mirrors/pipewire
1
0
Fork 0
mirror of https://gitlab.freedesktop.org/pipewire/pipewire.git synced 2026-05-24 21:38:05 -04:00
Code Issues Projects Releases Packages Wiki Activity Actions

channelmix: Convert matrices to float arrays

Browse source 
Convert matrix_orig and matrix to float arrays and use variable size 2d
arrays to access the elements of the matrices. This removes the need for
storing pointers to matrix rows.
This commit is contained in:
Jonas Holmberg 2026-05-12 14:47:57 +02:00
parent 7cfcb46fbf
commit a7994882b9
6 changed files with 161 additions and 134 deletions
Download patch file Download diff file Expand all files Collapse all files

92
spa/plugins/audioconvert/channelmix-ops-sse.c
View file

@ -144,11 +144,12 @@ static inline void sub_sse(float *d, const float *s0, const float *s1, uint32_t
void channelmix_copy_sse(struct channelmix *mix, void * SPA_RESTRICT dst[],
const void * SPA_RESTRICT src[], uint32_t n_samples)
{
CHANNELMIX_DEF_MATRIX(matrix, mix, matrix);
uint32_t i, n_dst = mix->dst_chan;
float **d = (float **)dst;
const float **s = (const float **)src;
for (i = 0; i < n_dst; i++)
vol_sse(d[i], s[i], mix->matrix[i][i], n_samples);
vol_sse(d[i], s[i], matrix[i][i], n_samples);
}
static void lr4_process_sse(struct lr4 *lr4, float *dst, const float *src, const float vol, int samples)
@ -346,6 +347,7 @@ void
channelmix_f32_n_m_sse(struct channelmix *mix, void * SPA_RESTRICT dst[],
const void * SPA_RESTRICT src[], uint32_t n_samples)
{
CHANNELMIX_DEF_MATRIX(matrix, mix, matrix);
float **d = (float **) dst;
const float **s = (const float **) src;
uint32_t i, j, n_dst = mix->dst_chan, n_src = mix->src_chan;
@ -357,9 +359,9 @@ channelmix_f32_n_m_sse(struct channelmix *mix, void * SPA_RESTRICT dst[],
uint32_t n_j = 0;
for (j = 0; j < n_src; j++) {
if (mix->matrix[i][j] == 0.0f)
if (matrix[i][j] == 0.0f)
continue;
mj[n_j] = mix->matrix[i][j];
mj[n_j] = matrix[i][j];
sj[n_j++] = s[j];
}
if (n_j == 0) {
@ -377,13 +379,14 @@ void
channelmix_f32_2_3p1_sse(struct channelmix *mix, void * SPA_RESTRICT dst[],
const void * SPA_RESTRICT src[], uint32_t n_samples)
{
CHANNELMIX_DEF_MATRIX(matrix, mix, matrix);
uint32_t i, n, unrolled, n_dst = mix->dst_chan;
float **d = (float **)dst;
const float **s = (const float **)src;
const float v0 = mix->matrix[0][0];
const float v1 = mix->matrix[1][1];
const float v2 = (mix->matrix[2][0] + mix->matrix[2][1]) * 0.5f;
const float v3 = (mix->matrix[3][0] + mix->matrix[3][1]) * 0.5f;
const float v0 = matrix[0][0];
const float v1 = matrix[1][1];
const float v2 = (matrix[2][0] + matrix[2][1]) * 0.5f;
const float v3 = (matrix[3][0] + matrix[3][1]) * 0.5f;
if (SPA_FLAG_IS_SET(mix->flags, CHANNELMIX_FLAG_ZERO)) {
for (i = 0; i < n_dst; i++)
@ -395,8 +398,8 @@ channelmix_f32_2_3p1_sse(struct channelmix *mix, void * SPA_RESTRICT dst[],
vol_sse(d[1], s[1], v1, n_samples);
avg_sse(d[2], s[0], s[1], n_samples);
} else {
const __m128 mv0 = _mm_set1_ps(mix->matrix[0][0]);
const __m128 mv1 = _mm_set1_ps(mix->matrix[1][1]);
const __m128 mv0 = _mm_set1_ps(matrix[0][0]);
