Crash on FreeBSD #99
Comments
|
Interesting, apparently |
|
|
|
I'm not really sure what the proper fix to this is. Since the upper bytes are zero maybe it'd work to truncate it, but that seems messy. I could maybe ask rustix folks to define their |
diff --git a/src/cap_wlr_screencopy.rs b/src/cap_wlr_screencopy.rs
index 0ead4d0..ab66cbf 100644
--- a/src/cap_wlr_screencopy.rs
+++ b/src/cap_wlr_screencopy.rs
@@ -93,6 +93,11 @@ impl Dispatch<ZwpLinuxDmabufFeedbackV1, ()> for State<CapWlrScreencopy> {
) {
use wayland_protocols::wp::linux_dmabuf::zv1::client::zwp_linux_dmabuf_feedback_v1::Event;
if let Event::MainDevice { device } = event {
+ // HACK for https://github.com/russelltg/wl-screenrec/issues/99
+ // dev_t is the wrong size in rust's libc crate on freebsd
+ // just hope that the final bytes are zero....
+ #[cfg(target_os = "freebsd")]
+ let device = &device[0..std::mem::size_of::<dev_t>()];
let dev = dev_t::from_ne_bytes(device.try_into().unwrap());
let node = DrmNode::from_dev_id(dev).unwrap();
let node = node Can you test this hack patch? |
|
Can you test this hack patch?
It doesn't help:
```
$ RUST_BACKTRACE=1 wl-screenrec
Opening libva device from DRM device /dev/dri/card0
Xlib: extension "DRI2" missing on display ":0".
[AVHWDeviceContext @ 0x350668a192c0] libva: vaGetDriverNames() failed with operation failed
[AVHWDeviceContext @ 0x350668a192c0] Failed to initialise VAAPI connection: 1 (operation failed).
failed to create encoder(s): Failed to load vaapi device: Input/output error
This is likely *not* a bug in wl-screenrec, but an issue with your vaapi installation. Follow your distribution's instructions. If you're pretty sure you've done this correctly, create a new issue with the output of `vainfo` and if `wf-recorder -c h264_vaapi -d /dev/dri/card0` works.
thread 'main' panicked at src/main.rs:904:51:
enc left in intermediate state
stack backtrace:
0: rust_begin_unwind
1: core::panicking::panic_fmt
2: wl_screenrec::State<S>::on_copy_src_ready
3: wl_screenrec::cap_wlr_screencopy::<impl wayland_client::event_queue::Dispatch<wayland_protocols_wlr::screencopy::v1::generated::client::zwlr_screencopy_frame_v1::ZwlrScreencopyFrameV1,()> for wl_screenrec::State<wl_screenrec::cap_wlr_screencopy::CapWlrScreencopy>>::event
4: wayland_client::event_queue::queue_callback
5: wayland_client::event_queue::EventQueue<State>::dispatching_impl
6: wayland_client::event_queue::EventQueue<State>::blocking_dispatch
7: wl_screenrec::main
note: Some details are omitted, run with `RUST_BACKTRACE=full` for a verbose backtrace.
```
|
|
What if you pass --dri-drvice /dev/dri/renderD128 (or whatever is the correct render node) |
|
I confirm, passing `--dri-device` makes the hack work:
```
$ wl-screenrec --dri-device /dev/dri/renderD128
Opening libva device from DRM device /dev/dri/renderD128
[h264_vaapi @ 0x548888305800] Driver does not support any RC mode compatible with selected options (supported modes: CQP).
failed to open encoder in low_power mode (Invalid argument), trying non low_power mode. if you have an intel iGPU, set enable_guc=2 in the i915 module to use the fixed function encoder. pass --low-power=off to suppress this warning
1 fps
2 fps
2 fps
8 fps
3 fps
^C
```
|
|
Can I get the |
|
|
Hmm OK. The hack is working as intended, but isn't properly converting it into a render node ( |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Regressed by a1599ee. Other
zwp_linux_dmabuf_v1consumers likempv --vo=dmabuf-waylandwork fine.The text was updated successfully, but these errors were encountered: