linux-kernel - Re: [PATCH v2 1/3] iio: light: Add driver for ap3216c

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] [day] [month] [year] [list]

Date:   Sat, 9 Mar 2019 17:33:46 +0000
From:   Jonathan Cameron <jic23@...nel.org>
To:     Sven Van Asbroeck <thesven73@...il.com>
Cc:     Robert Eshleman <bobbyeshleman@...il.com>,
        Hartmut Knaack <knaack.h@....de>,
        Lars-Peter Clausen <lars@...afoo.de>,
        Peter Meerwald-Stadler <pmeerw@...erw.net>,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
        linux-iio@...r.kernel.org
Subject: Re: [PATCH v2 1/3] iio: light: Add driver for ap3216c

On Mon, 4 Mar 2019 11:25:01 -0500
Sven Van Asbroeck <thesven73@...il.com> wrote:

> On Sun, Mar 3, 2019 at 9:38 AM Jonathan Cameron <jic23@...nel.org> wrote:
> >
> > Hmm. Just been thinking a bit about the events on here and wondered
> > if it is possible to mask them through careful use of the threshold
> > values - i.e. can we stop the hardware generating the interrupts for
> > the ones we don't want. It would be unusual for hardware to be
> > designed where this wasn't possible.  
> 
> Excellent point! People with power / battery constraints take a dim view of
> receiving interrupts when no-one wants them. So disabling them in h/w
> is definitely the way to go, if possible.
> 
> And yes, this also makes a non-issue of thresh_en visibility concerns, if any.
> 
> >
> > Alternatively if you have a scope or equivalent to verify if it is doing
> > these as a multi byte read and working that would be even better.
> > It is not uncommon for hardware to implement fairly standard i2c features
> > like this and not document them because they weren't what the test code
> > the docs writer got given does! (may not be true here of course)  
> 
> Or alternatively, the current chip rev supports undocumented multi-reads,
> and the next revision silently drops support, thereby breaking the driver...
> Been there, done that, got the T-shirt :(

Indeed it's a risk.  Sadly hardware guys never have a 'we mustn't break
software' rule like we do for userspace! 

Jonathan