mirrors/pipewire
1
0
Fork 0
mirror of https://gitlab.freedesktop.org/pipewire/pipewire.git synced 2025-10-31 22:25:38 -04:00
Code Issues Projects Releases Packages Wiki Activity Actions

fmt: unrolled loop only works for stereo

Browse source
This commit is contained in:
Wim Taymans 2018-08-21 10:02:16 +02:00
parent 898bb4b059
commit e6977fa178
1 changed files with 21 additions and 19 deletions
Download patch file Download diff file Expand all files Collapse all files

40
spa/plugins/audioconvert/fmt-ops-sse.c
View file

@ -47,31 +47,33 @@ conv_s16_to_f32d_2_sse(void *data, int n_dst, void *dst[n_dst], const void *src,
{
const int16_t *s = src;
float **d = (float **) dst;
int n, n_samples, unrolled;
int n = 0, n_samples, unrolled;
__m128i in, t[2];
__m128 out[2], factor = _mm_set1_ps(1.0f / S16_SCALE);
n_samples = n_bytes / (sizeof(int16_t) * n_dst);
unrolled = n_samples / 4;
n_samples = n_samples & 3;
if (n_dst == 2) {
unrolled = n_samples / 4;
n_samples = n_samples & 3;
for(n = 0; unrolled--; n += 4) {
in = _mm_loadu_si128((__m128i*)s);
for(; unrolled--; n += 4) {
in = _mm_loadu_si128((__m128i*)s);
t[0] = _mm_slli_epi32(in, 16);
t[0] = _mm_srai_epi32(t[0], 16);
t[1] = _mm_srai_epi32(in, 16);
t[0] = _mm_slli_epi32(in, 16);
t[0] = _mm_srai_epi32(t[0], 16);
t[1] = _mm_srai_epi32(in, 16);
out[0] = _mm_cvtepi32_ps(t[0]);
out[0] = _mm_mul_ps(out[0], factor);
out[1] = _mm_cvtepi32_ps(t[1]);
out[1] = _mm_mul_ps(out[1], factor);
out[0] = _mm_cvtepi32_ps(t[0]);
out[0] = _mm_mul_ps(out[0], factor);
out[1] = _mm_cvtepi32_ps(t[1]);
out[1] = _mm_mul_ps(out[1], factor);
_mm_storeu_ps(&d[0][n], out[0]);
_mm_storeu_ps(&d[1][n], out[1]);
_mm_storeu_ps(&d[0][n], out[0]);
_mm_storeu_ps(&d[1][n], out[1]);
s += 4*n_dst;
s += 4*n_dst;
}
}
for(; n_samples--; n++) {
out[0] = _mm_cvtsi32_ss(out[0], s[0]);
@ -121,10 +123,10 @@ conv_f32d_to_s32_1_sse(void *data, void *dst, int n_src, const void *src[n_src],
out[2] = _mm_shuffle_epi32(out[0], _MM_SHUFFLE(1, 0, 3, 2));
out[3] = _mm_shuffle_epi32(out[0], _MM_SHUFFLE(2, 1, 0, 3));
*(d + 0*n_src) = _mm_cvtsi128_si32(out[0]);
*(d + 1*n_src) = _mm_cvtsi128_si32(out[1]);
*(d + 2*n_src) = _mm_cvtsi128_si32(out[2]);
*(d + 3*n_src) = _mm_cvtsi128_si32(out[3]);
d[0*n_src] = _mm_cvtsi128_si32(out[0]);
d[1*n_src] = _mm_cvtsi128_si32(out[1]);
d[2*n_src] = _mm_cvtsi128_si32(out[2]);
d[3*n_src] = _mm_cvtsi128_si32(out[3]);
d += 4*n_src;
}
for(; n_samples--; n++) {