linux-kernel - Fwd: [amd-sfh] Interrupt disabling (b300667b) disables working HP ProBook x360 435 G7 accelerometer

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 for Android: free password hash cracker in your pocket

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

Message-ID: <14e3c405-5f80-4b73-b291-e84bdd6b240f@gmail.com>
Date:   Sun, 5 Nov 2023 20:38:07 +0700
From:   Bagas Sanjaya <bagasdotme@...il.com>
To:     Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
        Linux Regressions <regressions@...ts.linux.dev>,
        Linux Input Devices <linux-input@...r.kernel.org>
Cc:     Basavaraj Natikar <Basavaraj.Natikar@....com>,
        Jiri Kosina <jkosina@...e.cz>,
        Benjamin Tissoires <benjamin.tissoires@...hat.com>,
        Mauricio Collares <mauricio@...lares.org>
Subject: Fwd: [amd-sfh] Interrupt disabling (b300667b) disables working HP
 ProBook x360 435 G7 accelerometer

Hi,

I notice a regression report on Bugzilla [1]. Quoting from it:

> I have a HP ProBook x360 435 G7 (Ryzen 4700U). On GNOME Wayland with
> the Screen Autorotate extension
> (https://extensions.gnome.org/extension/4191/screen-autorotate/), I
> confirmed by bisection that b300667b33b2b5a2c8e5f8f22826befb3d7f4f2b
> caused autorotation to stop working.
> 
> More specifically, before this commit (on aa0b724a2bf041036e56cbb3b4b3afde7c5e7c9e):
> 
> * `sudo iio-sensor-proxy -v -r` stays running (doesn't exit immediately) and prints nothing
> * `monitor-sensor` starts up by saying "=== Has accelerometer (orientation: normal)" and prints "    Accelerometer orientation changed: bottom-up" when I physically rotate the screen.
> * `udevadm info --export-db` contains the accelerometer/gyroscope sensor info I've pasted in the attachment.
> 
> After b300667b33b2b5a2c8e5f8f22826befb3d7f4f2b:
> 
> * `sudo iio-sensor-proxy -v -r` exits immediately.
> * `monitor-sensor` prints "    Waiting for iio-sensor-proxy to appear" on startup and does nothing when I rotate the screen.
> * The sensors in the attachment are no longer listed in `udevadm info --export-db`.
> * The `hid_sensor_gyro_3d` and `hid_sensor_accel_3d` kernel modules are no longer loaded automatically, and loading them does nothing.
> 
> Reverting b300667b33b2b5a2c8e5f8f22826befb3d7f4f2b on top of 6.6 also makes screen autorotation work again. Whenever screen autorotation works, `iio_generic_buffer --device-num 2 -A -c 100` reports good accelerometer values.
> 
> (The following is essentially the contents of bug #212625 comment #46)
> 
>> If you revert that commit on top of 6.6 does it start working?  Also; can
>> you still get into the deepest s2idle sleep state in 6.6 with it reverted? 
>> You can check with /sys/power/suspend_stats/last_hw_sleep.
> 
> I don't have a /sys/power/suspend_stats/last_hw_sleep file (tested on stock 6.6, as well as on 6.6 + revert of b300667b33b2b5a2c8e5f8f22826befb3d7f4f2b), but if I leave my computer idle for a while it enters standby and I see
> 
> Nov 03 08:47:29 hp kernel: PM: suspend entry (deep)
> 
> in journalctl, and the computer enters deep sleep (need to press power button to wake it up). Also, the content of /sys/power/mem_sleep is "s2idle [deep]". In bug ug #212615 I also attached the report produced by amd_s2idle.py run on stock 6.6 in case it is relevant (I can also attach it for 6.6 + revert if desired).
> 

See Bugzilla for the full thread and attached udevadm output.

Anyway, I'm adding this regression to regzbot:

#regzbot introduced: b300667b33b2b5 https://bugzilla.kernel.org/show_bug.cgi?id=218104

Thanks.

[1]: https://bugzilla.kernel.org/show_bug.cgi?id=218104

-- 
An old man doll... just what I always wanted! - Clara