linux-kernel - Re: [lm-sensors] lm90 driver no longer working on PCs in 3.13

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]

Date:	Mon, 27 Jan 2014 10:34:37 +0000
From:	Mark Brown <broonie@...nel.org>
To:	Guenter Roeck <linux@...ck-us.net>
Cc:	Jean Delvare <khali@...ux-fr.org>,
	Liam Girdwood <lgirdwood@...il.com>, Wei Ni <wni@...dia.com>,
	linux-kernel@...r.kernel.org,
	LM Sensors <lm-sensors@...sensors.org>
Subject: Re: [lm-sensors] lm90 driver no longer working on PCs in 3.13

On Sun, Jan 26, 2014 at 08:16:37PM -0800, Guenter Roeck wrote:
> On 01/26/2014 03:51 PM, Mark Brown wrote:
> >On Sun, Jan 26, 2014 at 02:04:06PM -0800, Guenter Roeck wrote:

> >>I think I have a better idea: Surround the regulator code, or at least
> >>its error handling, in the lm90 driver with

> >>	if (IS_ENABLED(CONFIG_OF)) {
> >>	}

> >>Would that be ok ? If yes I'll submit a patch. I'll do the same in
> >>another driver I am working on.

> >That's not going to have the desired effect in cases where DT is built
> >into the kernel but not in use on the current system (which is a
> >configuration that gets used) and will remove error handling for non-DT
> >systems that do have regulators set up.  There's not the relationship
> >between this and DT that you seem think there is...

> >Besides, if we're going to do a bodge like that we should do it in the
> >core and not in individual callers.

> Then it appears the only remedy at this time is to revert the patch.

Why - that seems like a bit of a leap?  The first order problem with
what you're doing there is that conditionalising this on DT being built
into the kernel is the wrong conditional for several reasons, if you do
want to use an approach like this that at least needs to be changed.

Download attachment "signature.asc" of type "application/pgp-signature" (837 bytes)