guezoloic/t3dsr
1
0
Fork 0
mirror of https://github.com/guezoloic/t3dsr.git synced 2026-01-25 09:34:24 +00:00
Code Issues Packages Projects Releases Wiki Activity

refactor: rename common_math.h to mconfig.h

Browse Source
This commit is contained in:
Loïc GUEZO
2025-06-21 22:11:45 +02:00
parent 3f005a2bbb
commit 3415dc15a3
5 changed files with 30 additions and 28 deletions
Download Patch File Download Diff File Expand all files Collapse all files

17
src/math/common_math.h → src/math/mconfig.h
View File

@@ -1,7 +1,7 @@
// Common math library // Common math library
#ifndef MATH_H #ifndef MCONFIG_H
#define MATH_H #define MCONFIG_H
#include <stdlib.h> #include <stdlib.h>
#include <string.h> #include <string.h>
@@ -10,11 +10,22 @@
#if defined(__x86_64__) || defined(__i386__) #if defined(__x86_64__) || defined(__i386__)
#define SIMD_X86 #define SIMD_X86
#include <xmmintrin.h> #include <xmmintrin.h>
#elif defined(__aarch64__) || defined(__arm64__) #elif defined(__aarch64__) || defined(__arm64__)
#define SIMD_ARCH #define SIMD_ARCH
#include <arm_neon.h> #include <arm_neon.h>
#else #else
#define SIMD_NONE #define SIMD_NONE
#endif #endif
#endif // MATH_H
#ifdef _MSC_VER
#define ALIGN16 __declspec(align(16))
#define RESTRICT __restrict__
#else
#define ALIGN16 __attribute__((aligned(16)))
#define RESTRICT __restrict
#endif
#endif // MCONFIG_H

2
src/math/vec3.c
View File

@@ -6,7 +6,7 @@
// //
#include "vec3.h" #include "vec3.h"
#include "common_math.h" #include "mconfig.h"
#define VEC_SIZE 3 #define VEC_SIZE 3

16
src/math/vec3.h
View File

@@ -8,11 +8,7 @@
#ifndef vec3_h #ifndef vec3_h
#define vec3_h #define vec3_h
#ifdef _MSC_VER #include "mconfig.h"
#define ALIGN16 __declspec(align(16))
#else
#define ALIGN16 __attribute__((aligned(16)))
#endif
typedef union typedef union
{ {
@@ -20,25 +16,25 @@ typedef union
float data[4]; float data[4];
} ALIGN16 Vec3f_t; } ALIGN16 Vec3f_t;
Vec3f_t vec3f_from_array(const float *restrict val); Vec3f_t vec3f_from_array(const float *RESTRICT val);
Vec3f_t vec3f(float x, float y, float z); Vec3f_t vec3f(float x, float y, float z);
// (f, f, f) // (f, f, f)
Vec3f_t vec3f_scalar(float f); Vec3f_t vec3f_scalar(float f);
// (0, 0, 0) // (0, 0, 0)
Vec3f_t vec3f_zero(void); Vec3f_t vec3f_zero(void);
inline static Vec3f_t vec3f_clone(const Vec3f_t *restrict v) inline static Vec3f_t vec3f_clone(const Vec3f_t *RESTRICT v)
{ {
return *v; return *v;
} }
Vec3f_t vec3f_add_r(Vec3f_t *restrict out, Vec3f_t a); Vec3f_t vec3f_add_r(Vec3f_t *RESTRICT out, Vec3f_t a);
Vec3f_t vec3f_add(Vec3f_t a, Vec3f_t b); Vec3f_t vec3f_add(Vec3f_t a, Vec3f_t b);
Vec3f_t vec3f_sub_r(Vec3f_t *restrict out, Vec3f_t a); Vec3f_t vec3f_sub_r(Vec3f_t *RESTRICT out, Vec3f_t a);
Vec3f_t vec3f_sub(Vec3f_t a, Vec3f_t b); Vec3f_t vec3f_sub(Vec3f_t a, Vec3f_t b);
Vec3f_t vec3f_scale_r(Vec3f_t *restrict out, float scale); Vec3f_t vec3f_scale_r(Vec3f_t *RESTRICT out, float scale);
Vec3f_t vec3f_scale(Vec3f_t a, float scale); Vec3f_t vec3f_scale(Vec3f_t a, float scale);
#endif /* vec3_h */ #endif /* vec3_h */