const __m128 mv1 = _mm_set1_ps(matrix[1][1]);
const __m128 mw = _mm_set1_ps(mix->widen);
const __m128 mh = _mm_set1_ps(0.5f);
__m128 t0[1], t1[1], w[1], c[1];
@ -437,11 +440,12 @@ void
channelmix_f32_2_5p1_sse(struct channelmix *mix, void * SPA_RESTRICT dst[],
const void * SPA_RESTRICT src[], uint32_t n_samples)
{
CHANNELMIX_DEF_MATRIX(matrix, mix, matrix);
uint32_t i, n_dst = mix->dst_chan;
float **d = (float **)dst;
const float **s = (const float **)src;
const float v4 = mix->matrix[4][0];
const float v5 = mix->matrix[5][1];
const float v4 = matrix[4][0];
const float v5 = matrix[5][1];
if (SPA_FLAG_IS_SET(mix->flags, CHANNELMIX_FLAG_ZERO)) {
for (i = 0; i < n_dst; i++)
@ -468,13 +472,14 @@ void
channelmix_f32_2_7p1_sse(struct channelmix *mix, void * SPA_RESTRICT dst[],
const void * SPA_RESTRICT src[], uint32_t n_samples)
{
CHANNELMIX_DEF_MATRIX(matrix, mix, matrix);
uint32_t i, n_dst = mix->dst_chan;
float **d = (float **)dst;
const float **s = (const float **)src;
const float v4 = mix->matrix[4][0];
const float v5 = mix->matrix[5][1];
const float v6 = mix->matrix[6][0];
const float v7 = mix->matrix[7][1];
const float v4 = matrix[4][0];
const float v5 = matrix[5][1];
const float v6 = matrix[6][0];
const float v7 = matrix[7][1];
if (SPA_FLAG_IS_SET(mix->flags, CHANNELMIX_FLAG_ZERO)) {
for (i = 0; i < n_dst; i++)
@ -504,12 +509,13 @@ void
channelmix_f32_3p1_2_sse(struct channelmix *mix, void * SPA_RESTRICT dst[],
const void * SPA_RESTRICT src[], uint32_t n_samples)
{
CHANNELMIX_DEF_MATRIX(matrix, mix, matrix);
float **d = (float **) dst;
const float **s = (const float **) src;
const float m0 = mix->matrix[0][0];
const float m1 = mix->matrix[1][1];
const float m2 = (mix->matrix[0][2] + mix->matrix[1][2]) * 0.5f;
const float m3 = (mix->matrix[0][3] + mix->matrix[1][3]) * 0.5f;
const float m0 = matrix[0][0];
const float m1 = matrix[1][1];
const float m2 = (matrix[0][2] + matrix[1][2]) * 0.5f;
const float m3 = (matrix[0][3] + matrix[1][3]) * 0.5f;
if (m0 == 0.0f && m1 == 0.0f && m2 == 0.0f && m3 == 0.0f) {
clear_sse(d[0], n_samples);
@ -554,15 +560,16 @@ void
channelmix_f32_5p1_2_sse(struct channelmix *mix, void * SPA_RESTRICT dst[],
const void * SPA_RESTRICT src[], uint32_t n_samples)
{
CHANNELMIX_DEF_MATRIX(matrix, mix, matrix);
uint32_t n, unrolled;
float **d = (float **) dst;
const float **s = (const float **) src;
const float m00 = mix->matrix[0][0];
const float m11 = mix->matrix[1][1];
const __m128 clev = _mm_set1_ps((mix->matrix[0][2] + mix->matrix[1][2]) * 0.5f);
const __m128 llev = _mm_set1_ps((mix->matrix[0][3] + mix->matrix[1][3]) * 0.5f);
const __m128 slev0 = _mm_set1_ps(mix->matrix[0][4]);
const __m128 slev1 = _mm_set1_ps(mix->matrix[1][5]);
const float m00 = matrix[0][0];
const float m11 = matrix[1][1];
const __m128 clev = _mm_set1_ps((matrix[0][2] + matrix[1][2]) * 0.5f);
const __m128 llev = _mm_set1_ps((matrix[0][3] + matrix[1][3]) * 0.5f);
const __m128 slev0 = _mm_set1_ps(matrix[0][4]);
const __m128 slev1 = _mm_set1_ps(matrix[1][5]);
__m128 in, ctr;
if (SPA_IS_ALIGNED(s[0], 16) &&
