checks: give both VM machines a virtio GPU, and say why when it is not enough
`nix build .#checks.x86_64-linux.vm-basic` failed because Hyprland 0.56.2
aborted with `CBackend::create() failed!` on every start. Reading aquamarine
0.15.1 rather than guessing at it, that abort had three independent causes on a
NixOS test machine, and none of them is visible in the journal until
`debug:disable_logs = false` is set -- Hyprland mutes aquamarine, so the abort
is all that survives. checks/lib/gpu-session.nix now supplies all three and
records the evidence for each:
* `-vga virtio`, for a DRM render node. `virtualisation.graphics = false`
still leaves QEMU's default bochs-drm adapter, which gives card0 and no
renderD*. Measured in the guest: `-vga virtio` -> card0 + renderD128;
`-device virtio-gpu-pci` only *adds* a card next to bochs-drm, so it would
need `-vga none` beside it.
* `hardware.graphics.enable`, so Mesa's GBM backends exist at all. Without it
gbm_create_device() fails with "MESA-LOADER: failed to open dri:
/run/opengl-driver/lib/gbm/dri_gbm.so" and aquamarine reports "Cannot open
backend: no allocator available".
* `SEATD_VTBOUND=0`, so libseat's embedded seatd hands out a seat with no VT.
`CHeadlessBackend::drmFD()` returns -1 unconditionally, so the *only*
source of an allocator is the DRM backend, which opens /dev/dri/card0
through libseat -- and the default VT-bound seat needs /dev/tty0, which the
unprivileged service user cannot open.
The diagnostics dump grows a "graphics devices" section (`ls -l /dev/dri`,
`cat /proc/fb`) and a "wayland socket vs WAYLAND_DISPLAY" section, so the next
machine that is missing a render node, or that binds a socket the frame server
is not waiting on, says so in three lines instead of timing out mutely.
vm-basic's `dump()` is now vm-hyprland's, which reports the exit status too.
vm-basic also gained `python3` in the guest: its last subtest shells out to a
python one-liner and had been dying on exit code 127 -- invisible until the
compositor got far enough to reach that subtest.
Neither check passes yet, and the two remaining blockers both live in
nix/module.nix, which this slice may not edit, so they are reported rather
than patched:
* vm-basic stops at "the module's two units come up". The compositor is up
and healthy; the frame server sits in ExecStartPre forever. The module
hardcodes `waylandDisplay = "wayland-0"`, but Hyprland's socket loop is
`// get socket, avoid using 0` over candidates 1..32 (src/Compositor.cpp),
so the session binds wayland-1. Forcing WAYLAND_DISPLAY=wayland-1 in the
guest was verified to carry vm-basic through every other subtest.
* vm-hyprland stops at "the compositor reports the configured headless
output" with "hyprctl never reported a HEADLESS-1 monitor". hyprctl reports
`Monitor Virtual-1` at 1280x800 -- the virtio-gpu output. A HEADLESS-N
output only exists after `hyprctl output create headless`, which the module
never issues; Hyprland's own fallback output is named FALLBACK. The
`monitor = HEADLESS-1, ...` line in hyprland.conf is a rule for a monitor
that is never created.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_018fJZYyS4pqxt2FC37JsFYd
This commit is contained in:
1 parent
db886faa38
commit
a37b3dd485
4 files changed
+133
-7
No files matched your search
@@ -0,0 +1,84 @@
|
||||
# What a NixOS test machine has to be given before Hyprland 0.56.2 will start.
|
||||
#
|
||||
# Imported by both VM checks. It is a machine definition, not a test: nothing
|
||||
# here relaxes anything either check asserts, it only supplies hardware and
|
||||
# session facts that a default `nodes.machine` does not have.
|
||||
#
|
||||
# The failure it fixes is the same one every time -- the compositor aborts with
|
||||
#
|
||||
# terminate called after throwing an instance of 'std::runtime_error'
|
||||
# what(): CBackend::create() failed!
|
||||
#
|
||||
# and systemd restarts it until the test's timeout expires. `CBackend::create()`
|
||||
# is aquamarine's, and reading aquamarine 0.15.1 rather than guessing at it, it
|
||||
# needs three separate things that a bare test machine does not provide. Each is
|
||||
# one of the blocks below, and each was confirmed in the guest by removing it
|
||||
# again and watching the compositor die on the next line down.
|
||||
#
|
||||
# A note for whoever hits this next: none of it is visible in the journal until
|
||||
# you also set `debug:disable_logs = false`, because Hyprland mutes aquamarine
|
||||
# by default and the abort is all that survives. That is how the three causes
|
||||
# below were separated.
|
||||
{ ... }:
