HP OmniBook X Flip charges at ~0.2 W without acpi_osi="Windows 2022" - carry DMI OSI quirk

[daviddelsol1998]

HP OmniBook X Flip charges at ~0.2 W without acpi_osi="Windows 2022" — carrying the DMI OSI quirk would fix all CachyOS users

Summary

On HP OmniBook X Flip laptops (2025, AMD Ryzen AI 5 340 — DMI board 8DA8, 103C_5335M8 HP OmniBook X), battery charging is throttled to ~0.2 W on stock Linux. Root cause is verified: the EC firmware only enables normal charging when the OS claims _OSI("Windows 2022"). The DSDT/SSDT contain OSI checks only up to Windows 2022 (confirmed via strings on /sys/firmware/acpi/tables/*), and the kernel's default claimed OSI strings stop before it — so the fast-charging branch never activates on Linux.

Measurements (65 W USB-C PD adapter)

Condition	Charging rate
Stock CachyOS	0.2 W (POWER_SUPPLY_POWER_NOW ≈ 200000 µW)
acpi_osi=! acpi_osi="Windows 2022" on cmdline	31–33 W (Windows parity; ~30 W on / ~60 W off is HP firmware design)
Windows (per HP spec)	~30 W while system on

Verified across kernels 7.1.6-1-cachyos and 6.18.42-1-cachyos-lts and multiple charge cycles (currently 27.8 W charging at 79%).

Workaround (confirmed, persistent)

Kernel cmdline: acpi_osi=! acpi_osi="Windows 2022" — persistent config:

# /etc/default/limine
KERNEL_CMDLINE[linux-cachyos]+=acpi_osi=! acpi_osi="Windows 2022"
KERNEL_CMDLINE[linux-cachyos-lts]+=acpi_osi=! acpi_osi="Windows 2022"

then sudo limine-update (also rebuilds the initramfs, so it takes a while) and reboot.
GRUB users: add the same to GRUB_CMDLINE_LINUX_DEFAULT in /etc/default/grub, then sudo grub-mkconfig -o /boot/grub/grub.cfg.

Upstream

Filed with full evidence + EC RAM attachments:

• kernel bugzilla 221866 (ACPI / Power-Battery) — requesting a DMI-scoped OSI quirk in drivers/acpi/osi.c for board 8DA8 (same mechanism as existing dmi_enable_osi_linux entries).
• Previous report 221776 (closed ANSWERED — the earlier UCSI-corruption theory was withdrawn; UCSI is not involved, the quirk request for ucsi_acpi should not be pursued).
• Same firmware pattern also documented on the Samsung Galaxy Book 4 (Arch forum thread 297580) and across the HP OmniBook X Flip line (HP Community).

Request

1. Carry the quirk in the CachyOS kernel (drivers/acpi/osi.c, DMI-scoped, board 8DA8 + siblings 8EA2 etc.) until it lands upstream — this is the single-line change that fixes every affected CachyOS user with zero config.
2. Document the workaround on the wiki (kernel cmdline + Limine config location) so users can self-help today.

Extra (separate, rare)

~1 in 4 plug/unplug events: PD contract sticks at 5 V/3 A until reboot (UCSI reload, suspend/resume, replug all fail to recover). Tracked in bugzilla 221866 as a secondary issue; a sink-side PDO request interface in UCSI would fix it OS-side.

Hardware: HP OmniBook X Flip Laptop 16-ar0xxx, SKU B5UJ4UA#ABA, serial VNY540CBJH, BIOS/EC 49.34, 68 Wh battery (CI04068XL), 65 W adapter.

[daviddelsol1998]

Maintainers: could you add the bug and upstream labels? This is an upstream kernel bug (bugzilla 221866) that needs a DMI-scoped OSI quirk in drivers/acpi/osi.c.

[daviddelsol1998]

Related: same HP Strix/Krackan Point family firmware-gap pattern as #705 (audio quirk, ACP 7.0). Our charging issue follows the same trajectory: upstream bugzilla 221866 + open HP warranty case #5163282638, awaiting an upstream DMI-scoped OSI quirk in drivers/acpi/osi.c.

[daviddelsol1998]

Update: HP has now provided its written position (warranty case #5163282638, supervisor email 2026-08-10): "HP does not officially recommend, support, or provide drivers for alternative operating systems" - no BIOS/EC update will be provided. Vendor path exhausted. This strengthens the case for carrying the DMI OSI quirk in the CachyOS kernel until upstream (bugzilla 221866, patch attached) merges.

[daviddelsol1998]

Documented the workaround in the CachyOS wiki (PR): CachyOS/wiki#616 — will link it in the Reddit writeup once merged.

[simiscoool-afk]

what is the status of this issue?