Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
MIME-Version: 1.0
References: <20200615063837.GA14668@zn.tnic> <20200714121955.GA2080@chrisdown.name>
 <20200714154728.GA3101@nazgul.tnic> <20200714160448.GC2080@chrisdown.name>
 <CACdnJuvfhjMNQUYNYWpPMfwTE3xHi7UNPm7HEwUMv_1F3KT4gA@mail.gmail.com>
 <20201117210018.GA4247@weirdfishes> <CACdnJusV0QBy16Njge3-KR+gxcdTVXF3aO4B-z-QN+e0uhDnkQ@mail.gmail.com>
In-Reply-To: <CACdnJusV0QBy16Njge3-KR+gxcdTVXF3aO4B-z-QN+e0uhDnkQ@mail.gmail.com>
From:   Matthew Garrett <mjg59@google.com>
Date:   Tue, 17 Nov 2020 13:22:49 -0800
Message-ID: <CACdnJuvXhVnoeK223pA3VmaRAQ=utWZei6RpQrkJyBcdpMxkOg@mail.gmail.com>
Subject: Re: [PATCH -v2.1] x86/msr: Filter MSR writes
To:     Mathieu Chouquet-Stringer <me@mathieu.digital>,
        Matthew Garrett <mjg59@google.com>,
        Chris Down <chris@chrisdown.name>,
        Borislav Petkov <bp@alien8.de>,
        Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
        sean.j.christopherson@intel.com, tony.luck@intel.com,
        Linus Torvalds <torvalds@linux-foundation.org>,
        "the arch/x86 maintainers" <x86@kernel.org>, kernel-team@fb.com,
        Len Brown <lenb@kernel.org>
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

On Tue, Nov 17, 2020 at 1:21 PM Matthew Garrett <mjg59@google.com> wrote:
>
> On Tue, Nov 17, 2020 at 1:00 PM Mathieu Chouquet-Stringer
> <me@mathieu.digital> wrote:
>
> > I'm late to the party but it seems allowing MSR_IA32_ENERGY_PERF_BIAS
> > has the downside of flagging the kernel as tainted without telling you
> > why if you use something like x86_energy_perf_policy (from
> > tools/power/x86/x86_energy_perf_policy) which itself is used by tuned.
>
> I initially pushed back against a kernel interface for
> MSR_IA32_ENERGY_PERF_BIAS (cc: Len Brown, who tried mightily to
> convince me I was wrong here) on the grounds that it was exporting an
> implementation detail rather than providing a generic interface, and
> that it was something that could be done via userland instead. I
> thought we'd end up with more examples of similar functionality and
> could tie it into something more reasonable - history has proven me
> wrong on that. I think it's probably reasonable to dust off the driver
> that Len submitted however many years ago and push that into the
> kernel now.

But ha ok based on Borislav's response it looks like someone's already
done that.