|
||||
{
|
||||
# 1. A DRM render node.
|
||||
#
|
||||
# `virtualisation.graphics = false` still leaves QEMU's default bochs-drm
|
||||
# adapter in the guest, which offers /dev/dri/card0 and *no* renderD*. Measured
|
||||
# in the guest, not assumed:
|
||||
#
|
||||
# (default) card0 (bochs-drm)
|
||||
# -vga virtio card0, renderD128 (virtio_gpu)
|
||||
# -device virtio-gpu-pci card0, card1, renderD128 (bochs + virtio)
|
||||
# -vga none -device virtio-gpu-pci card0, renderD128 (virtio_gpu)
|
||||
#
|
||||
# `-vga virtio` *replaces* the default adapter, so it is the one-option form of
|
||||
# the last line. `-device virtio-gpu-pci` on its own only *adds* a second card
|
||||
# next to bochs-drm, so it would need `-vga none` beside it to leave the guest
|
||||
# with a single DRM device.
|
||||
virtualisation.qemu.options = [ "-vga virtio" ];
|
||||
|
||||
# 2. Mesa, where the GBM loader looks for it.
|
||||
#
|
||||
# aquamarine's allocator is gbm_create_device() on that node. On NixOS the GBM
|
||||
# backends live under /run/opengl-driver, which only exists when
|
||||
# hardware.graphics is enabled -- a test machine has it off. Without it:
|
||||
#
|
||||
# MESA-LOADER: failed to open dri: /run/opengl-driver/lib/gbm/dri_gbm.so:
|
||||
# cannot open shared object file: No such file or directory
|
||||
# ERR from aquamarine ]: Couldn't open a GBM device at fd 21
|
||||
# CRIT from aquamarine ]: Cannot open backend: no allocator available
|
||||
hardware.graphics.enable = true;
|
||||
|
||||
# 3. A seat that is not bound to a VT.
|
||||
#
|
||||
# This one is not really the test machine's business, and it is flagged rather
|
||||
# than fixed: it belongs in nix/module.nix, which this slice may not edit.
|
||||
#
|
||||
# Only the DRM backend ever yields a buffer allocator --
|
||||
# `CHeadlessBackend::drmFD()` returns -1 unconditionally (aquamarine
|
||||
# src/backend/Headless.cpp), and CBackend::start() takes the allocator from the
|
||||
# first implementation with a drm fd. So "headless" here does not mean "no
|
||||
# GPU": the headless backend still rides on the DRM backend's allocator, and if
|
||||
# the DRM backend fails there is none, whatever /dev/dri holds.
|
||||
#
|
||||
# The DRM backend opens /dev/dri/card0 through libseat. For this unit libseat
|
||||
# finds no /run/seatd.sock and no logind session (it is a system service, not a
|
||||
# login), so it falls back to its embedded seatd -- whose seat is VT-bound by
|
||||
# default. It then tries to open /dev/tty0, which the unprivileged `hhstream`
|
||||
# user cannot, so the client never becomes active and every device open is
|
||||
# refused:
|
||||
#
|
||||
# ERR ]: [libseat] [common/terminal.c:162] Could not open target tty: Permission denied
|
||||
# ERR ]: [libseat] [seatd/seat.c:557] Could not open VT for client
|
||||
# ERR ]: libseat: Couldn't open device at /dev/dri/card0
|
||||
# ERR ]: drm: Skipping device /sys/devices/.../drm/card0, not a KMS device
|
||||
# ERR ]: drm: Found no gpus to use, cannot continue
|
||||
# ERR ]: DRM Backend failed
|
||||
#
|
||||
# SEATD_VTBOUND=0 (seatd/server.c) asks that embedded seatd for a seat with no
|
||||
# VT at all, which is the truth about this session: there is no terminal to
|
||||
# switch away from. A headless session run as a system user will need this on
|
||||
# any machine, not just in the test.
|
||||
systemd.services.hyprland-headless.environment.SEATD_VTBOUND = "0";
|
||||
}
|
||||
+16
-4
@@ -8,6 +8,7 @@
|
||||
testers,
|
||||
hhstream,
|
||||
testTools,
|
||||
python3,
|
||||
hhsModule,
|
||||
width ? 1280,
|
||||
height ? 720,
|
||||
@@ -22,7 +23,9 @@ testers.runNixOSTest {
|
||||
globalTimeout = 900;
|
||||
|
||||
nodes.machine = {
|
||||
imports = [ hhsModule ];
|
||||
# gpu-session.nix is what makes the compositor able to start at all; see
|
||||
# that file for the three things a bare test machine is missing.
|
||||
imports = [ hhsModule ./lib/gpu-session.nix ];
|
||||
|
||||
services.hyprland-headless-stream = {
|
||||
enable = true;