5
src/math/vec4.c
View File

@@ -1,5 +1,4 @@
#include "vec4.h" #include "vec4.h"
#include "common_math.h"
#define VEC_SIZE 4 #define VEC_SIZE 4
@@ -89,8 +88,8 @@ Vec4f_t vec4f_sub_r(Vec4f_t *restrict out, Vec4f_t a)
_mm_store_ps(out->data, vres); _mm_store_ps(out->data, vres);
#elif defined (SIMD_ARCH) #elif defined (SIMD_ARCH)
float32x4_t va = vld1q_f32(a.data); float32x4_t va = vld1q_f32(out->data);
float32x4_t vb = vld1q_f32(out->data); float32x4_t vb = vld1q_f32(a.data);
float32x4_t vres = vsubq_f32(va, vb); float32x4_t vres = vsubq_f32(va, vb);
vst1q_f32(out->data, vres); vst1q_f32(out->data, vres);

18
src/math/vec4.h
View File

@@ -1,12 +1,7 @@
#ifndef VECTOR4_H #ifndef VECTOR4_H
#define VECTOR4_H #define VECTOR4_H
#include "vec3.h" #include "mconfig.h"
#ifdef _MSC_VER
#define ALIGN16 __declspec(align(16))
#else
#define ALIGN16 __attribute__((aligned(16)))
#endif
// must be aligned by 16 Bytes (less instruction executed for SSE) // must be aligned by 16 Bytes (less instruction executed for SSE)
typedef union typedef union
@@ -15,24 +10,25 @@ typedef union
float data[4]; float data[4];
} ALIGN16 Vec4f_t; } ALIGN16 Vec4f_t;
Vec4f_t vec4f_from_array(const float *restrict val); Vec4f_t vec4f_from_array(const float *RESTRICT val);
Vec4f_t vec4f(float x, float y, float z, float w); Vec4f_t vec4f(float x, float y, float z, float w);
// (f, f, f, f) // (f, f, f, f)
Vec4f_t vec4f_scalar(float f); Vec4f_t vec4f_scalar(float f);
// (0, 0, 0, 0) // (0, 0, 0, 0)
Vec4f_t vec4f_zero(void); Vec4f_t vec4f_zero(void);
inline static Vec4f_t vec4f_clone(const Vec4f_t *restrict v) inline static Vec4f_t vec4f_clone(const Vec4f_t *RESTRICT v)
{ {
return *v; return *v;
} }
Vec4f_t vec4f_add_r(Vec4f_t *restrict out, Vec4f_t a);
Vec4f_t vec4f_add_r(Vec4f_t *RESTRICT out, Vec4f_t a);
Vec4f_t vec4f_add(Vec4f_t a, Vec4f_t b); Vec4f_t vec4f_add(Vec4f_t a, Vec4f_t b);
Vec4f_t vec4f_sub_r(Vec4f_t *restrict out, Vec4f_t a); Vec4f_t vec4f_sub_r(Vec4f_t *RESTRICT out, Vec4f_t a);
Vec4f_t vec4f_sub(Vec4f_t a, Vec4f_t b); Vec4f_t vec4f_sub(Vec4f_t a, Vec4f_t b);
Vec4f_t vec4f_scale_r(Vec4f_t *restrict out, float scale); Vec4f_t vec4f_scale_r(Vec4f_t *RESTRICT out, float scale);
Vec4f_t vec4f_scale(Vec4f_t a, float scale); Vec4f_t vec4f_scale(Vec4f_t a, float scale);
float vec4_dot(Vec4f_t a, Vec4f_t b); float vec4_dot(Vec4f_t a, Vec4f_t b);