linux-kernel - Re: [GIT PULL] platform-drivers-x86 for 5.11-1

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: <c24d791c-1f2f-7f94-84cd-450cbdba6ce3@redhat.com>
Date:   Wed, 16 Dec 2020 10:21:20 +0100
From:   Hans de Goede <hdegoede@...hat.com>
To:     Linus Torvalds <torvalds@...ux-foundation.org>,
        Alexander Duyck <alexander.h.duyck@...ux.intel.com>,
        "David E. Box" <david.e.box@...ux.intel.com>,
        Lee Jones <lee.jones@...aro.org>
Cc:     Andy Shevchenko <andy@...radead.org>,
        Mark Gross <mark.gross@...el.com>,
        LKML <linux-kernel@...r.kernel.org>
Subject: Re: [GIT PULL] platform-drivers-x86 for 5.11-1

Hi,

On 12/16/20 1:17 AM, Linus Torvalds wrote:
> On Mon, Dec 14, 2020 at 4:43 AM Hans de Goede <hdegoede@...hat.com> wrote:
>>
>> - New Intel PMT telemetry and crashlog drivers
> 
> These have _very_ annoying Kconfig setups.
> 
> First it asks about INTEL_PMT support.
> 
> If you say no, it then asks about INTEL_PMT_CLASS, INTEL_PMT_TELEMETRY
> and INTEL_PMT_CRASHLOG support.

Yeah that is bad, sorry for not spotting that (I answer M/Y everywhere
so that I at least compile test everything).

> I've pulled this,

Thanks.

> but I really wish the PMT support understood that
> whole "No means no" thing.

Agreed. I will make sure to get a fix to you for this sometime
during the 5.11 cycle.

Alexander and/or David I believe that the best solution for this
would be to replace the:

        select INTEL_PMT_CLASS

In the INTEL_PMT_TELEMETRY and INTEL_PMT_CRASHLOG Kconfig entries
with "depends on". Also as a general comment for the future, select
should typically be used with non-user-selectable options. E.g.
you could also have made INTEL_PMT_CLASS non-user-selectable by
changing:

config INTEL_PMT_CLASS
        tristate "Intel Platform Monitoring Technology (PMT) Class driver"

To:

config INTEL_PMT_CLASS
        tristate

Having both INTEL_PMT_CLASS user-selectable and using select on
it is bad. I believe that in this case changing the 2 "select"-s
to "depend on"-s is best.

Regards,

Hans