|
||||
@@ -51,12 +54,19 @@ testers.runNixOSTest {
|
||||
environment.systemPackages = [
|
||||
testTools
|
||||
hhstream
|
||||
# The "client dying mid-stream" subtest opens the socket from a python
|
||||
# one-liner in the guest. hhstream's own interpreter is reachable only
|
||||
# through its wrapper, so without this the subtest dies on exit code 127
|
||||
# (`python3: command not found`) rather than on anything it asserts.
|
||||
python3
|
||||
];
|
||||
|
||||
# Mesa and the compositor together need more than the fake capture alone
|
||||
# did; the GPU itself comes from gpu-session.nix.
|
||||
virtualisation = {
|
||||
memorySize = 1024;
|
||||
memorySize = 2048;
|
||||
cores = 2;
|
||||
diskSize = 3072;
|
||||
diskSize = 4096;
|
||||
graphics = false;
|
||||
};
|
||||
};
|
||||
@@ -68,7 +78,9 @@ testers.runNixOSTest {
|
||||
|
||||
def dump(label):
|
||||
print(f"########## diagnostics: {label} ##########")
|
||||
print(machine.execute("hhs-hyprland-diagnostics")[1])
|
||||
status, out = machine.execute("hhs-hyprland-diagnostics", timeout=120)
|
||||
print(out)
|
||||
print(f"########## end diagnostics (exit {status}) ##########")
|
||||
|
||||
|
||||
start_all()
|
||||
|
||||
@@ -24,7 +24,9 @@ testers.runNixOSTest {
|
||||
globalTimeout = 1800;
|
||||
|
||||
nodes.machine = {
|
||||
imports = [ hhsModule ];
|
||||
# gpu-session.nix is what makes the compositor able to start at all; see
|
||||
# that file for the three things a bare test machine is missing.
|
||||
imports = [ hhsModule ./lib/gpu-session.nix ];
|
||||
|
||||
services.hyprland-headless-stream = {
|
||||
enable = true;
|
||||
@@ -53,8 +55,9 @@ testers.runNixOSTest {
|
||||
];
|
||||
|
||||
# The compositor needs more room than the fake capture did, and its software
|
||||
# renderer wants a CPU. There is still no GPU: Hyprland runs on its headless
|
||||
# backend over llvmpipe.
|
||||
# renderer wants a CPU. The virtio GPU from gpu-session.nix is there for its
|
||||
# render node and its DRM backend, not for acceleration: rendering is still
|
||||
# Mesa's software path.
|
||||
virtualisation = {
|
||||
memorySize = 2048;
|
||||
cores = 2;
|
||||
|
||||
@@ -32,6 +32,15 @@ done
|
||||
|
||||
run systemctl --no-pager --full list-dependencies "$SERVICE_UNIT"
|
||||
|
||||
say "graphics devices"
|
||||
# The one piece of hardware a headless Hyprland still needs. A NixOS test
|
||||
# machine gets QEMU's default bochs-drm adapter, which gives a card0 and *no*
|
||||
# renderD*; aquamarine's headless backend wants a render node and aborts with
|
||||
# "CBackend::create() failed!" without one. So "card0 but no renderD128" here
|
||||
# is the signature of a machine that needs a virtio GPU in qemu.options.
|
||||
ls -l /dev/dri 2>&1 || echo "[no /dev/dri: this machine has no DRM device at all]"
|
||||
run cat /proc/fb
|
||||
|
||||
say "socket and runtime directories"
|
||||
ls -la /run/hyprland-headless-stream 2>&1 || echo "[no /run/hyprland-headless-stream]"
|
||||
ls -la /run/user 2>&1 || echo "[no /run/user]"
|
||||
@@ -39,6 +48,24 @@ for d in /run/user/*; do
|
||||
[ -d "$d" ] && ls -la "$d" 2>&1
|
||||
done
|
||||
|
||||
say "wayland socket vs WAYLAND_DISPLAY"
|
||||
# The other silent killer: the frame server waits for
|
||||
# $XDG_RUNTIME_DIR/$WAYLAND_DISPLAY, so if the compositor bound a different
|
||||
# socket name the unit just sits in ExecStartPre until the test times out, with
|
||||
# a healthy compositor next to it. Print both names side by side.
|
||||
systemctl show -p Environment --value "$SERVICE_UNIT" 2>&1 \
|
||||
| tr ' ' '\n' | grep -E '^(WAYLAND_DISPLAY|XDG_RUNTIME_DIR)=' \
|
||||
|| echo "[no WAYLAND_DISPLAY/XDG_RUNTIME_DIR in the unit environment]"
|
||||
for d in /run/hyprland-headless-stream /run/user/*; do
|
||||
[ -d "$d" ] || continue
|
||||
found_sock=$(find "$d" -xdev -maxdepth 1 -name 'wayland-*' ! -name '*.lock' 2>/dev/null)
|
||||
if [ -n "$found_sock" ]; then
|
||||
echo "bound sockets in $d: $found_sock"
|
||||
else
|
||||
echo "bound sockets in $d: [none]"
|
||||
fi
|
||||
done
|
||||
|
||||
say "hyprland logs"
|
||||
# hyprland.log lives beside the compositor's IPC sockets; -xdev keeps this out
|
||||
# of /nix/store and /proc.
|
||||
|
||||
Reference in new issue
Block a user