linux-kernel - Re: [PATCH] arm64: dts: renesas: ulcb-kf: add 9-asix sensor device

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:   Wed, 26 Jan 2022 18:27:58 +0300
From:   Nikita Yushchenko <nikita.yoush@...entembedded.com>
To:     Geert Uytterhoeven <geert@...ux-m68k.org>
Cc:     Magnus Damm <magnus.damm@...il.com>,
        Rob Herring <robh+dt@...nel.org>,
        Linux-Renesas <linux-renesas-soc@...r.kernel.org>,
        "open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS" 
        <devicetree@...r.kernel.org>,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] arm64: dts: renesas: ulcb-kf: add 9-asix sensor device

>> Interrupt definition is not added yet, because the interrupt lines of
>> lsm9ds0 are pulled to VCC on the board, which implies need for
>> active-low configuration. But st_sensors drivers currently can't work
>> with active-low interrupts on this chip.
> 
> That's unfortunate, as DT describes hardware, not limitations of the
> software stack.

Unfortunately, if interrupt definition is added, driver does wrong things and causes board hang.

>> +                               vdd-supply = <&accel_3v3>;
>> +                               vddio-supply = <&accel_3v3>;
> 
> According to the bindings, the supplies are not required, so you can
> leave them out? Or are the bindings wrong?

If supplies are not defined, warning messages about dummy regulator are logged.

> (The bindings also say "interrupts: maxItems 2", while the "interrupts:
>   description" says up to three interrupts, doh...)

Chip has 3 interrupt outputs. On KF board, all those are ANDed together and result connected to SoC's 
gpio that is expected to be used as a shared active-low interrupt. Driver currently claims that this 
chip does not support active-low interrupts. Per datasheet, this is not true. But driver's way to set up 
interrupt registers does not scale to the case when interrupts have to be configured by different bits 
in several registers, that part of the driver has to be somehow rewritten. I guess nobody has ever tried 
to make these drivers (st_*) to drive a compound device (accel+gyro) with interrupts.

At the same time, the device is perfectly useful without interrupts, and that is how it is enabled in 
the vendor BSP.

Nikita