linux-kernel - Re: [tpmdd-devel] tpm, tpm_tis: fix tpm_tis ACPI detection issue with TPM 2.0

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: <20151023130121.GB5705@fau-staff-11-84.rrze.uni-erlangen.de>
Date:	Fri, 23 Oct 2015 15:01:21 +0200
From:	Valentin Rothberg <rothberg@...fau.de>
To:	Jarkko Sakkinen <jarkko.sakkinen@...ux.intel.com>
Cc:	Andreas Ziegler <andreas.ziegler@....de>,
	Paul Bolle <pebolle@...cali.nl>,
	"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
	tpmdd-devel@...ts.sourceforge.net
Subject: Re: [tpmdd-devel] tpm, tpm_tis: fix tpm_tis ACPI detection issue
 with TPM 2.0

On Oct 21 '15 18:58, Jarkko Sakkinen wrote:
> On Tue, Oct 20, 2015 at 05:58:35PM +0300, Jarkko Sakkinen wrote:
> > On Tue, Oct 20, 2015 at 01:49:02PM +0200, Andreas Ziegler wrote:
> > > Hi Jarkko,
> > > 
> > > your patch "tpm, tpm_tis: fix tpm_tis ACPI detection issue with TPM 2.0"
> > > showed up as commit 399235dc6e95 in linux-next today (that is,
> > > next-20151020). I noticed it because we (a research group from
> > > Erlangen[0]) are running daily checks on linux-next.
> > > 
> > > Your commit creates the following structure of #ifdef blocks in
> > > drivers/char/tpm/tpm_tis.c following line 1088:
> > > 
> > >  #ifdef CONFIG_ACPI
> > >  ...
> > >  #ifdef CONFIG_PNP
> > >  ...
> > >  #endif
> > >  ...
> > >  #endif
> > > 
> > > Looking at the definition of CONFIG_ACPI at drivers/acpi/Kconfig, line
> > > 5, we see that ACPI unconditionally selects PNP, meaning that CONFIG_PNP
> > > is always enabled if CONFIG_ACPI has been enabled.
> > > Thus, the inner #ifdef statement can never evaluate to 'false' if the
> > > outer #ifdef evaluates to true (i.e., CONFIG_ACPI is enabled), and
> > > hence, the #ifdef is unnecessary.
> > > 
> > > The same situation holds for the nested structure following line 1124,
> > > where the #ifdef CONFIG_PNP at line 1129 is unnecessary.
> > > 
> > > Is this correct or did we miss something?
> > 
> > Good catch. Shoud I send a separate fix for this? Thanks for pointing
> > this out.
> 
> In all I would cases do a separate fix and do not fixup the original
> patchs because I wouldn't consider this a regression.
> 
> The next question is: will it always be like this? Can I safely assume
> that ACPI will always select PNP unconditionally? This is so minor
> cosmetic glitch in the code that I'm getting second thoughts whether I
> should anything to this or not.

Any option has certain benefits.  In the current form, a reader might
(falsely) assume that the block is configurable.  However, if the
constraints of ACPI or PNP change, the situation might look different.

Best regards,
 Valentin
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/