mirror of
https://gitlab.freedesktop.org/emersion/libliftoff.git
synced 2024-12-25 21:59:11 +01:00
c05f290f95
Instead of exposing to consumers write streaming log data into a per-device buffer that gets created on request and destroyed after the buffer has been written to the log. Passing a NULL string to the internal continuous log function triggers a write of the buffer to the log. BREAKING CHANGE: continuous log function is moved from log header to private header. |
||
---|---|---|
example | ||
include | ||
test | ||
.build.yml | ||
.editorconfig | ||
.gitignore | ||
alloc.c | ||
device.c | ||
layer.c | ||
LICENSE | ||
list.c | ||
log.c | ||
meson.build | ||
output.c | ||
plane.c | ||
README.md |
libliftoff
Lightweight hardware composer library for libdrm.
libliftoff eases the use of KMS planes from userspace without standing in your way. Users create "virtual planes" called layers, set KMS properties on them, and libliftoff will allocate planes for these layers if possible.
See the blog post introducing the project for more context.
Building
Depends on libdrm. Requires universal planes and atomic.
meson build/
ninja -C build/
Usage
See liftoff.h
. Here's the general idea:
struct liftoff_device *device;
struct liftoff_output *output;
struct liftoff_layer *layer;
drmModeAtomicReq *req;
int ret;
device = liftoff_device_create(drm_fd);
output = liftoff_output_create(device, crtc_id);
layer = liftoff_layer_create(output);
liftoff_layer_set_property(layer, "FB_ID", fb_id);
/* Probably setup more properties and more layers */
req = drmModeAtomicAlloc();
if (!liftoff_output_apply(output, req)) {
perror("liftoff_output_apply");
exit(1);
}
ret = drmModeAtomicCommit(drm_fd, req, DRM_MODE_ATOMIC_NONBLOCK, NULL);
if (ret < 0) {
perror("drmModeAtomicCommit");
exit(1);
}
drmModeAtomicFree(req);
License
MIT