linux-kernel - Re: [RESEND PATCH v3 1/2] power: supply: add input voltage limit property

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: <20190502210138.ekrjvg4jex5x2tbo@earth.universe>
Date:   Thu, 2 May 2019 23:01:38 +0200
From:   Sebastian Reichel <sebastian.reichel@...labora.com>
To:     Enric Balletbo i Serra <enric.balletbo@...labora.com>
Cc:     Pavel Machek <pavel@....cz>, linux-pm@...r.kernel.org,
        Sameer Nanda <snanda@...omium.org>, bleung@...omium.org,
        rjw@...ysocki.net, gwendal@...omium.org,
        linux-kernel@...r.kernel.org, Len Brown <len.brown@...el.com>,
        groeck@...omium.org, Adam.Thomson.Opensource@...semi.com,
        kernel@...labora.com
Subject: Re: [RESEND PATCH v3 1/2] power: supply: add input voltage limit
 property

Hi,

On Tue, Apr 16, 2019 at 10:42:30AM +0200, Enric Balletbo i Serra wrote:
> On 16/4/19 9:19, Pavel Machek wrote:
> >> This patch exposes a new property, similar to input current limit, to
> >> re-configure the maximum voltage from the external supply at runtime
> >> based on system-level knowledge or user input.
> > 
> > Well, and I suspect it should expose input power limit, not input
> > voltage limit.
> 
> Oh, ok, I thought we were agree that input voltage had sense after had some
> discussion in v3. Seems that no, let me try to give you another example...
> 
> > DC-DC convertor efficiency normally does not much depend on input
> > voltage....
> 
> As we said we have a heat "problem" due the internal voltage conversions.
> 
> Lets assume you have a linear regulator instead with a Vin range from 5V to 9V
> and we want an output of 3.3V/1A
>
> For 9V:
>  Input power : P(in) = 9V x 1A = 9W
>  Output power: P(out) = 3.3V x 1A = 3.3W
>  Regulator power dissipated: P(reg) = P(in) - P(out) = 9W - 3.3W = 5.7W
> 
> For 5V:
>  Input power : P(in) = 5V x 1A = 5W
>  Output power: P(out) = 3.3V x 1A = 3.3W
>  Regulator power dissipated: P(reg) = P(in) - P(out) = 5W - 3.3W = 1,7W
> 
> In the first case the regulator needs to dissipate more power, hence the
> temperature is greater than the second case.

I would be surprised, if a linear regulator is being used in this
place :) That would basically render functionality of higher voltage
completley useless and a good reason to always limit to 5V. For the
generic case I agree with Pavel, that control over the power (voltage
* current) is the better choice. Still I believe it makes sense to
have a control knob for the voltage available, since some hardware
designs suck.

For example the bad hardware design might be the remote side,
that has issues providing some voltages and this would make it
possible to debug that.

-- Sebastian

Download attachment "signature.asc" of type "application/pgp-signature" (834 bytes)