YourWishes/cglm
1
0
Fork 0
mirror of https://github.com/recp/cglm.git synced 2026-02-17 03:39:05 +00:00
Code Issues Packages Projects Releases Wiki Activity

re-organise files, remove cglm prefix from file to make them more clean

Browse Source
This commit is contained in:
Recep Aslantas
2017-07-11 18:05:10 +03:00
parent 3b159cdf02
commit 9d61668e04
49 changed files with 977 additions and 847 deletions
Download Patch File Download Diff File Expand all files Collapse all files

79
include/cglm/simd/sse2/affine.h Normal file
View File

@@ -0,0 +1,79 @@
/*
* Copyright (c), Recep Aslantas.
*
* MIT License (MIT), http://opensource.org/licenses/MIT
* Full license can be found in the LICENSE file
*/
#ifndef cglm_affine_mat_sse2_h
#define cglm_affine_mat_sse2_h
#if defined( __SSE__ ) || defined( __SSE2__ )
#include "../../common.h"
#include "../intrin.h"
CGLM_INLINE
void
glm_mul_sse2(mat4 m1, mat4 m2, mat4 dest) {
/* D = R * L (Column-Major) */
__m128 l0, l1, l2, l3, r;
l0 = _mm_load_ps(m1[0]);
l1 = _mm_load_ps(m1[1]);
l2 = _mm_load_ps(m1[2]);
l3 = _mm_load_ps(m1[3]);
r = _mm_load_ps(m2[0]);
_mm_store_ps(dest[0],
_mm_add_ps(_mm_add_ps(_mm_mul_ps(_mm_shuffle1_ps1(r, 0), l0),
_mm_mul_ps(_mm_shuffle1_ps1(r, 1), l1)),
_mm_mul_ps(_mm_shuffle1_ps1(r, 2), l2)));
r = _mm_load_ps(m2[1]);
_mm_store_ps(dest[1],
_mm_add_ps(_mm_add_ps(_mm_mul_ps(_mm_shuffle1_ps1(r, 0), l0),
_mm_mul_ps(_mm_shuffle1_ps1(r, 1), l1)),
_mm_mul_ps(_mm_shuffle1_ps1(r, 2), l2)));
r = _mm_load_ps(m2[2]);
_mm_store_ps(dest[2],
_mm_add_ps(_mm_add_ps(_mm_mul_ps(_mm_shuffle1_ps1(r, 0), l0),
_mm_mul_ps(_mm_shuffle1_ps1(r, 1), l1)),
_mm_mul_ps(_mm_shuffle1_ps1(r, 2), l2)));
r = _mm_load_ps(m2[3]);
_mm_store_ps(dest[3],
_mm_add_ps(_mm_add_ps(_mm_mul_ps(_mm_shuffle1_ps1(r, 0), l0),
_mm_mul_ps(_mm_shuffle1_ps1(r, 1), l1)),
_mm_add_ps(_mm_mul_ps(_mm_shuffle1_ps1(r, 2), l2),
_mm_mul_ps(_mm_shuffle1_ps1(r, 3), l3))));
}
CGLM_INLINE
void
glm_inv_tr_sse2(mat4 mat) {
__m128 r0, r1, r2, r3, x0, x1;
r0 = _mm_load_ps(mat[0]);
r1 = _mm_load_ps(mat[1]);
r2 = _mm_load_ps(mat[2]);
r3 = _mm_load_ps(mat[3]);
x1 = _mm_set_ps(1.0f, 0.0f, 0.0f, 0.0f);
_MM_TRANSPOSE4_PS(r0, r1, r2, x1);
x0 = _mm_add_ps(_mm_mul_ps(r0, _mm_shuffle1_ps(r3, 0, 0, 0, 0)),
_mm_mul_ps(r1, _mm_shuffle1_ps(r3, 1, 1, 1, 1)));
x0 = _mm_add_ps(x0, _mm_mul_ps(r2, _mm_shuffle1_ps(r3, 2, 2, 2, 2)));
x0 = _mm_xor_ps(x0, _mm_set1_ps(-0.f));
x0 = _mm_add_ps(x0, x1);
_mm_store_ps(mat[0], r0);
_mm_store_ps(mat[1], r1);
_mm_store_ps(mat[2], r2);
_mm_store_ps(mat[3], x0);
}
#endif
#endif /* cglm_affine_mat_sse2_h */