External/cglm
1
0
Fork 0
mirror of https://github.com/recp/cglm.git synced 2025-10-04 09:08:53 +00:00
Code Issues Packages Projects Releases Wiki Activity

avx: replace binary constants with hex

Browse Source
This commit is contained in:
Recep Aslantas
2018-10-12 09:05:42 +03:00
parent aa2b0f2631
commit 20360f2296
3 changed files with 9 additions and 7 deletions
Download Patch File Download Diff File Expand all files Collapse all files

7
include/cglm/simd/avx/affine.h
View File

@@ -27,13 +27,14 @@ glm_mul_avx(mat4 m1, mat4 m2, mat4 dest) {
y2 = glmm_load256(m1[0]); /* h g f e d c b a */ y2 = glmm_load256(m1[0]); /* h g f e d c b a */
y3 = glmm_load256(m1[2]); /* p o n m l k j i */ y3 = glmm_load256(m1[2]); /* p o n m l k j i */
y4 = _mm256_permute2f128_ps(y2, y2, 0b00000011); /* d c b a h g f e */ /* 0x03: 0b00000011 */
y5 = _mm256_permute2f128_ps(y3, y3, 0b00000000); /* l k j i l k j i */ y4 = _mm256_permute2f128_ps(y2, y2, 0x03); /* d c b a h g f e */
y5 = _mm256_permute2f128_ps(y3, y3, 0x03); /* l k j i l k j i */
/* f f f f a a a a */ /* f f f f a a a a */
/* g g g g c c c c */ /* g g g g c c c c */
/* e e e e b b b b */ /* e e e e b b b b */
y7 = _mm256_permute_ps(y0, 0b10101010); y7 = _mm256_permute_ps(y0, 0xAA); /* 0xAA: 0b10101010 */
y6 = _mm256_permutevar_ps(y0, _mm256_set_epi32(1, 1, 1, 1, 0, 0, 0, 0)); y6 = _mm256_permutevar_ps(y0, _mm256_set_epi32(1, 1, 1, 1, 0, 0, 0, 0));
y8 = _mm256_permutevar_ps(y0, _mm256_set_epi32(0, 0, 0, 0, 1, 1, 1, 1)); y8 = _mm256_permutevar_ps(y0, _mm256_set_epi32(0, 0, 0, 0, 1, 1, 1, 1));

5
include/cglm/simd/avx/mat4.h
View File

@@ -27,8 +27,9 @@ glm_mat4_mul_avx(mat4 m1, mat4 m2, mat4 dest) {
y2 = glmm_load256(m1[0]); /* h g f e d c b a */ y2 = glmm_load256(m1[0]); /* h g f e d c b a */
y3 = glmm_load256(m1[2]); /* p o n m l k j i */ y3 = glmm_load256(m1[2]); /* p o n m l k j i */
y4 = _mm256_permute2f128_ps(y2, y2, 0b00000011); /* d c b a h g f e */ /* 0x03: 0b00000011 */
y5 = _mm256_permute2f128_ps(y3, y3, 0b00000011); /* l k j i p o n m */ y4 = _mm256_permute2f128_ps(y2, y2, 0x03); /* d c b a h g f e */
y5 = _mm256_permute2f128_ps(y3, y3, 0x03); /* l k j i p o n m */
/* f f f f a a a a */ /* f f f f a a a a */
/* h h h h c c c c */ /* h h h h c c c c */