From: Paolo Bonzini <pbonzini@redhat.com>
To: "Maciej S. Szmigiero" <mail@maciej.szmigiero.name>,
Sean Christopherson <seanjc@google.com>
Cc: Vitaly Kuznetsov <vkuznets@redhat.com>,
Wanpeng Li <wanpengli@tencent.com>,
Jim Mattson <jmattson@google.com>,
Igor Mammedov <imammedo@redhat.com>,
Marc Zyngier <maz@kernel.org>, James Morse <james.morse@arm.com>,
Julien Thierry <julien.thierry.kdev@gmail.com>,
Suzuki K Poulose <suzuki.poulose@arm.com>,
Huacai Chen <chenhuacai@kernel.org>,
Aleksandar Markovic <aleksandar.qemu.devel@gmail.com>,
Paul Mackerras <paulus@ozlabs.org>,
Christian Borntraeger <borntraeger@de.ibm.com>,
Janosch Frank <frankja@linux.ibm.com>,
David Hildenbrand <david@redhat.com>,
Cornelia Huck <cohuck@redhat.com>,
Claudio Imbrenda <imbrenda@linux.ibm.com>,
Joerg Roedel <joro@8bytes.org>,
kvm@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH 2/2] KVM: Scalable memslots implementation
Date: Wed, 3 Feb 2021 14:41:02 +0100 [thread overview]
Message-ID: <4bdcb44c-c35d-45b2-c0c1-e857e0fd383e@redhat.com> (raw)
In-Reply-To: <9e6ca093-35c3-7cca-443b-9f635df4891d@maciej.szmigiero.name>
On 02/02/21 23:42, Maciej S. Szmigiero wrote:
>> I'm not opposed to using more sophisticated storage for the gfn
>> lookups, but only if there's a good reason for doing so. IMO, the
>> rbtree isn't simpler, just different.
And it also has worse cache utilization than an array, due to memory
footprint (as you point out below) but also pointer chasing.
>> Memslot modifications are
>> unlikely to be a hot path (and if it is, x86's "zap everything"
>> implementation is a far bigger problem), and it's hard to beat the
>> memory footprint of a raw array. That doesn't leave much
>> motivation for such a big change to some of KVM's scariest (for me)
>> code.
>>
>
> Improvements can be done step-by-step,
> kvm_mmu_invalidate_zap_pages_in_memslot() can be rewritten, too in
> the future, if necessary. After all, complains are that this change
> alone is too big.
>
> I think that if you look not at the patch itself but at the
> resulting code the new implementation looks rather straightforward,
> there are comments at every step in kvm_set_memslot() to explain
> exactly what is being done. Not only it already scales well, but it
> is also flexible to accommodate further improvements or even new
> operations.
>
> The new implementation also uses standard kernel {rb,interval}-tree
> and hash table implementation as its basic data structures, so it
> automatically benefits from any generic improvements to these.
>
> All for the low price of just 174 net lines of code added.
I think the best thing to do here is to provide a patch series that
splits the individual changes so that they can be reviewed and their
separate merits evaluated.
Another thing that I dislike about KVM_SET_USER_MEMORY_REGION is that
IMO userspace should provide all memslots at once, for an atomic switch
of the whole memory array. (Or at least I would like to see the code;
it might be a bit tricky because you'll need code to compute the
difference between the old and new arrays and invoke
kvm_arch_prepare/commit_memory_region). I'm not sure how that would
interact with the active/inactive pair that you introduce here.
Paolo
next prev parent reply other threads:[~2021-02-03 13:43 UTC|newest]
Thread overview: 19+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-02-01 8:13 [PATCH 1/2] KVM: x86/mmu: Make HVA handler retpoline-friendly Maciej S. Szmigiero
2021-02-01 8:13 ` [PATCH 2/2] KVM: Scalable memslots implementation Maciej S. Szmigiero
2021-02-02 1:33 ` Sean Christopherson
2021-02-02 22:42 ` Maciej S. Szmigiero
2021-02-02 23:43 ` Sean Christopherson
2021-02-03 10:44 ` Maciej S. Szmigiero
2021-02-03 14:21 ` Vitaly Kuznetsov
2021-02-03 20:10 ` Maciej S. Szmigiero
2021-02-03 13:43 ` Paolo Bonzini
2021-02-03 13:41 ` Paolo Bonzini [this message]
2021-02-03 13:44 ` David Hildenbrand
2021-02-03 13:46 ` Paolo Bonzini
2021-02-03 13:52 ` David Hildenbrand
2021-02-03 14:33 ` Paolo Bonzini
2021-02-03 20:09 ` Maciej S. Szmigiero
2021-02-01 8:21 ` [PATCH 1/2] KVM: x86/mmu: Make HVA handler retpoline-friendly Paolo Bonzini
2021-02-01 15:19 ` Maciej S. Szmigiero
2021-02-03 14:43 ` Paolo Bonzini
2021-02-03 20:09 ` Maciej S. Szmigiero
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=4bdcb44c-c35d-45b2-c0c1-e857e0fd383e@redhat.com \
--to=pbonzini@redhat.com \
--cc=aleksandar.qemu.devel@gmail.com \
--cc=borntraeger@de.ibm.com \
--cc=chenhuacai@kernel.org \
--cc=cohuck@redhat.com \
--cc=david@redhat.com \
--cc=frankja@linux.ibm.com \
--cc=imammedo@redhat.com \
--cc=imbrenda@linux.ibm.com \
--cc=james.morse@arm.com \
--cc=jmattson@google.com \
--cc=joro@8bytes.org \
--cc=julien.thierry.kdev@gmail.com \
--cc=kvm@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mail@maciej.szmigiero.name \
--cc=maz@kernel.org \
--cc=paulus@ozlabs.org \
--cc=seanjc@google.com \
--cc=suzuki.poulose@arm.com \
--cc=vkuznets@redhat.com \
--cc=wanpengli@tencent.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).