Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Tue, 12 Jul 2022 19:10:56 +0000
From:   Sean Christopherson <seanjc@google.com>
To:     Vitaly Kuznetsov <vkuznets@redhat.com>
Cc:     Maxim Levitsky <mlevitsk@redhat.com>, kvm@vger.kernel.org,
        Paolo Bonzini <pbonzini@redhat.com>,
        Wanpeng Li <wanpengli@tencent.com>,
        Jim Mattson <jmattson@google.com>, linux-kernel@vger.kernel.org
Subject: Re: [PATCH] KVM: nVMX: Always enable TSC scaling for L2 when it was
 enabled for L1
Message-ID: <Ys3HQDuAO63uOV3f@google.com>
References: <20220712135009.952805-1-vkuznets@redhat.com>
 <ee479e42605d3ed3276b66da69179dbfbcb05dbc.camel@redhat.com>
 <871quqpcq4.fsf@redhat.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <871quqpcq4.fsf@redhat.com>
Precedence: bulk

On Tue, Jul 12, 2022, Vitaly Kuznetsov wrote:
> Maxim Levitsky <mlevitsk@redhat.com> writes:
> 
> > On Tue, 2022-07-12 at 15:50 +0200, Vitaly Kuznetsov wrote:
> >> Windows 10/11 guests with Hyper-V role (WSL2) enabled are observed to
> >> hang upon boot or shortly after when a non-default TSC frequency was
> >> set for L1. The issue is observed on a host where TSC scaling is
> >> supported. The problem appears to be that Windows doesn't use TSC
> >> frequency
> 
> ^^^ scaling ^^^
> 
> >> for its guests even when the feature is advertised and KVM
> >> filters SECONDARY_EXEC_TSC_SCALING out when creating L2 controls from
> >> L1's. This leads to L2 running with the default frequency (matching
> >> host's) while L1 is running with an altered one.
> >
> > Ouch.
> >
> > I guess that needs a Fixes tag?
> >
> > Fixes: d041b5ea93352b ("KVM: nVMX: Enable nested TSC scaling")
> >
> 
> I dismissed that because prior to d041b5ea93352b SECONDARY_EXEC_TSC_SCALING
> was filtered out in nested_vmx_setup_ctls_msrs() but now I think I was
> wrong, SECONDARY_EXEC_TSC_SCALING was likely kept in VMCS02 regardless
> of that. Will add in v2.

Yep, it would have been kept in vmcs02 even though the feature wasn't advertised
to the L1 VMM.  A Cc for stable is warranted as well.

I added this (with the tags and s/frequency/scaling) to the queue of patches for
5.20 I have lined up for Paolo to consume on his return.  Paolo and I haven't
hashed out how we'll actually manage anything, i.e. my list is speculative, but
unless you hear otherwise, no need to send a v2.