Received-SPF: pass (google.com: domain of linux-kernel+bounces-89134-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:40f1:3f00::1 as permitted sender) client-ip=2604:1380:40f1:3f00::1;
Date: Fri, 1 Mar 2024 21:13:26 +0000
From: Mark Brown <broonie@kernel.org>
To: Oliver Upton <oliver.upton@linux.dev>
Cc: Marc Zyngier <maz@kernel.org>, James Morse <james.morse@arm.com>,
	Suzuki K Poulose <suzuki.poulose@arm.com>,
	Catalin Marinas <catalin.marinas@arm.com>,
	Will Deacon <will@kernel.org>, Joey Gouly <joey.gouly@arm.com>,
	linux-arm-kernel@lists.infradead.org, kvmarm@lists.linux.dev,
	linux-kernel@vger.kernel.org
Subject: Re: [PATCH] KVM: arm64: Only save S1PIE registers when dirty
Message-ID: <562f5e62-c26c-41d9-9ab9-aac02c91c7ae@sirena.org.uk>
References: <20240301-kvm-arm64-defer-regs-v1-1-401e3de92e97@kernel.org>
 <ZeItTLQxdxxICw01@linux.dev>
Precedence: bulk
MIME-Version: 1.0
Content-Type: multipart/signed; micalg=pgp-sha512;
	protocol="application/pgp-signature"; boundary="tCYQ/jJS3BIUwGub"
Content-Disposition: inline
In-Reply-To: <ZeItTLQxdxxICw01@linux.dev>


--tCYQ/jJS3BIUwGub
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline

On Fri, Mar 01, 2024 at 07:32:28PM +0000, Oliver Upton wrote:
> On Fri, Mar 01, 2024 at 06:05:53PM +0000, Mark Brown wrote:

> > I don't have a good sense if this is a good idea or not, or if this is a
> > desirable implementation of the concept - the patch is based on some
> > concerns about the cost of the system register context switching.  We
> > should be able to do something similar for some of the other registers.

> Is there any data beyond a microbenchmark to suggest save elision
> benefits the VM at all? The idea of baking the trap configuration based
> on what KVM _thinks_ the guest will do isn't particularly exciting. This
> doesn't seem to be a one-size-fits-all solution.

No, and as I said above I'm really not confident about this.  There's no
hardware with these registers yet as far as I know so I don't know how
meaningful any benchmark would be anyway, and as you suggest even with a
benchmark a new guest could always come along and blow performance up
with a change in access patterns.

> The overheads of guest exits are extremely configuration dependent, and
> on VHE the save/restore of EL1 state happens at vcpu_load() / vcpu_put()
> rather than every exit. There isn't a whole lot KVM can do to lessen the
> blow of sharing EL1 in the nVHE configuration.

> Looking a bit further out, the cost of traps will be dramatically higher
> when running as a guest hypervisor, so we'd want to avoid them if
> possible...

Indeed, but OTOH I got some complaints about adding more system register
switching in __sysreg_save_el1_state() for one of my other serieses that
specifically mentioned nested virt and there don't seem to be a huge
range of other options for reducing what we're doing with context
switching without using traps to figure out what's in use, especially in
the nVHE case.  I figured I'd send the patch so the idea could be
considered.

--tCYQ/jJS3BIUwGub
Content-Type: application/pgp-signature; name="signature.asc"

-----BEGIN PGP SIGNATURE-----

iQEzBAABCgAdFiEEreZoqmdXGLWf4p/qJNaLcl1Uh9AFAmXiRPYACgkQJNaLcl1U
h9Dfuwf/bgBH01WZBV1RX0Vk8R30mTChMdhuVWMiQhutnkILMJwYVIpKOzni84N+
wdMI7nOMHwqFa1XrlzB8Cgmukf//RTNSXEV0RAKT+a8xwzZMhd8LpU6iTUqDX2Au
Mmlrc7A5zHPEsWmSyit1+417x5g9Ki/dAC8yk7IpzgC/PJVc/0uqjHqmd3c2vzuc
vQQww4Iv9hVYlM/eAOmt8cl5+TOenP+VDEiHJ4HM8st5UD08RFu1w04VBsvmOw5S
WxZXwRknDVscRutJixjhF9+jtWO9LbNWriH92Qor73z4fixi6II7ID2afKaWrCnl
gSSUfnaTNmUoOzlL6Ax/ytrFsotx/w==
=MPnZ
-----END PGP SIGNATURE-----

--tCYQ/jJS3BIUwGub--