ASUS ExpertBook Ultra on Linux,
without the broken bits.

The kernel's HID descriptor parser inflates ABS_MT_PRESSURE max to 2601 — literally the Y-axis max value, suggesting a parser typo — for this PixArt haptic touchpad. Real hardware values top out around 1000. libinput's pressure thresholds are calibrated against the kernel-reported max, so real touches register at 1–6% of the bogus max — well below the activation threshold. Result: every motion is rejected as a "kernel bug: Touch jump."

$ sudo dmesg | grep "Touch jump" | wc -l
1873      ← without the module
0         ← with the module

Files installed:

• /etc/udev/hwdb.d/61-pixart-4f05-pressure-fix.hwdb — clamps the pressure axes via EVDEV_ABS_18 / EVDEV_ABS_3A
• /etc/libinput/local-overrides.quirks — tells libinput to ignore pressure entirely (AttrEventCode=-ABS_MT_PRESSURE;-ABS_PRESSURE)

1. Cirrus CS35L56 amps boot in FIRMWARE_MISSING because linux-firmware-cirrus < 20260519 doesn't ship per-OEM blobs for PCI subsystem 1043:15e4.
2. The combined sidecar codec cs35l56+cs42l43-spk has no UCM in alsa-ucm-conf < 1.2.16, and its SpeakerCodec regex drops the -spk suffix — so the UCM fails to open. (Fixed upstream in alsa-ucm-conf 1.2.16.)
3. WirePlumber then falls back to stereo-fallback, which plays to the Jack PCM, never the speaker output PCM.

$ sudo dmesg | grep cs35l56
cs35l56 sdw:0:2:01fa:3556:01:0: FIRMWARE_MISSING                    ← without
cs35l56 sdw:0:2:01fa:3556:01:1: FIRMWARE_MISSING                    ← without
─────────────────────────────────────────────────────────────────────────
cs35l56 sdw:0:2:01fa:3556:01:0: Calibration applied                 ← with
cs35l56 sdw:0:2:01fa:3556:01:0: Tuning PID: 0x23134, SID: 0x470200  ← with

Files installed (v2.1.0):

• /lib/firmware/cirrus/cs35l56-b0-dsp1-misc-104315e4-l2u{0,1}.bin — per-OEM tuning blobs (fallback for linux-firmware-cirrus < 20260519)
• /lib/firmware/cirrus/cs35l56-b0-dsp1-misc-104315e4-l2u{0,1}.wmfw — generic CS35L56 Rev 3.13.4 firmware patch (chip ROM upgrades 3.4.4 → 3.13.4)
• /etc/wireplumber/wireplumber.conf.d/52-disable-bt-sco-offload.conf — silences the dead SSP2-BT offload PCM that spams the kernel log
• HiFi UCM — only on alsa-ucm-conf < 1.2.16 (1.2.16+ ships it in the package): sof-soundwire.conf (fixed -spk regex), the cs35l56+cs42l43-spk / cs42l43-spk+cs35l56 speaker confs, and codecs/cs42l43-spk+cs35l56/init.conf

The core problem is 802.11be (EHT / Wi-Fi 7) itself on the Intel BE211 (8086:e440) under iwlwifi/iwlmld: the EHT RX path collapses to MCS0 / NSS1 and MLO sessions tear down, so the "Wi-Fi 7" link is slower and flakier than plain Wi-Fi 6. Not fixed upstream as of Linux 7.1-rc7. Two secondary irritants pile on: iwlmld defaults to power_scheme=2 (beacon-loss recovery churn), and the iwlwifi TX-segmentation offload bug can throw Microcode SW error freezes under heavy traffic.

$ sudo dmesg | grep -E "missed beacons|Microcode SW error" | wc -l
2046      ← with EHT on
0         ← with disable_11be=Y (Wi-Fi 6 fallback)

Core fix — drop the broken 802.11be layer so the radio runs as rock-solid Wi-Fi 6 (HE). Same approach Omarchy ships; verified at ~2.1 Gbit/s over 160 MHz 6 GHz HE here.

Files installed:

• /etc/modprobe.d/iwlwifi-disable-eht.conf — options iwlwifi disable_11be=Y; disables EHT / Wi-Fi 7, link falls back to stable Wi-Fi 6 / HE
• /etc/modprobe.d/iwlmld-active.conf — options iwlmld power_scheme=1 (secondary; trims HE-mode instability)
• /etc/tmpfiles.d/pcie-aspm-performance.conf — global ASPM = performance at boot (secondary)
• /etc/NetworkManager/dispatcher.d/90-iwlwifi-no-offload — ethtool -K $iface tso off gso off gro off on every iwlwifi up (secondary)

This is a deliberate Wi-Fi 7 → Wi-Fi 6 downgrade — EHT is the problem on this silicon, and the secondary tunables do not keep Wi-Fi 7 alive. Bluetooth coex (bt_coex_active=Y) is left untouched, so Bluetooth keeps working.

