linux-kernel - Re: [PATCH v2 1/7] iio: adc: max1027: Add debugfs register read support

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, 7 Oct 2019 12:00:01 +0200
From:   Miquel Raynal <miquel.raynal@...tlin.com>
To:     Jonathan Cameron <jic23@...nel.org>
Cc:     Hartmut Knaack <knaack.h@....de>,
        Lars-Peter Clausen <lars@...afoo.de>,
        Peter Meerwald-Stadler <pmeerw@...erw.net>,
        Rob Herring <robh+dt@...nel.org>,
        Mark Rutland <mark.rutland@....com>,
        <devicetree@...r.kernel.org>, linux-iio@...r.kernel.org,
        linux-kernel@...r.kernel.org,
        Thomas Petazzoni <thomas.petazzoni@...tlin.com>
Subject: Re: [PATCH v2 1/7] iio: adc: max1027: Add debugfs register read
 support

Hi Jonathan,

Jonathan Cameron <jic23@...nel.org> wrote on Sun, 6 Oct 2019 11:04:24
+0100:

> On Thu,  3 Oct 2019 19:33:55 +0200
> Miquel Raynal <miquel.raynal@...tlin.com> wrote:
> 
> > Until now, only write operations were supported. Force two bytes read
> > operation when reading from this register (might be wrong when reading
> > the temperature, but will work with any other value).  
> 
> That's worrying as comments go.  Just return an error on the temperature
> register if it's going to do the wrong thing.

Actually the debugfs_reg_access hook is supposedly stateless. When
reading registers I don't know what I am reading because the "source" is
selected during the write operation, so I have no reliable way to know
what I am reading.

I set the read length to 2 bytes because most of the "atomic"reads are
two bytes and it allows us to test various commands directly from
userspace and read meaningful values. This is a limitation as:
* Voltage 'atomic' reads are 2 bytes
* Temperature 'atomic' reads are 2 bytes but never come alone (usually
  one voltage input of 2B will follow).
* Any other 'condensed' input will be more than 2 bytes, ie. several
  voltage values in one go.

In any case, doing a software reset of the chip will turn it back
into a working state no matter what was requested/read.

For me, 2-byte reads is a "good enough" solution that will work with
almost all the simplest ('atomic') SPI operations, but if you think
limiting to 2-bytes access is a problem (right now there is only write
access, which is kind of useless on its own) then let's drop the patch.
But I wanted to contribute it because it really helped me during the
development. 

Thanks,
Miquèl