Lenovo ThinkPad T480 and T480s on Omarchy

Verdict

Silver. The 2018 ThinkPad T480 and T480s are Kaby Lake-R machines with Intel UHD 620 graphics, an Intel 8265 Wi-Fi card and HDA audio. Every one of those has been in the kernel for years, and nothing in the Omarchy issue tracker suggests they stop working under 4.x. This is the cheap used laptop that mostly behaves.

Two problems are real and neither is fixed in v4.0.4. Omarchy reads only one battery, which matters because the T480’s selling point is two of them. And the fingerprint reader in most units cannot be driven by stock libfprint at all, so the setup wizard fails every time. A third, newer one is worth watching: the bespoke kernel that shipped to everyone in v4.0.4 has a first report of display flicker on this exact GPU.

Everything below was checked against the v4.0.4 source tree and against issues filed on 4.0.x. Where a report is from the 3.x era it says so.

What works

The clearest evidence is a full omarchy debug dump from a T480s (machine type 20L8S27E04, i5-8350U, 8 GB) running 4.0.3 posted in issue #11721. In it:

• Graphics come up on i915 with the UHD 620 at PCI ID 8086:5917, driving the internal 1920x1080 eDP panel under Hyprland 0.56.2.
• Wi-Fi is the Intel 8265/8275 on iwlwifi, interface up. Wired is the Intel I219-LM on e1000e.
• Bluetooth is the Intel btusb adapter, hci0 up at Bluetooth 4.2.
• Audio is Sunrise Point-LP HD Audio on snd_hda_intel, with PipeWire, pipewire-pulse and WirePlumber all active. The T480 predates the SOF-based codecs that cause so much trouble on newer ThinkPads, so the sof-firmware package Omarchy installs on every Intel audio controller is not what makes sound work here.
• Both cameras, the regular Chicony one and the IR camera on IR-equipped units, bind to uvcvideo. That is a driver binding, not a capture test, so the camera row above stays at unknown.
• Suspend states freeze,mem,disk are available and the machine reports deep as the configured suspend mode, so the T480 still has real S3 rather than s2idle only.

That is one machine, so treat the Bluetooth and audio lines as “came up on a live system” rather than as a broad guarantee. There are no open T480 issues against any of them.

What breaks

The second battery is invisible. A T480 has an internal BAT0 and a hot-swappable rear BAT1. omarchy-battery-status in v4.0.4 still starts with upower -e | grep BAT | head -n 1, so the power panel describes BAT0 only, while the bar icon uses UPower’s aggregate DisplayDevice. The two disagree, sometimes wildly: in #6885 the panel showed 94 percent while the icon showed 24 percent on the same machine at the same moment. #7214 and #10244 are the same bug reported again, and a commenter on #7214 adds that the power draw reads 0 W while the laptop is actually discharging, because a T480 runs down the rear pack before it touches the internal one, and the script only reads the internal one. PR #10384 and PR #8864 both propose reading DisplayDevice and summing wattage across packs. Both were still open when this page was written. #5780 reported the same mismatch against the Waybar notification in May and was closed when Quattro replaced Waybar, not fixed, so the script carried the assumption into the new shell. Detail: /fix/battery-drains-fast/ and /hardware/battery-power/.

The fingerprint reader does not enroll. The T480, T480s, T490 and X1 Carbon 6th generation commonly carry the Synaptics Metallica reader 06cb:009a. Omarchy’s omarchy-hw-fingerprint sees the USB device and the wizard proceeds, installs libfprint-git and fprintd, then dies with Impossible to enroll: GDBus.Error:net.reactivated.Fprint.Error.NoSuchDevice: No devices available. In #11721 BeameX points out that libfprint’s own hwdb file lists usb:v06CBp009A* under known unsupported devices, so this is a missing driver, not a flaky sensor. #7006 covers the other half: if you already run the AUR python-validity plus open-fprintd plus fprintd-clients-git stack, rerunning the wizard hits an unresolvable package conflict and can leave you with the reader gone. The same complaint goes back to 3.x in #2046, closed in favor of #2064 with the note that the missing driver is a limitation beyond Omarchy’s scope. v4.0.3 shipped an “improve fingerprint-reader support during setup” change, and #11721 was filed against 4.0.3 anyway, so first-run enrollment is still broken. See /fix/fingerprint-enrollment-fails/ and /hardware/fingerprint/.

Display flicker on the 4.0.4 kernel. #12238 reports random flicker with repeated i915 [drm] *ERROR* CPU pipe A FIFO underrun on a T480 after linux-omarchy 7.2.5-3 replaced stock Arch linux 7.2.3. The reporter suspects i915 power features on Kaby Lake and has not bisected it. One report, filed the day this page was checked, so treat it as a lead, not a verdict. The migration keeps your previous kernel installed, so Limine can still boot it.

