libliftoff/example/multi-output.c
Simon Ser ceb4a1ff9e
Replace liftoff_device_apply with liftoff_output_apply
Compositors need to drive multiple connectors, each with its own vblank timings.
For each device, there's one separate rendering loop per output.

It's not possible to call liftoff_device_apply each time we want to submit a new
frame to one output, because this could touch another's output state, submitting
a new frame there in the process. When the other output will submit a new frame,
it'll get EBUSY (can't submit two frames without waiting for vblank).

Closes: https://github.com/emersion/libliftoff/issues/21
2019-11-24 18:13:40 +01:00

187 lines
4.7 KiB
C

/* Multiple outputs: create a few layers on each output and display as many of
* them as possible. Layers that don't make it into a plane won't be dispayed.
* Demonstrates how the library distributes planes across CRTCs. */
#define _POSIX_C_SOURCE 200809L
#include <drm_fourcc.h>
#include <fcntl.h>
#include <libliftoff.h>
#include <stdio.h>
#include <string.h>
#include <unistd.h>
#include <xf86drm.h>
#include <xf86drmMode.h>
#include "common.h"
#define MAX_OUTPUTS 32
#define LAYERS_PER_OUTPUT 4
/* ARGB 8:8:8:8 */
static const uint32_t colors[] = {
0xFFFF0000, /* red */
0xFF00FF00, /* green */
0xFF0000FF, /* blue */
0xFFFFFF00, /* yellow */
};
static struct liftoff_layer *add_layer(int drm_fd, struct liftoff_output *output,
int x, int y, int width, int height,
bool with_alpha)
{
static bool first = true;
static size_t color_idx = 0;
struct dumb_fb fb = {0};
uint32_t color;
struct liftoff_layer *layer;
uint32_t format = with_alpha ? DRM_FORMAT_ARGB8888 : DRM_FORMAT_XRGB8888;
if (!dumb_fb_init(&fb, drm_fd, format, width, height)) {
fprintf(stderr, "failed to create framebuffer\n");
return NULL;
}
printf("Created FB %d with size %dx%d\n", fb.id, width, height);
if (first) {
color = 0xFFFFFFFF;
first = false;
} else {
color = colors[color_idx];
color_idx = (color_idx + 1) % (sizeof(colors) / sizeof(colors[0]));
}
dumb_fb_fill(&fb, drm_fd, color);
layer = liftoff_layer_create(output);
liftoff_layer_set_property(layer, "FB_ID", fb.id);
liftoff_layer_set_property(layer, "CRTC_X", x);
liftoff_layer_set_property(layer, "CRTC_Y", y);
liftoff_layer_set_property(layer, "CRTC_W", width);
liftoff_layer_set_property(layer, "CRTC_H", height);
liftoff_layer_set_property(layer, "SRC_X", 0);
liftoff_layer_set_property(layer, "SRC_Y", 0);
liftoff_layer_set_property(layer, "SRC_W", width << 16);
liftoff_layer_set_property(layer, "SRC_H", height << 16);
return layer;
}
int main(int argc, char *argv[])
{
int drm_fd;
struct liftoff_device *device;
drmModeRes *drm_res;
drmModeConnector *connector;
drmModeCrtc *crtcs[MAX_OUTPUTS], *crtc;
struct liftoff_output *outputs[MAX_OUTPUTS], *output;
struct liftoff_layer *layers[MAX_OUTPUTS * LAYERS_PER_OUTPUT];
size_t outputs_len, layers_len;
drmModeAtomicReq *req;
int ret;
size_t i, j;
drm_fd = open("/dev/dri/card0", O_RDWR | O_CLOEXEC);
if (drm_fd < 0) {
perror("open");
return 1;
}
if (drmSetClientCap(drm_fd, DRM_CLIENT_CAP_UNIVERSAL_PLANES, 1) < 0) {
perror("drmSetClientCap(UNIVERSAL_PLANES)");
return 1;
}
if (drmSetClientCap(drm_fd, DRM_CLIENT_CAP_ATOMIC, 1) < 0) {
perror("drmSetClientCap(ATOMIC)");
return 1;
}
device = liftoff_device_create(drm_fd);
if (device == NULL) {
perror("liftoff_device_create");
return 1;
}
drm_res = drmModeGetResources(drm_fd);
outputs_len = 0;
for (i = 0; i < (size_t)drm_res->count_connectors; i++) {
connector = drmModeGetConnector(drm_fd, drm_res->connectors[i]);
if (connector->connection != DRM_MODE_CONNECTED) {
drmModeFreeConnector(connector);
continue;
}
crtc = pick_crtc(drm_fd, drm_res, connector);
if (crtc == NULL || !crtc->mode_valid) {
drmModeFreeConnector(connector);
continue;
}
output = liftoff_output_create(device, crtc->crtc_id);
printf("Using connector %d, CRTC %d\n", connector->connector_id,
crtc->crtc_id);
drmModeFreeConnector(connector);
crtcs[outputs_len] = crtc;
outputs[outputs_len] = output;
outputs_len++;
}
drmModeFreeResources(drm_res);
if (outputs_len == 0) {
fprintf(stderr, "no connector found\n");
return 1;
}
layers_len = 0;
for (i = 0; i < outputs_len; i++) {
output = outputs[i];
crtc = crtcs[i];
layers[layers_len++] = add_layer(drm_fd, output, 0, 0,
crtc->mode.hdisplay,
crtc->mode.vdisplay, false);
for (j = 1; j < LAYERS_PER_OUTPUT; j++) {
layers[layers_len++] = add_layer(drm_fd, output,
100 * j, 100 * j,
256, 256, j % 2);
}
}
for (i = 0; i < layers_len; i++) {
liftoff_layer_set_property(layers[i], "zpos", i);
}
req = drmModeAtomicAlloc();
for (i = 0; i < outputs_len; i++) {
if (!liftoff_output_apply(outputs[i], req)) {
perror("liftoff_output_apply");
return 1;
}
}
ret = drmModeAtomicCommit(drm_fd, req, DRM_MODE_ATOMIC_NONBLOCK, NULL);
if (ret < 0) {
perror("drmModeAtomicCommit");
return false;
}
for (i = 0; i < layers_len; i++) {
printf("Layer %zu got assigned to plane %u\n", i,
liftoff_layer_get_plane_id(layers[i]));
}
sleep(1);
drmModeAtomicFree(req);
for (i = 0; i < layers_len; i++) {
liftoff_layer_destroy(layers[i]);
}
for (i = 0; i < outputs_len; i++) {
liftoff_output_destroy(outputs[i]);
drmModeFreeCrtc(crtcs[i]);
}
liftoff_device_destroy(device);
return 0;
}