linux-kernel - Re: [PATCH 2/5] hwmon: (ina3221) Return -ENODATA for two alarms attributes

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]

Message-ID: <20181017203917.GB15941@Asurada-Nvidia.nvidia.com>
Date:   Wed, 17 Oct 2018 13:39:17 -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
Subject: Re: [PATCH 2/5] hwmon: (ina3221) Return -ENODATA for two alarms
 attributes

On Wed, Oct 17, 2018 at 12:46:05PM -0700, Guenter Roeck wrote:
> On Tue, Oct 16, 2018 at 06:24:23PM -0700, Nicolin Chen wrote:
> > There is nothing critically wrong to read these two attributes
> > without having a is_enabled() check at this point. But reading
> > the MASK_ENABLE register would clear the CVRF bit according to
> > the datasheet. So it'd be safer to fence for disabled channels
> > in order to add pm runtime feature.
> > 
> > Signed-off-by: Nicolin Chen <nicoleotsuka@...il.com>
> > ---
> >  drivers/hwmon/ina3221.c | 2 ++
> >  1 file changed, 2 insertions(+)
> > 
> > diff --git a/drivers/hwmon/ina3221.c b/drivers/hwmon/ina3221.c
> > index d61688f04594..3e98b59108ee 100644
> > --- a/drivers/hwmon/ina3221.c
> > +++ b/drivers/hwmon/ina3221.c
> > @@ -200,6 +200,8 @@ static int ina3221_read_curr(struct device *dev, u32 attr,
> >  		return 0;
> >  	case hwmon_curr_crit_alarm:
> >  	case hwmon_curr_max_alarm:
> > +		if (!ina3221_is_enabled(ina, channel))
> > +			return -ENODATA;
> 
> Makes sense, but can you check what the sensors command does with this ?

Not quite understanding the question. Do you mean the user case
causing the race condition -- wiping out the CVRF bit?

> If it bails out I'd rather have the code return 0 and no error (after all,
> the sensor is disabled, so any alarm would be bogus).

That's true. Since they are alert flags, should return 0. I will
fix it.

Thanks