linux-kernel - Re: [PATCH v2 2/2] hwmon: ibmpowernv: Add attributes to enable/disable sensor groups

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:   Wed, 4 Jul 2018 14:42:56 -0700
From:   Guenter Roeck <linux@...ck-us.net>
To:     Shilpasri G Bhat <shilpa.bhat@...ux.vnet.ibm.com>,
        mpe@...erman.id.au
Cc:     linuxppc-dev@...ts.ozlabs.org, linux-hwmon@...r.kernel.org,
        linux-kernel@...r.kernel.org, ego@...ux.vnet.ibm.com
Subject: Re: [PATCH v2 2/2] hwmon: ibmpowernv: Add attributes to
 enable/disable sensor groups

On 07/04/2018 09:53 AM, Shilpasri G Bhat wrote:
> Hi Guenter,
> 
> Thanks for reviewing the patch.
> On 07/04/2018 08:16 PM, Guenter Roeck wrote:
>>> +        /* Disable if last sensor in the group */
>>> +        send_command = true;
>>> +        for (i = 0; i < sg->nr_sensor; i++) {
>>> +            struct sensor_data *sd = sg->sensors[i];
>>> +
>>> +            if (sd->enable) {
>>> +                send_command = false;
>>> +                break;
>>> +            }
>>
>> This is weird. So there are situations where a request to disable
>> a sensor is accepted, but effectively ignored ? Shouldn't that
>> return, say, -EBUSY ?
> 
> This is because we do not support per-sensor enable/disable. We can only
> enable/disable at a sensor-group level.
> 
> This patch follows the semantic to disable a sensor group iff all the sensors
> belonging to that group have been disabled. Otherwise the sensor alone is marked
> to be disabled and returns -ENODATA on reading it.
> 
> And a sensor group will be enabled if any of the sensor in that group is enabled.
> 

In similar situations, where setting one attribute affects others, a common solution
is to make only the first attribute writable and have it affect all the others.
I think that would make sense here as well, and it would be much simpler to implement.

Guenter

> I will make changes to the remaining code according to your suggestion.
> 
> Thanks and Regards,
> Shilpa
> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-hwmon" in
> the body of a message to majordomo@...r.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>