v0.41.0 fails DMABUF on FreeBSD

[jbeich]

Bug or Regression?

Regression

System Info and Version

System/Version info

$ hyprctl systeminfo -c
Hyprland, built from branch main at commit ea2501d4556f84d3de86a4ae2f4b22a474555b9f portbld (?).
Date: Mon Jun 10 14:23:06 UTC 2024
Tag: v0.41.0, commits: ?

flags: (if any)


System Information:
System name: FreeBSD
Node name: foo.bar
Release: 15.0-CURRENT
Version: FreeBSD 15.0-CURRENT #0 main-n270588-480ad4055314: Tue Jun  4 21:21:33 UTC 2024     test@foo.bar:/usr/obj/usr/src/amd64.amd64/sys/MYKERNEL


GPU information: 
vgapci1@pci0:0:2:0:	class=0x030000 rev=0x06 hdr=0x00 vendor=0x8086 device=0x1912 subvendor=0x1462 subdevice=0x7968
    vendor     = 'Intel Corporation'
    device     = 'HD Graphics 530'
    class      = display
    subclass   = VGA

os-release: NAME=FreeBSD
VERSION="15.0-CURRENT"
VERSION_ID="15.0"
ID=freebsd
ANSI_COLOR="0;31"
PRETTY_NAME="FreeBSD 15.0-CURRENT"
CPE_NAME="cpe:/o:freebsd:freebsd:15.0"
HOME_URL="https://FreeBSD.org/"
BUG_REPORT_URL="https://bugs.FreeBSD.org/"


plugins:

======Config-Start======


Config File: /tmp/hyprland.conf: Read Succeeded
bind = SUPER, BackSpace, exit
bind = SUPER, Return, exec, vkcube-wayland

debug {
      enable_stdout_logs = true
      disable_logs = false
}

======Config-End========

Description

Regressed by #6268

# Error out if GPU rendering fails instead of falling back
$ sudo rm /usr/lib/dri/{swrast,zink}_dri.so

$ mpv --no-config --vo=gpu-next --gpu-context=wayland /path/to/foo.mp4
[...]
libEGL warning: wayland-egl: we could not map the format table so we won't be able to use this batch of dma-buf feedback events.
libEGL warning: MESA-LOADER: failed to open swrast: Cannot open "/usr/lib/dri/swrast_dri.so" (search paths /usr/local/lib/dri, suffix _dri)

[vo/gpu-next] Failed initializing any suitable GPU context!
Error opening/initializing the selected video_out (--vo) device.

$ vkcube-wayland
Selected GPU 0: Intel(R) HD Graphics 530 (SKL GT2), type: IntegratedGpu
Can't find our preferred formats... Falling back to first exposed format. Rendering may be incorrect.

$ Hyprland
[...]
[ERR] [LinuxDMABUF] FIXME: surface feedback stub
[LOG] [LinuxDMABUF] Creating a dmabuf, with id 20: size [Vector2D: x: 500, y: 500], fmt 0, planes 1
[LOG] [LinuxDMABUF]  | plane 0: mod 72057594037927935 fd 41 stride 512 offset 0
[EGL] command: eglCreateImageKHR, error: EGL_BAD_MATCH (0x3009), message: "unknown drm fourcc format"
[ERR] EGL: EGLCreateImageKHR failed: 12297
[ERR] [LinuxDMABUF] Possibly compositor bug: buffer failed to create
[...]

How to reproduce

Unknown. I can't test on Linux or with an AMD GPU.

Crash reports, logs, images, videos

hyprland.log

[vaxerski]

hm, looks like the format table got fucked.

[vaxerski]

Hyprland/src/protocols/LinuxDMABUF.hpp

Lines 33 to 39 in ea2501d

	#pragma pack(push, 1)
	struct SDMABUFFeedbackTableEntry {
	uint32_t fmt = 0;
	char pad[4];
	uint64_t modifier = 0;
	};
	#pragma pack(pop)

Is there something wrong here? I remember wlroots having some comment complaining about BSD.

I'm guessing this is some bsd issue.

[vaxerski]

generally from the log hyprland seems to collect the dmabuf formats correctly, but the table in linux-dmabuf is malformed.

Most of the logic is here: https://github.com/hyprwm/Hyprland/blob/811429bfd4a46f33c7788580f72038b0c3c1c2b1/src/protocols/LinuxDMABUF.cpp

and it was tested under nvidia, amd and intel (albeit on linux) and was working.