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

jack: improve mix function

Browse source 
Avoid many reads and writes to temp buffer.
This commit is contained in:
Wim Taymans 2022-07-10 21:59:05 +02:00
parent 7a0445cb28
commit 23984f8790
1 changed files with 44 additions and 35 deletions
Download patch file Download diff file Expand all files Collapse all files

79
pipewire-jack/src/pipewire-jack.c
View file

@ -67,7 +67,7 @@
#define JACK_PORT_TYPE_SIZE 32 #define JACK_PORT_TYPE_SIZE 32
#define MONITOR_EXT " Monitor" #define MONITOR_EXT " Monitor"
#define MAX_MIDI_MIX 1024 #define MAX_MIX 1024
#define MAX_BUFFER_FRAMES 8192 #define MAX_BUFFER_FRAMES 8192
#define MAX_ALIGN 16 #define MAX_ALIGN 16
@ -106,9 +106,9 @@ static bool mlock_warned = false;
#define OBJECT_CHUNK 8 #define OBJECT_CHUNK 8
#define RECYCLE_THRESHOLD 128 #define RECYCLE_THRESHOLD 128
typedef void (*mix2_func) (float *dst, float *src1, float *src2, int n_samples); typedef void (*mix_func) (float *dst, float *src[], uint32_t n_src, bool aligned, uint32_t n_samples);
static mix2_func mix2; static mix_func mix_function;
struct object { struct object {
struct spa_list link; struct spa_list link;
@ -736,38 +736,40 @@ static struct buffer *dequeue_buffer(struct client *c, struct mix *mix)
#if defined (__SSE__) #if defined (__SSE__)
#include <xmmintrin.h> #include <xmmintrin.h>
static void mix2_sse(float *dst, float *src1, float *src2, int n_samples) static void mix_sse(float *dst, float *src[], uint32_t n_src, bool aligned, uint32_t n_samples)
{ {
int n, unrolled; uint32_t i, n, unrolled;
__m128 in[2]; __m128 in[1];
if (SPA_IS_ALIGNED(src1, 16) && if (SPA_IS_ALIGNED(dst, 16) && aligned)
SPA_IS_ALIGNED(src2, 16) && unrolled = n_samples & ~3;
SPA_IS_ALIGNED(dst, 16))
unrolled = n_samples / 4;
else else
unrolled = 0; unrolled = 0;
for (n = 0; unrolled--; n += 4) { for (n = 0; n < unrolled; n += 4) {
in[0] = _mm_load_ps(&src1[n]), in[0] = _mm_load_ps(&src[0][n]);
in[1] = _mm_load_ps(&src2[n]), for (i = 1; i < n_src; i++)
in[0] = _mm_add_ps(in[0], in[1]); in[0] = _mm_add_ps(in[0], _mm_load_ps(&src[i][n]));
_mm_store_ps(&dst[n], in[0]); _mm_store_ps(&dst[n], in[0]);
} }
for (; n < n_samples; n++) { for (; n < n_samples; n++) {
in[0] = _mm_load_ss(&src1[n]), in[0] = _mm_load_ss(&src[0][n]);
in[1] = _mm_load_ss(&src2[n]), for (i = 1; i < n_src; i++)
in[0] = _mm_add_ss(in[0], in[1]); in[0] = _mm_add_ss(in[0], _mm_load_ss(&src[i][n]));
_mm_store_ss(&dst[n], in[0]); _mm_store_ss(&dst[n], in[0]);
} }
} }
#endif #endif
static void mix2_c(float *dst, float *src1, float *src2, int n_samples) static void mix_c(float *dst, float *src[], uint32_t n_src, bool aligned, uint32_t n_samples)
{ {
int i; uint32_t n, i;
for (i = 0; i < n_samples; i++) for (n = 0; n < n_samples; n++) {
dst[i] = src1[i] + src2[i]; float t = src[0][n];
for (i = 1; i < n_src; i++)
t += src[i][n];
dst[n] = t;
}
} }
SPA_EXPORT SPA_EXPORT
