eden-emu/eden
24
3
Fork 15
Code Issues 138 Pull requests 56 Projects 1 Releases Packages Wiki Activity Actions

Minor updates
All checks were successful
eden-license / license-header (pull_request) Successful in 13s
Details

Browse source
This commit is contained in:
Gamer64 2025-08-01 20:13:16 +02:00 • committed by crueter
parent c897d998b7
commit 37a410558e
1 changed files with 13 additions and 15 deletions
Download patch file Download diff file Expand all files Collapse all files

28
src/common/vector_math.h
View file

@ -648,24 +648,22 @@ template <typename T>
return a.x * b.x + a.y * b.y + a.z * b.z + a.w * b.w; return a.x * b.x + a.y * b.y + a.z * b.z + a.w * b.w;
} }
#ifdef __ARM_NEON__ template <>
// NEON-accelerated overload for float Vec4 dot product [[nodiscard]] inline float Dot(const Vec4<float>& a, const Vec4<float>& b) {
inline float Dot(const Vec4<float>& a, const Vec4<float>& b) { #ifdef __ARM_NEON
// Load 4 floats into NEON registers
float32x4_t va = vld1q_f32(&a.x); float32x4_t va = vld1q_f32(&a.x);
float32x4_t vb = vld1q_f32(&b.x); float32x4_t vb = vld1q_f32(&b.x);
// Element-wise multiply float32x4_t result = vmulq_f32(va, vb);
float32x4_t prod = vmulq_f32(va, vb); #if defined(__aarch64__) // Use vaddvq_f32 in ARMv8 architectures
return vaddvq_f32(result);
// Horizontal add across the vector #else // Use manual addition for older architectures
#if defined(__aarch64__) float32x2_t sum2 = vadd_f32(vget_high_f32(result), vget_low_f32(result));
return vaddvq_f32(prod); return vget_lane_f32(vpadd_f32(sum2, sum2), 0);
#else
float32x2_t sum2 = vadd_f32(vget_high_f32(prod), vget_low_f32(prod));
return vget_lane_f32(vpadd_f32(sum2, sum2), 0);
#endif
}
#endif #endif
#else
return a.x * b.x + a.y * b.y + a.z * b.z + a.w * b.w;
#endif
}
template <typename T> template <typename T>
[[nodiscard]] constexpr Vec3<decltype(T{} * T{} - T{} * T{})> Cross(const Vec3<T>& a, [[nodiscard]] constexpr Vec3<decltype(T{} * T{} - T{} * T{})> Cross(const Vec3<T>& a,