linux-kernel - Re: BMI160 accelerometer on AyaNeo tablet

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

Date:   Mon, 18 Oct 2021 10:40:33 +0300
From:   Andy Shevchenko <andy.shevchenko@...il.com>
To:     Jonathan Cameron <jic23@...nel.org>,
        Hans de Goede <hdegoede@...hat.com>
Cc:     Maxim Levitsky <maximlevitsky@...il.com>,
        linux-iio <linux-iio@...r.kernel.org>,
        Lars-Peter Clausen <lars@...afoo.de>,
        Andy Shevchenko <andriy.shevchenko@...ux.intel.com>,
        LKML <linux-kernel@...r.kernel.org>
Subject: Re: BMI160 accelerometer on AyaNeo tablet

+Cc: Hans

On Mon, Oct 18, 2021 at 6:41 AM Jonathan Cameron <jic23@...nel.org> wrote:
>
> On Sat, 16 Oct 2021 19:27:50 +0300
> Maxim Levitsky <maximlevitsky@...il.com> wrote:
>
> >  BMI160: AYA NEA accelometer ID

accelerometer

> >     On AYA NEO, the accelerometer is BMI160 but it is exposed
> >     via ACPI as 10EC5280
> >
> >     Signed-off-by: Maxim Levitsky <mlevitsk@...hat.com>
>
> I guess it is hopelessly optimistic to hope that we could let someone
> at the supplier know that's a totally invalid ACPI id and that they
> should clean up their act.
>
> Curiously it looks like a valid PCI ID pair though for a realtek device.
>
> Ah well.  Applied to the iio-togreg branch of iio.git and pushed out
> as testing to see if 0-day can find any issues with it.

NAK. And I explain below why and how to make progress with it.

The commit message should contain at least the link to the DSDT and
official technical description of the platform. Besides that, it
should have a corresponding comment near to the ID in the code.

On top of that, in particular to this case, the ID is very valid from
the ACPI specification point of view, but in this case it's a
representation of the PCI ID 10ec:5280 which is Realtek owned. So, we
need to hear (okay in reasonable time) from Realtek (I believe they
are active in the Linux kernel) and that OEM.

I hardly believe that Realtek has issued a special ID from the range
where mostly PCIe ports or so are allocated, although it's possible.
We need proof.

What I believe is the case here is that OEMs are just quite diletants
in ACPI and firmware and they messed up with BIOS somehow that it
issued the ID for the device.
There are also two other possibilities: OEM stole the ID (deliberately
or accidentally), or the device is not just gyro, but something which
contains gyro.

As to the last paragraph, see above, we must see DSDT. Without it I
have a strong NAK.

P.S. Jonathan, please do not be so fast next time with ACPI IDs.

-- 
With Best Regards,
Andy Shevchenko