Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Date:   Mon, 10 Dec 2018 19:13:28 +0100
From:   Borislav Petkov <bp@alien8.de>
To:     Paolo Bonzini <pbonzini@redhat.com>
Cc:     Eduardo Habkost <ehabkost@redhat.com>,
        =?utf-8?B?SsO2cmcgUsO2ZGVs?= <joro@8bytes.org>,
        Andre Przywara <andre.przywara@arm.com>,
        kvm ML <kvm@vger.kernel.org>,
        lkml <linux-kernel@vger.kernel.org>,
        Tom Lendacky <thomas.lendacky@amd.com>
Subject: Re: kvm: RDTSCP on AMD
Message-ID: <20181210181328.GA762@zn.tnic>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Disposition: inline
User-Agent: Mutt/1.10.1 (2018-07-13)
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

Reviving an old thread here.

On Wed, Jul 06, 2016 at 11:27:16PM +0200, Paolo Bonzini wrote:
> On 06/07/2016 19:34, Eduardo Habkost wrote:
> >> > Nothing is needed in the kernel actually.  You can skip the intercept
> >> > by running the guest with MSR_TSC_AUX set to the guest's expected value.
> >> >  Which KVM does, except that it's botched so I need to apply the
> >> > patch in https://lkml.org/lkml/2016/4/13/802.
> > Do you mean -cpu Opteron_G*,+rdtscp will be buggy on Linux v4.5?
> > (v4.5 reports rdtscp as supported in GET_SUPPORTED_CPUID)
> > 
> > Can we do something to make QEMU detect the buggy kernel before
> > allowing rdtscp to be enabled, or should we just tell people to
> > upgrade their kernel?
> 
> We usually just tell people to use the latest stable kernel.
> 
> Adding new CPU models is not a big deal, in fact it's almost easier than
> getting compat properties right. :)

Ok, can we finally revert

  33b5e8c03ae7 ("target-i386: Disable rdtscp on Opteron_G* CPU models")

in the qemu tree?

Three years should be enough by now for

  46896c73c1a4 ("KVM: svm: add support for RDTSCP")

to have percolated downstream.

Btw, its commit message talks about Linux not using RDTSCP but that will
change soon. :-)

Oh, and the EPYC qemu CPU model has CPUID_EXT2_RDTSCP and works just
fine when patching in RDTSCP:

[    0.543197] apply_alternatives: feat: 3*32+18, old: (read_tsc+0x0/0x10 (ffffffff8101d1c0) len: 5), repl: (ffffffff824e6d33, len: 5), pad: 3
[    0.544448] ffffffff8101d1c0: old_insn: 0f 31 90 90 90
[    0.545023] ffffffff824e6d33: rpl_insn: 0f ae e8 0f 31
[    0.545598] ffffffff8101d1c0: final_insn: 0f ae e8 0f 31
[    0.546193] apply_alternatives: feat: 1*32+27, old: (read_tsc+0x0/0x10 (ffffffff8101d1c0) len: 5), repl: (ffffffff824e6d38, len: 3), pad: 3
[    0.547195] ffffffff8101d1c0: old_insn: 0f ae e8 0f 31
[    0.547775] ffffffff824e6d38: rpl_insn: 0f 01 f9
[    0.548307] ffffffff8101d1c0: final_insn: 0f 01 f9 66 90

That final_insn which gets patched to by the alternatives is

ffffffff8101d1c0:       0f 01 f9                rdtscp
ffffffff8101d1c3:       66 90                   xchg %ax,%ax

Thx.

-- 
Regards/Gruss,
    Boris.

Good mailing practices for 400: avoid top-posting and trim the reply.