linux-kernel - [PATCH v2 0/1] platform/x86: asus-wmi: disable USB0 hub on ROG Ally before suspend

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [thread-next>] [day] [month] [year] [list]

Date:   Mon, 27 Nov 2023 12:05:20 +1300
From:   "Luke D. Jones" <luke@...nes.dev>
To:     hdegoede@...hat.com
Cc:     ilpo.jarvinen@...ux.intel.com, corentin.chary@...il.com,
        platform-driver-x86@...r.kernel.org, linux-kernel@...r.kernel.org,
        "Luke D. Jones" <luke@...nes.dev>
Subject: [PATCH v2 0/1] platform/x86: asus-wmi: disable USB0 hub on ROG Ally before suspend

This is a fix for the ROG Ally not being able to use the N-Key device after a suspend/resume cycle.

The root of the issue is that ASUS changed the MCU firmware to dfisconnect the USB0 hub when the
screen is switched off during the s2idle suspend path. I tried many many different tactics to try
and get this s2idle part to work but it seems there are races between this and other subsystems.

What has so far been reliable and consistent is a manual call to the CSEE method that is called in
ACPI by the Microsoft DSM screen off path followed by a short sleep in asus-wmi. The PM prepare
hook looks to be the earliest possible place. A sleep that is too long ends up with USB subsystem
registering a disconnect, and thus on resume the device paths change. Too short and it is unreliable.

Some discussion regarding this mess is at https://gitlab.freedesktop.org/drm/amd/-/issues/2719#note_2181402

Changelog:
- v2:
  - Emit dev_err(), but don't return error codes
  - Add check for MCU power save mode being enabled and disable if active
  - General cleanup and rename some vars/funcs
- v1: https://lore.kernel.org/all/20231124082749.23353-1-luke@ljones.dev/

Luke D. Jones (1):
  platform/x86: asus-wmi: disable USB0 hub on ROG Ally before suspend

 drivers/platform/x86/asus-wmi.c            | 50 ++++++++++++++++++++++
 include/linux/platform_data/x86/asus-wmi.h |  3 ++
 2 files changed, 53 insertions(+)

-- 
2.43.0