Received-SPF: pass (google.com: domain of linux-kernel+bounces-103678-linux.lists.archive=gmail.com@vger.kernel.org designates 147.75.80.249 as permitted sender) client-ip=147.75.80.249;
Date: Thu, 14 Mar 2024 11:38:08 -0700
In-Reply-To: <CABgObfa3By9GU9_8FmqHQK-AxWU3ocbBkQK0xXwx2XRDP828dg@mail.gmail.com>
Precedence: bulk
Mime-Version: 1.0
References: <20240308223702.1350851-1-seanjc@google.com> <20240308223702.1350851-5-seanjc@google.com>
 <CABgObfa3By9GU9_8FmqHQK-AxWU3ocbBkQK0xXwx2XRDP828dg@mail.gmail.com>
Message-ID: <ZfNEEFmTkx-RVuix@google.com>
Subject: Re: [GIT PULL] KVM: x86: MMU changes for 6.9
From: Sean Christopherson <seanjc@google.com>
To: Paolo Bonzini <pbonzini@redhat.com>
Cc: kvm@vger.kernel.org, linux-kernel@vger.kernel.org
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: quoted-printable

On Thu, Mar 14, 2024, Paolo Bonzini wrote:
> On Fri, Mar 8, 2024 at 11:37=E2=80=AFPM Sean Christopherson <seanjc@googl=
e.com> wrote:
> >
> >  - Zap TDP MMU roots at 4KiB granularity to minimize the delay in yield=
ing if
> >    a reschedule is needed, e.g. if a high priority task needs to run.  =
Because
> >    KVM doesn't support yielding in the middle of processing a zapped no=
n-leaf
> >    SPTE, zapping at 1GiB granularity can result in multi-millisecond la=
g when
> >    attempting to schedule in a high priority.
> >
>=20
> Would 2 MiB provide a nice middle ground?

Not really?

Zapping at 2MiB definitely fixes the worst of the tail latencies, but there=
 is
still a measurable difference between 2MiB and 4KiB.  And on the other side=
 of the
coing, I was unable to observe a meaningful difference in total runtime by =
zapping
at 2MiB, or even 1GiB, versus 4KiB.

In other words, AFAICT, there's no need to shoot for a middle ground becaus=
e trying
to zap at larger granularities doesn't buy us anything.