Resume is not instant. On a T480s, #8447 describes a lock screen that stays frozen for roughly 30 seconds after S3 resume. Authentication actually succeeds immediately; only the picture is stale. The reporter ruled out lock-screen plugins and hyprmoncfgd, and found no i915 errors. Open.

Docked monitor modes. #10266 is a T480 in clamshell mode on a dock: a monitor with an explicit mode line in monitors.lua comes back at 1920x1080 instead of 2560x1440 after a dock reconnect. hyprctl reload restores it. See /fix/multi-monitor-layout-not-saved/.

TPM left on. Two 3.x-era T480 reports are the same failure: a hang after the LUKS prompt followed by a reboot in #4296, and a two-minute stall at the login screen after the 3.3.0 update in #4120. The logs show systemd’s TPM2 and PCR services timing out on the T480’s TPM, a systemd 259 bug. The fix in both threads was turning TPM off in the BIOS, which the Getting Started chapter of the manual still requires. One T480 in #4120 was stuck in a TPM “MFG mode” it could not disable, and booted only after masking systemd-tpm2-setup, systemd-pcrphase and the other pcr services on the kernel command line. A separate 3.5 offline ISO bug, an archinstall Limine crash that left a T480 unbootable in #5330, was fixed in ISO 3.5.1-2.

What Omarchy does for this model

Nothing by name. bin/omarchy-hw-match is the DMI matcher, and it greps /sys/class/dmi/id/product_name and product_family. No caller in the v4.0.4 tree passes a ThinkPad or T480 pattern, and install/hardware/lenovo/ contains exactly one script, for Yoga Pro 7 bass speakers. Your T480 gets only what install/hardware/all.sh hands every machine of its shape: thermald enabled (Kaby Lake-R clears the Sandy Bridge or newer test), intel-media-driver plus libvpl and vpl-gpu-rt for video acceleration, sof-firmware on the strength of an Intel audio controller being present, and a generic Synaptics InterTouch touchpad nudge that does nothing while the installer runs under arch-chroot. intel-lpmd is skipped, since that script only matches Alder Lake and newer CPU model numbers.

On DMI, ThinkPads put the machine type in product_name and the friendly name in product_version and product_family. Machine types seen in the cited issues are 20L6S77E00 for a T480 and 20L8S27E04, 20L7S2QV00 and 20MF000DUS for the T480s; the T480s family string reads ThinkPad T480s. If you write your own quirk, match on the family, not the machine type, because every configuration has a different type code.

Variants

Prefer the plain T480 if you want the second battery and user-replaceable RAM in both slots. Prefer the T480s if you want a lighter machine and can live with one pack, which incidentally sidesteps the battery reporting bug entirely.

Avoid the MX150 discrete graphics option on the T480 unless you specifically need it. There is no T480 NVIDIA report in the data behind this page, so the hybrid path here is unverified, and hybrid graphics is the single largest source of laptop breakage elsewhere in the tracker. See /hardware/hybrid-gpu/ before you buy one.

Libreboot or Coreboot units without an EFI payload install but never boot: #4208 is a T480 in exactly that state, the installer finished and the first reboot failed because it assumes UEFI. The response was that nobody on the team runs that setup and a patch would be welcome.

Before you install

• Check the fingerprint reader with lsusb | grep -iE '06cb|138a|04f3' first. If it says 06cb:009a, plan on password login.
• Put in 16 GB. The 8 GB unit in the #11721 dump sat at 3 GB used with a fairly idle desktop.
• Confirm UEFI, not Libreboot or a BIOS-only Coreboot payload.
• Turn off Secure Boot and TPM in the BIOS before the first boot. The manual words it as “Secure Boot and/or TPM”, which reads as optional and is not, and #4296 and #4120 are what a T480 does with TPM left on.
• Leave the old kernel entry in Limine alone after updating to 4.0.4, in case of #12238.
• Expect the power panel number to be wrong if you have two packs, and check with upower -i /org/freedesktop/UPower/devices/DisplayDevice when it matters.
• Read the sleep and fingerprint chapters of the official manual: system sleep and hardware authentication.

Related

/hardware/fingerprint/ · /hardware/battery-power/ · /hardware/intel-gpu/ · /hardware/suspend-sleep/ · /hardware/multi-monitor/ · /hardware/lenovo-thinkpad-x1-carbon/ · /reference/commands/omarchy-battery-status/ · /releases/v4.0.4/ · /hardware/submit/