATIVE) && defined(SIMDE_ARM_NEON_FP16)
|
|
return vrnda_f16(a);
|
|
#else
|
|
simde_float16x4_private
|
|
r_,
|
|
a_ = simde_float16x4_to_private(a);
|
|
|
|
SIMDE_VECTORIZE
|
|
for (size_t i = 0 ; i < (sizeof(r_.values) / sizeof(r_.values[0])) ; i++) {
|
|
r_.values[i] = simde_vrndah_f16(a_.values[i]);
|
|
}
|
|
|
|
return simde_float16x4_from_private(r_);
|
|
#endif
|
|
}
|
|
#if defined(SIMDE_ARM_NEON_A32V8_ENABLE_NATIVE_ALIASES)
|
|
#undef vrnda_f16
|
|
#define vrnda_f16(a) simde_vrnda_f16(a)
|
|
#endif
|
|
|
|
SIMDE_FUNCTION_ATTRIBUTES
|
|
simde_float32x2_t
|
|
simde_vrnda_f32(simde_float32x2_t a) {
|
|
#if defined(SIMDE_ARM_NEON_A32V8_NATIVE)
|
|
return vrnda_f32(a);
|
|
#else
|
|
simde_float32x2_private
|
|
r_,
|
|
a_ = simde_float32x2_to_private(a);
|
|
|
|
SIMDE_VECTORIZE
|
|
for (size_t i = 0 ; i < (sizeof(r_.values) / sizeof(r_.values[0])) ; i++) {
|
|
r_.values[i] = simde_math_roundf(a_.values[i]);
|
|
}
|
|
|
|
return simde_float32x2_from_private(r_);
|
|
#endif
|
|
}
|
|
#if defined(SIMDE_ARM_NEON_A32V8_ENABLE_NATIVE_ALIASES)
|
|
#undef vrnda_f32
|
|
#define vrnda_f32(a) simde_vrnda_f32(a)
|
|
#endif
|
|
|
|
SIMDE_FUNCTION_ATTRIBUTES
|
|
simde_float64x1_t
|
|
simde_vrnda_f64(simde_float64x1_t a) {
|
|
#if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
|
|
return vrnda_f64(a);
|
|
#else
|
|
simde_float64x1_private
|
|
r_,
|
|
a_ = simde_float64x1_to_private(a);
|
|
|
|
SIMDE_VECTORIZE
|
|
for (size_t i = 0 ; i < (sizeof(r_.values) / sizeof(r_.values[0])) ; i++) {
|
|
r_.values[i] = simde_math_round(a_.values[i]);
|
|
}
|
|
|
|
return simde_float64x1_from_private(r_);
|
|
#endif
|
|
}
|
|
#if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
|
|
#undef vrnda_f64
|
|
#define vrnda_f64(a) simde_vrnda_f64(a)
|
|
#endif
|
|
|
|
SIMDE_FUNCTION_ATTRIBUTES
|
|
simde_float16x8_t
|
|
simde_vrndaq_f16(simde_float16x8_t a) {
|
|
#if defined(SIMDE_ARM_NEON_A32V8_NATIVE) && defined(SIMDE_ARM_NEON_FP16)
|
|
return vrndaq_f16(a);
|
|
#else
|
|
simde_float16x8_private
|
|
r_,
|
|
a_ = simde_float16x8_to_private(a);
|
|
|
|
SIMDE_VECTORIZE
|
|
for (size_t i = 0 ; i < (sizeof(r_.values) / sizeof(r_.values[0])) ; i++) {
|
|
r_.values[i] = simde_vrndah_f16(a_.values[i]);
|
|
}
|
|
|
|
return simde_float16x8_from_private(r_);
|
|
#endif
|
|
}
|
|
#if defined(SIMDE_ARM_NEON_A32V8_ENABLE_NATIVE_ALIASES)
|
|
#undef vrndaq_f16
|
|
#define vrndaq_f16(a) simde_vrndaq_f16(a)
|
|
#endif
|
|
|
|
SIMDE_FUNCTION_ATTRIBUTES
|
|
simde_float32x4_t
|
|
simde_vrndaq_f32(simde_float32x4_t a) {
|
|
#if defined(SIMDE_ARM_NEON_A32V8_NATIVE)
|
|
return vrndaq_f32(a);
|
|
#else
|
|
simde_float32x4_private
|
|
r_,
|
|
a_ = simde_float32x4_to_private(a);
|
|
|
|
SIMDE_VECTORIZE
|
|
for (size_t i = 0 ; i < (sizeof(r_.values) / sizeof(r_.values[0])) ; i++) {
|
|
r_.values[i] = simde_math_roundf(a_.values[i]);
|
|
}
|
|
|
|
return simde_float32x4_from_private(r_);
|
|
#endif
|
|
}
|
|
#if defined(SIMDE_ARM_NEON_A32V8_ENABLE_NATIVE_ALIASES)
|
|
#undef vrndaq_f32
|
|
#define vrndaq_f32(a) simde_vrndaq_f32(a)
|
|
#endif
|
|
|
|
SIMDE_FUNCTION_ATTRIBUTES
|
|
simde_float64x2_t
|
|
simde_vrndaq_f64(simde_float64x2_t a) {
|
|
#if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
|
|
return vrndaq_f64(a);
|
|
#else
|
|
simde_float64x2_private
|
|
r_,
|
|
a_ = simde_float64x2_to_private(a);
|
|
|
|
SIMDE_VECTORIZE
|
|
for (size_t i = 0 ; i < (sizeof(r_.values) / sizeof(r_.values[0])) ; i++) {
|
|
r_.values[i] = simde_math_round(a_.values[i]);
|
|
}
|
|
|
|
return simde_float64x2_from_private(r_);
|
|
#endif
|
|
}
|
|
#if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
|
|
#undef vrndaq_f64
|
|
#define vrndaq_f64(a) simde_vrndaq_f64(a)
|
|
#endif
|
|
|
|
SIMDE_END_DECLS_
|
|
HEDLEY_DIAGNOSTIC_POP
|
|
|
|
#endif /* !defined(SIMDE_ARM_NEON_RNDA_H) */
|