The Samsung Display Corp panel reports IEEE OUI 00:aa:01 in DPCD register 0x300 and supports Panel Replay Selective Update (Early Transport). The xe driver's PSR idle wait times out on this panel firmware:

xe 0000:00:02.0: [drm] *ERROR* Timed out waiting PSR idle state
xe 0000:00:02.0: [drm] *ERROR* [CRTC:151:pipe A] DSB 0 timed out waiting for idle
kwin_wayland: Pageflip timed out! This is a bug in the xe kernel driver

Once wedged, only a reboot recovers — modeset cycle, GPU reset, and runtime PSR-disable via debugfs all fail to unstick the display state buffer. Worse than a black panel: when the PSR2 selective-fetch path deadlocks the DSB during heavy compositing (a screen capture is enough), it can take the whole kernel down — a silent hard hang with no oops, no MCE, empty pstore/BERT. Same pattern Intel addresses for Dell XPS 14/16 in mainline (per-device disable in intel_quirks.c) — our upstream-patches/0001 ports the same disable to our subsystem ID.

There is no dedicated upstream tracker for this Panther Lake PSR2 selective-fetch / DSB hang; it's reproduced locally on linux-cachyos 7.0.11 and linux-cachyos-rc 7.1-rc7. (drm/xe #7513 — "Lunar lake, rare shutdown under load" — is a distinct Lunar Lake PMC-firmware bug that leaves a BERT: [Hardware Error] record and is not cured by disabling PSR, so it is not this one.)

Files installed:

• /etc/modprobe.d/xe-disable-psr.conf — belt-and-suspenders for late module load
• (managed block) /etc/default/limine — appends xe.enable_psr=0 xe.enable_psr2_sel_fetch=0 xe.enable_panel_replay=0 to kernel cmdline; the post-install hook calls limine-update

Without a userspace thermal daemon, the kernel's only lever is "cap CPU frequency". On Panther Lake's hybrid topology (P + E + LP-E cores) a P/E-aware throttle is far smarter — it can park work on slower cores instead of slowing everything down. Without intel-lpmd, idle work spreads; with it, all idle work concentrates on a single LP-E core and P-cores deep-sleep.

Packages installed (no payload files):

• thermald (extra repo) — thermald.service enabled
• intel-lpmd (extra / cachyos repo, plain pacman) — intel_lpmd.service enabled

Both daemons coexist with the existing power-profiles-daemon: PPD handles user profile, thermald handles thermal, intel-lpmd handles idle topology. Different layers, no conflict.

Windows Studio Effects on Copilot+ PCs runs background blur, smart framing, eye-contact correction, and voice focus on the NPU. None of these are shipped on Linux out of the box, even though the Intel Panther Lake NPU itself is fully supported by the kernel (intel_vpu driver, /dev/accel/accel0 exposed) and the userspace stack (OpenVINO 2026, level-zero) is available via the AUR.

This module bundles the userspace plumbing: a virtual camera target via v4l2loopback, OBS Studio as the capture + filter pipeline, and the obs-backgroundremoval plugin which runs ONNX models. On Linux that plugin is CPU-only — its execution providers are CUDA / ROCm / MIGraphX, and there is no OpenVINO/NPU path in the OBS plugin (installing openvino does not add an "NPU" device). CPU is fine for 720p30 background blur.

Files installed:

• /etc/modules-load.d/v4l2loopback.conf — auto-load at boot
• /etc/modprobe.d/v4l2loopback-options.conf — devices=1 video_nr=10 card_label='AI Camera' exclusive_caps=1

Packages installed:

• v4l2loopback-dkms (extra) — kernel module
• obs-studio (extra) — orchestrator
• obs-backgroundremoval (AUR) — ML segmentation plugin (ONNX); CPU-only on Linux

Use: open OBS, add a Video Capture Device source pointing at the real webcam, attach the Background Removal filter, and start the virtual camera (Tools → Start Virtual Camera). Apps then see "AI Camera" as a webcam choice.

Reality check — no NPU offload in OBS on Linux. For an actual NPU route see ericjchang/linux-studio-effects (OpenVINO + v4l2loopback), but it's validated on Arrow Lake, not yet Panther Lake, and installs via git + pip.

The backlight is not dead — the Fn hotkeys toggle it fine (handled by the EC in hardware, bypassing the buggy ACPI path). This module is optional: it only restores software control (the KDE slider / brightnessctl / sysfs).

