linux-kernel - Re: [PATCH 2/2] hwmon: (ina3221) Add operating mode 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] [day] [month] [year] [list]

Message-ID: <20181011195321.GA5567@Asurada-Nvidia.nvidia.com>
Date:   Thu, 11 Oct 2018 12:53:23 -0700
From:   Nicolin Chen <nicoleotsuka@...il.com>
To:     Guenter Roeck <linux@...ck-us.net>
Cc:     jdelvare@...e.com, linux-hwmon@...r.kernel.org,
        linux-kernel@...r.kernel.org, corbet@....net,
        linux-doc@...r.kernel.org
Subject: Re: [PATCH 2/2] hwmon: (ina3221) Add operating mode support

On Thu, Oct 11, 2018 at 12:50:25PM -0700, Guenter Roeck wrote:
> Hi Nicolin,
> 
> On Thu, Oct 11, 2018 at 12:36:59PM -0700, Nicolin Chen wrote:
> > On Thu, Oct 11, 2018 at 12:31:52PM -0700, Guenter Roeck wrote:
> > 
> > > > One more question here, and this might sound a bit abuse of using
> > > > the existing hwmon ABI: would it sound plausible to you that the
> > > > driver powers down the chip when all three channels get disabled
> > > > via in[123]_enable nodes? :)
> > > > 
> > > 
> > > I would not call that an abuse, no.
> > 
> > Hmm..do you mean that you aren't in favor of powering down the chip
> > after all channels get disabled?
> > 
> No, I was trying to say that I would be ok with powering down the chip.

Great!

> > I was thinking about having pm_runtime_get_sync()/put() for channel
> > enabling/disabling routine of in[123]_enable.
> > 
> 
> Not sure if that would work. It might end up waking the chip when a
> sysfs attribute is accessed. It might be worth a try, though.
> 
> It might also be possible to utilize userspace runtime attributes,
> like setting runtime_enabled and setting the idle time before the sensor
> shuts down. It would probably be necessary to implement not only
> activating the sensor, though - we would also need to to ensure that
> the first reading after activation waits until the first read is
> complete.

That's true. Thanks for the input!

Nicolin