/* 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:
*   2021      Atharva Nimbalkar <atharvakn@gmail.com>
*   2023      Yi-Yen Chung <eric681@andestech.com> (Copyright owned by Andes Technology)
*/

#if !defined(SIMDE_ARM_NEON_FMA_H)
#define SIMDE_ARM_NEON_FMA_H

#include "add.h"
#include "mul.h"

HEDLEY_DIAGNOSTIC_PUSH
SIMDE_DISABLE_UNWANTED_DIAGNOSTICS
SIMDE_BEGIN_DECLS_

SIMDE_FUNCTION_ATTRIBUTES
simde_float16_t
simde_vfmah_f16(simde_float16_t a, simde_float16_t b, simde_float16_t c) {
  #if defined(SIMDE_ARM_NEON_A32V8_NATIVE) && defined(SIMDE_ARCH_ARM_FMA) && defined(SIMDE_ARM_NEON_FP16)
    return vfmah_f16(a, b, c);
  #else
    return simde_vaddh_f16(a, simde_vmulh_f16(b, c));
  #endif
}
#if defined(SIMDE_ARM_NEON_A32V8_ENABLE_NATIVE_ALIASES)
  #undef vfmah_f16
  #define vfmah_f16(a, b, c) simde_vfmah_f16(a, b, c)
#endif

SIMDE_FUNCTION_ATTRIBUTES
simde_float32x2_t
simde_vfma_f32(simde_float32x2_t a, simde_float32x2_t b, simde_float32x2_t c) {
  #if defined(SIMDE_ARM_NEON_A32V7_NATIVE) && defined(SIMDE_ARCH_ARM_FMA)
    return vfma_f32(a, b, c);
  #else
    return simde_vadd_f32(a, simde_vmul_f32(b, c));
  #endif
}
#if defined(SIMDE_ARM_NEON_A32V7_ENABLE_NATIVE_ALIASES)
  #undef vfma_f32
  #define vfma_f32(a, b, c) simde_vfma_f32(a, b, c)
#endif

SIMDE_FUNCTION_ATTRIBUTES
simde_float64x1_t
simde_vfma_f64(simde_float64x1_t a, simde_float64x1_t b, simde_float64x1_t c) {
  #if defined(SIMDE_ARM_NEON_A64V8_NATIVE) && defined(SIMDE_ARCH_ARM_FMA)
    return vfma_f64(a, b, c);
  #else
    return simde_vadd_f64(a, simde_vmul_f64(b, c));
  #endif
}
#if defined(SIMDE_ARM_NEON_A32V7_ENABLE_NATIVE_ALIASES)
  #undef vfma_f64
  #define vfma_f64(a, b, c) simde_vfma_f64(a, b, c)
#endif

SIMDE_FUNCTION_ATTRIBUTES
simde_float16x4_t
simde_vfma_f16(simde_float16x4_t a, simde_float16x4_t b, simde_float16x4_t c) {
  #if defined(SIMDE_ARM_NEON_A32V8_NATIVE) && defined(SIMDE_ARCH_ARM_FMA) && defined(SIMDE_ARM_NEON_FP16)
    return vfma_f16(a, b, c);
  #else
    return simde_vadd_f16(a, simde_vmul_f16(b, c));
  #endif
}
#if defined(SIMDE_ARM_NEON_A32V8_ENABLE_NATIVE_ALIASES)
  #undef vfma_f16
  #define vfma_f16(a, b, c) simde_vfma_f16(a, b, c)
#endif

SIMDE_FUNCTION_ATTRIBUTES
simde_float16x8_t
simde_vfmaq_f16(simde_float16x8_t a, simde_float16x8_t b, simde_float16x8_t c) {
  #if defined(SIMDE_ARM_NEON_A32V8_NATIVE) && defined(SIMDE_ARCH_ARM_FMA) && defined(SIMDE_ARM_NEON_FP16)
    return vfmaq_f16(a, b, c);
  #else
    return simde_vaddq_f16(a, simde_vmulq_f16(b, c));
  #endif
}
#if defined(SIMDE_ARM_NEON_A32V8_ENABLE_NATIVE_ALIASES)
  #undef vfmaq_f16
  #define vfmaq_f16(a, b, c) simde_vfmaq_f16(a, b, c)
#endif

SIMDE_FUNCTION_ATTRIBUTES
simde_float32x4_t
simde_vfmaq_f32(simde_float32x4_t a, simde_float32x4_t b, simde_float32x4_t c) {
  #if defined(SIMDE_ARM_NEON_A32V7_NATIVE) && defined(SIMDE_ARCH_ARM_FMA)
    return vfmaq_f32(a, b, c);
  #elif defined(SIMDE_POWER_ALTIVEC_P6_NATIVE)
    return vec_madd(b, c, a);
  #elif \
      defined(SIMDE_X86_FMA_NATIVE)
    simde_float32x4_private
      r_,
      a_ = simde_float32x4_to_private(a),
      b_ = simde_float32x4_to_private(b),
      c_ = simde_float32x4_to_private(c);

    #if defined(SIMDE_X86_FMA_NATIVE)
      r_.m128 = _mm_fmadd_ps(b_.m128, c_.m128, a_.m128);
    #endif

    return simde_float32x4_from_private(r_);
  #else
    return simde_vaddq_f32(a, simde_vmulq_f32(b, c));
  #endif
}
#if defined(SIMDE_ARM_NEON_A32V7_ENABLE_NATIVE_ALIASES)
  #undef vfmaq_f32
  #define vfmaq_f32(a, b, c) simde_vfmaq_f32(a, b, c)
#endif

SIMDE_FUNCTION_ATTRIBUTES
simde_float64x2_t
simde_vfmaq_f64(simde_float64x2_t a, simde_float64x2_t b, simde_float64x2_t c) {
  #if defined(SIMDE_ARM_NEON_A64V8_NATIVE) && defined(SIMDE_ARCH_ARM_FMA)
    return vfmaq_f64(a, b, c);
  #elif defined(SIMDE_POWER_ALTIVEC_P7_NATIVE) || defined(SIMDE_ZARCH_ZVECTOR_13_NATIVE)
    return vec_madd(b, c, a);
  #elif \
      defined(SIMDE_X86_FMA_NATIVE)
    simde_float64x2_private
      r_,
      a_ = simde_float64x2_to_private(a),
      b_ = simde_float64x2_to_private(b),
      c_ = simde_float64x2_to_private(c);

    #if defined(SIMDE_X86_FMA_NATIVE)
      r_.m128d = _mm_fmadd_pd(b_.m128d, c_.m128d, a_.m128d);
    #endif

    return simde_float64x2_from_private(r_);
  #else
    return simde_vaddq_f64(a, simde_vmulq_f64(b, c));
  #endif
}
#if defined(SIMDE_ARM_NEON_A32V7_ENABLE_NATIVE_ALIASES)
  #undef vfmaq_f64
  #define vfmaq_f64(a, b, c) simde_vfmaq_f64(a, b, c)
#endif

SIMDE_END_DECLS_
HEDLEY_DIAGNOSTIC_POP

#endif /* !defined(SIMDE_ARM_NEON_CMLA_H) */