@ -3281,13 +3283,13 @@ jack_client_t * jack_client_open (const char *client_name,
support = pw_context_get_support(client->context.context, &n_support); support = pw_context_get_support(client->context.context, &n_support);
mix2 = mix2_c; mix_function = mix_c;
cpu_iface = spa_support_find(support, n_support, SPA_TYPE_INTERFACE_CPU); cpu_iface = spa_support_find(support, n_support, SPA_TYPE_INTERFACE_CPU);
if (cpu_iface) { if (cpu_iface) {
#if defined (__SSE__) #if defined (__SSE__)
uint32_t flags = spa_cpu_get_flags(cpu_iface); uint32_t flags = spa_cpu_get_flags(cpu_iface);
if (flags & SPA_CPU_FLAG_SSE) if (flags & SPA_CPU_FLAG_SSE)
mix2 = mix2_sse; mix_function = mix_sse;
#endif #endif
} }
client->context.old_thread_utils = client->context.old_thread_utils =
@ -4416,13 +4418,14 @@ static void *get_buffer_input_float(struct port *p, jack_nframes_t frames)
{ {
struct mix *mix; struct mix *mix;
struct buffer *b; struct buffer *b;
int layer = 0;
void *ptr = NULL; void *ptr = NULL;
float *mix_ptr[MAX_MIX], *np;
uint32_t n_ptr = 0;
bool ptr_aligned = true;
spa_list_for_each(mix, &p->mix, port_link) { spa_list_for_each(mix, &p->mix, port_link) {
struct spa_data *d; struct spa_data *d;
uint32_t offset, size; uint32_t offset, size;
void *np;
pw_log_trace_fp("%p: port %s mix %d.%d get buffer %d", pw_log_trace_fp("%p: port %s mix %d.%d get buffer %d",
p->client, p->object->port.name, p->port_id, mix->id, frames); p->client, p->object->port.name, p->port_id, mix->id, frames);
@ -4436,14 +4439,20 @@ static void *get_buffer_input_float(struct port *p, jack_nframes_t frames)
if (size / sizeof(float) < frames) if (size / sizeof(float) < frames)
continue; continue;
np = SPA_PTROFF(d->data, offset, void); np = SPA_PTROFF(d->data, offset, float);
if (layer++ == 0) { if (!SPA_IS_ALIGNED(np, 16))
ptr = np; ptr_aligned = false;
} else {
mix2(p->emptyptr, ptr, np, frames); mix_ptr[n_ptr++] = np;
ptr = p->emptyptr; if (n_ptr == MAX_MIX)
p->zeroed = false; break;
} }
if (n_ptr == 1) {
ptr = mix_ptr[0];
} else {
ptr = p->emptyptr;
mix_function(ptr, mix_ptr, n_ptr, ptr_aligned, frames);
p->zeroed = false;
} }
if (ptr == NULL) if (ptr == NULL)
ptr = init_buffer(p); ptr = init_buffer(p);
@ -4454,7 +4463,7 @@ static void *get_buffer_input_midi(struct port *p, jack_nframes_t frames)
{ {
struct mix *mix; struct mix *mix;
void *ptr = p->emptyptr; void *ptr = p->emptyptr;
struct spa_pod_sequence *seq[MAX_MIDI_MIX]; struct spa_pod_sequence *seq[MAX_MIX];
uint32_t n_seq = 0; uint32_t n_seq = 0;
jack_midi_clear_buffer(ptr); jack_midi_clear_buffer(ptr);
@ -4478,7 +4487,7 @@ static void *get_buffer_input_midi(struct port *p, jack_nframes_t frames)
continue; continue;
seq[n_seq++] = pod; seq[n_seq++] = pod;
if (n_seq == MAX_MIDI_MIX) if (n_seq == MAX_MIX)
break; break;
} }
convert_to_midi(seq, n_seq, ptr, p->client->fix_midi_events); convert_to_midi(seq, n_seq, ptr, p->client->fix_midi_events);