orbit-plus/include/cglm/affine-mat.h
2023-02-11 23:18:45 -05:00

179 lines
5.0 KiB
C

/*
* Copyright (c), Recep Aslantas.
*
* MIT License (MIT), http://opensource.org/licenses/MIT
* Full license can be found in the LICENSE file
*/
/*
Functions:
CGLM_INLINE void glm_mul(mat4 m1, mat4 m2, mat4 dest);
CGLM_INLINE void glm_inv_tr(mat4 mat);
*/
#ifndef cglm_affine_mat_h
#define cglm_affine_mat_h
#include "common.h"
#include "mat4.h"
#include "mat3.h"
#ifdef CGLM_SSE_FP
# include "simd/sse2/affine.h"
#endif
#ifdef CGLM_AVX_FP
# include "simd/avx/affine.h"
#endif
#ifdef CGLM_NEON_FP
# include "simd/neon/affine.h"
#endif
/*!
* @brief this is similar to glm_mat4_mul but specialized to affine transform
*
* Matrix format should be:
* R R R X
* R R R Y
* R R R Z
* 0 0 0 W
*
* this reduces some multiplications. It should be faster than mat4_mul.
* if you are not sure about matrix format then DON'T use this! use mat4_mul
*
* @param[in] m1 affine matrix 1
* @param[in] m2 affine matrix 2
* @param[out] dest result matrix
*/
CGLM_INLINE
void
glm_mul(mat4 m1, mat4 m2, mat4 dest) {
#ifdef __AVX__
glm_mul_avx(m1, m2, dest);
#elif defined( __SSE__ ) || defined( __SSE2__ )
glm_mul_sse2(m1, m2, dest);
#elif defined(CGLM_NEON_FP)
glm_mul_neon(m1, m2, dest);
#else
float a00 = m1[0][0], a01 = m1[0][1], a02 = m1[0][2], a03 = m1[0][3],
a10 = m1[1][0], a11 = m1[1][1], a12 = m1[1][2], a13 = m1[1][3],
a20 = m1[2][0], a21 = m1[2][1], a22 = m1[2][2], a23 = m1[2][3],
a30 = m1[3][0], a31 = m1[3][1], a32 = m1[3][2], a33 = m1[3][3],
b00 = m2[0][0], b01 = m2[0][1], b02 = m2[0][2],
b10 = m2[1][0], b11 = m2[1][1], b12 = m2[1][2],
b20 = m2[2][0], b21 = m2[2][1], b22 = m2[2][2],
b30 = m2[3][0], b31 = m2[3][1], b32 = m2[3][2], b33 = m2[3][3];
dest[0][0] = a00 * b00 + a10 * b01 + a20 * b02;
dest[0][1] = a01 * b00 + a11 * b01 + a21 * b02;
dest[0][2] = a02 * b00 + a12 * b01 + a22 * b02;
dest[0][3] = a03 * b00 + a13 * b01 + a23 * b02;
dest[1][0] = a00 * b10 + a10 * b11 + a20 * b12;
dest[1][1] = a01 * b10 + a11 * b11 + a21 * b12;
dest[1][2] = a02 * b10 + a12 * b11 + a22 * b12;
dest[1][3] = a03 * b10 + a13 * b11 + a23 * b12;
dest[2][0] = a00 * b20 + a10 * b21 + a20 * b22;
dest[2][1] = a01 * b20 + a11 * b21 + a21 * b22;
dest[2][2] = a02 * b20 + a12 * b21 + a22 * b22;
dest[2][3] = a03 * b20 + a13 * b21 + a23 * b22;
dest[3][0] = a00 * b30 + a10 * b31 + a20 * b32 + a30 * b33;
dest[3][1] = a01 * b30 + a11 * b31 + a21 * b32 + a31 * b33;
dest[3][2] = a02 * b30 + a12 * b31 + a22 * b32 + a32 * b33;
dest[3][3] = a03 * b30 + a13 * b31 + a23 * b32 + a33 * b33;
#endif
}
/*!
* @brief this is similar to glm_mat4_mul but specialized to affine transform
*
* Right Matrix format should be:
* R R R 0
* R R R 0
* R R R 0
* 0 0 0 1
*
* this reduces some multiplications. It should be faster than mat4_mul.
* if you are not sure about matrix format then DON'T use this! use mat4_mul
*
* @param[in] m1 affine matrix 1
* @param[in] m2 affine matrix 2
* @param[out] dest result matrix
*/
CGLM_INLINE
void
glm_mul_rot(mat4 m1, mat4 m2, mat4 dest) {
#if defined( __SSE__ ) || defined( __SSE2__ )
glm_mul_rot_sse2(m1, m2, dest);
#elif defined(CGLM_NEON_FP)
glm_mul_rot_neon(m1, m2, dest);
#else
float a00 = m1[0][0], a01 = m1[0][1], a02 = m1[0][2], a03 = m1[0][3],
a10 = m1[1][0], a11 = m1[1][1], a12 = m1[1][2], a13 = m1[1][3],
a20 = m1[2][0], a21 = m1[2][1], a22 = m1[2][2], a23 = m1[2][3],
a30 = m1[3][0], a31 = m1[3][1], a32 = m1[3][2], a33 = m1[3][3],
b00 = m2[0][0], b01 = m2[0][1], b02 = m2[0][2],
b10 = m2[1][0], b11 = m2[1][1], b12 = m2[1][2],
b20 = m2[2][0], b21 = m2[2][1], b22 = m2[2][2];
dest[0][0] = a00 * b00 + a10 * b01 + a20 * b02;
dest[0][1] = a01 * b00 + a11 * b01 + a21 * b02;
dest[0][2] = a02 * b00 + a12 * b01 + a22 * b02;
dest[0][3] = a03 * b00 + a13 * b01 + a23 * b02;
dest[1][0] = a00 * b10 + a10 * b11 + a20 * b12;
dest[1][1] = a01 * b10 + a11 * b11 + a21 * b12;
dest[1][2] = a02 * b10 + a12 * b11 + a22 * b12;
dest[1][3] = a03 * b10 + a13 * b11 + a23 * b12;
dest[2][0] = a00 * b20 + a10 * b21 + a20 * b22;
dest[2][1] = a01 * b20 + a11 * b21 + a21 * b22;
dest[2][2] = a02 * b20 + a12 * b21 + a22 * b22;
dest[2][3] = a03 * b20 + a13 * b21 + a23 * b22;
dest[3][0] = a30;
dest[3][1] = a31;
dest[3][2] = a32;
dest[3][3] = a33;
#endif
}
/*!
* @brief inverse orthonormal rotation + translation matrix (ridig-body)
*
* @code
* X = | R T | X' = | R' -R'T |
* | 0 1 | | 0 1 |
* @endcode
*
* @param[in,out] mat matrix
*/
CGLM_INLINE
void
glm_inv_tr(mat4 mat) {
#if defined( __SSE__ ) || defined( __SSE2__ )
glm_inv_tr_sse2(mat);
#elif defined(CGLM_NEON_FP)
glm_inv_tr_neon(mat);
#else
CGLM_ALIGN_MAT mat3 r;
CGLM_ALIGN(8) vec3 t;
/* rotate */
glm_mat4_pick3t(mat, r);
glm_mat4_ins3(r, mat);
/* translate */
glm_mat3_mulv(r, mat[3], t);
glm_vec3_negate(t);
glm_vec3_copy(t, mat[3]);
#endif
}
#endif /* cglm_affine_mat_h */