@ -616,6 +623,7 @@ void
channelmix_f32_5p1_3p1_sse(struct channelmix *mix, void * SPA_RESTRICT dst[],
const void * SPA_RESTRICT src[], uint32_t n_samples)
{
CHANNELMIX_DEF_MATRIX(matrix, mix, matrix);
uint32_t i, n, unrolled, n_dst = mix->dst_chan;
float **d = (float **) dst;
const float **s = (const float **) src;
@ -639,10 +647,10 @@ channelmix_f32_5p1_3p1_sse(struct channelmix *mix, void * SPA_RESTRICT dst[],
clear_sse(d[i], n_samples);
}
else {
const __m128 v0 = _mm_set1_ps(mix->matrix[0][0]);
const __m128 v1 = _mm_set1_ps(mix->matrix[1][1]);
const __m128 slev0 = _mm_set1_ps(mix->matrix[0][4]);
const __m128 slev1 = _mm_set1_ps(mix->matrix[1][5]);
const __m128 v0 = _mm_set1_ps(matrix[0][0]);
const __m128 v1 = _mm_set1_ps(matrix[1][1]);
const __m128 slev0 = _mm_set1_ps(matrix[0][4]);
const __m128 slev1 = _mm_set1_ps(matrix[1][5]);
for(n = 0; n < unrolled; n += 4) {
_mm_store_ps(&d[0][n], _mm_add_ps(
@ -662,8 +670,8 @@ channelmix_f32_5p1_3p1_sse(struct channelmix *mix, void * SPA_RESTRICT dst[],
_mm_mul_ss(_mm_load_ss(&s[1][n]), v1),
_mm_mul_ss(_mm_load_ss(&s[5][n]), slev1)));
}
vol_sse(d[2], s[2], mix->matrix[2][2], n_samples);
vol_sse(d[3], s[3], mix->matrix[3][3], n_samples);
vol_sse(d[2], s[2], matrix[2][2], n_samples);
vol_sse(d[3], s[3], matrix[3][3], n_samples);
}
}
@ -672,11 +680,12 @@ void
channelmix_f32_5p1_4_sse(struct channelmix *mix, void * SPA_RESTRICT dst[],
const void * SPA_RESTRICT src[], uint32_t n_samples)
{
CHANNELMIX_DEF_MATRIX(matrix, mix, matrix);
uint32_t i, n_dst = mix->dst_chan;
float **d = (float **) dst;
const float **s = (const float **) src;
const float v4 = mix->matrix[2][4];
const float v5 = mix->matrix[3][5];
const float v4 = matrix[2][4];
const float v5 = matrix[3][5];
if (SPA_FLAG_IS_SET(mix->flags, CHANNELMIX_FLAG_ZERO)) {
for (i = 0; i < n_dst; i++)
@ -695,17 +704,18 @@ void
channelmix_f32_7p1_2_sse(struct channelmix *mix, void * SPA_RESTRICT dst[],
const void * SPA_RESTRICT src[], uint32_t n_samples)
{
CHANNELMIX_DEF_MATRIX(matrix, mix, matrix);
uint32_t n, unrolled;
float **d = (float **) dst;
const float **s = (const float **) src;
const __m128 v0 = _mm_set1_ps(mix->matrix[0][0]);
const __m128 v1 = _mm_set1_ps(mix->matrix[1][1]);
const __m128 clev = _mm_set1_ps((mix->matrix[0][2] + mix->matrix[1][2]) * 0.5f);
const __m128 llev = _mm_set1_ps((mix->matrix[0][3] + mix->matrix[1][3]) * 0.5f);
const __m128 slev0 = _mm_set1_ps(mix->matrix[0][4]);
const __m128 slev1 = _mm_set1_ps(mix->matrix[1][5]);
const __m128 rlev0 = _mm_set1_ps(mix->matrix[0][6]);
const __m128 rlev1 = _mm_set1_ps(mix->matrix[1][7]);
const __m128 v0 = _mm_set1_ps(matrix[0][0]);
const __m128 v1 = _mm_set1_ps(matrix[1][1]);
const __m128 clev = _mm_set1_ps((matrix[0][2] + matrix[1][2]) * 0.5f);
const __m128 llev = _mm_set1_ps((matrix[0][3] + matrix[1][3]) * 0.5f);
const __m128 slev0 = _mm_set1_ps(matrix[0][4]);
const __m128 slev1 = _mm_set1_ps(matrix[1][5]);
const __m128 rlev0 = _mm_set1_ps(matrix[0][6]);
const __m128 rlev1 = _mm_set1_ps(matrix[1][7]);
__m128 in, ctr;
if (SPA_IS_ALIGNED(s[0], 16) &&