On BIOS B9406CAA.304, the SLKB ACPI method that sets keyboard-backlight brightness has a broken branch. Disassembled from the live DSDT:

Method (SLKB, 1, NotSerialized) {
    If    ((Arg0 >= 0x0100) && (Arg0 <= 0x0106)) { Local0 = (Arg0 - 0x0100) }
    ElseIf((Arg0 >= 0x80)   && (Arg0 <= 0x83))   { Local0 = (Arg0 - 0x80) * 0x21 ... }
    ElseIf((Arg0 >= Zero)   && (Arg0 <= 0x03))   { Local0 = Zero }   // ← BUG
    STBC (Zero, Local0)
    Return (One)
}

The mainline asus-wmi Linux driver writes the standard kernel range 0..3 to set keyboard brightness. That hits the third branch which unconditionally clamps Local0 to zero. So every software write — KDE / brightnessctl / direct /sys/class/leds/ — is silently turned into "set brightness 0" and no-ops. (The Fn hotkeys are unaffected; they go through the EC, not this method.) The OEM-tested 0x100..0x103 range works correctly.

Workaround: the userspace daemon asusd (shipped in asusctl) translates standard kernel-level brightness writes into the OEM range before invoking ACPI, side-stepping the buggy branch. Once asusd is running, KDE PowerDevil's keyboard-brightness control reaches the EC correctly.

Files installed:

• /usr/share/dbus-1/system-services/xyz.ljones.Asusd.service — D-Bus activation entry that ASUS doesn't ship; without it asusd.service (Type=dbus) never auto-starts
• /etc/modules-load.d/acpi_call.conf — auto-load acpi_call for any direct ACPI fallback debugging
• /etc/asusd/ directory — asusd refuses to start without it; pacman package leaves it absent

Packages installed:

• asusctl (extra) — provides the asusd daemon
• acpi_call-dkms (AUR, manual) — kernel module for direct ACPI invocations during debugging

If a future ASUS BIOS revision fixes the SLKB branch, this entire module becomes redundant — uninstall it.

Most of it transfers. The audio-fix firmware blobs are matched on PCI subsystem 1043:15e4 (this exact laptop). Sibling subsystems 104315d4 and 104315f4 ship different per-OEM tuning files in upstream linux-firmware. The touchpad-fix, wifi-fix, display-fix, intel-perf-fix, webcam-ai-fix, and keyboard-backlight-fix modules are hardware-agnostic or match by family-level identifiers and apply more broadly. PRs adding sibling-model entries are welcome.

No. The Wi-Fi module deliberately leaves iwlwifi.bt_coex_active=Y alone. Bluetooth audio, HID, and file transfer keep working as usual.

Yes — on purpose. 802.11be / EHT is broken on the BE211 (RX collapses to MCS0/NSS1, MLO tears down), so wifi-fix disables it (disable_11be=Y) and the link runs as stable Wi-Fi 6 / HE instead — ~2.1 Gbit/s over 160 MHz 6 GHz here, faster in practice than the flaky EHT link. Drop the module (or set disable_11be=N) once Intel fixes the iwlwifi EHT path upstream.

The HiFi UCM (active since audio-fix v2.0.0, and upstream in alsa-ucm-conf 1.2.16) drives the mic-mute LED (platform::micmute) correctly. The speaker-mute LED (F1) stays in its EC default state because this laptop exposes no speaker-mute LED device to Linux at all — there's nothing for the UCM SetLED hook to bind to. It's a missing-device limitation, not a profile issue.

That's the goal — see upstream-patches/. Two of the audio pieces already landed: alsa-ucm-conf 1.2.16 ships the cs42l43-spk+cs35l56 codec dir and linux-firmware-cirrus >= 20260519 ships the OEM blobs, so on an up-to-date system audio-fix is already down to a single SSP2-BT topology-noise drop-in. As the remaining kernel/quirk patches land and your distro picks them up, every module here becomes deletable.

Yes. Every module is reversible (./patch.sh uninstall <mod> removes the files it placed and reverts service enables). All file targets are well-known config paths (/etc/modprobe.d, /etc/udev/hwdb.d, /etc/libinput, /lib/firmware, /usr/share/alsa, /etc/wireplumber/wireplumber.conf.d, /etc/NetworkManager/dispatcher.d) — no kernel binary replacement, no pacman lock manipulation, no boot partition rewriting except the cmdline injection in display-fix's case (which is also fully reversible via the uninstall hook).

./patch.sh diff <module>
# shows source vs installed for every file the module manages,
# marked unchanged / would update / would create

MIT for code (scripts, configs, patches). Firmware blobs under audio-fix/ are redistributed verbatim from upstream linux-firmware under their original Cirrus Logic redistribution license; see NOTICE.