public inbox for linuxppc-dev@ozlabs.org 
 help / color / mirror / Atom feed
From: Sean Christopherson <seanjc@google•com>
To: Marc Zyngier <maz@kernel•org>
Cc: Juergen Gross <jgross@suse•com>,
	Alexandru Elisei <alexandru.elisei@arm•com>,
	Anup Patel <anup.patel@wdc•com>,
	Janosch Frank <frankja@linux•ibm.com>,
	kvm@vger•kernel.org,
	Christian Borntraeger <borntraeger@de•ibm.com>,
	Huacai Chen <chenhuacai@kernel•org>,
	David Hildenbrand <david@redhat•com>,
	linux-mips@vger•kernel.org, Nicholas Piggin <npiggin@gmail•com>,
	Atish Patra <atish.patra@wdc•com>,
	Aleksandar Markovic <aleksandar.qemu.devel@gmail•com>,
	Paul Mackerras <paulus@samba•org>,
	James Morse <james.morse@arm•com>,
	Paolo Bonzini <pbonzini@redhat•com>,
	kernel-team@android•com,
	Claudio Imbrenda <imbrenda@linux•ibm.com>,
	linuxppc-dev@lists•ozlabs.org, kvmarm@lists•cs.columbia.edu,
	Suzuki K Poulose <suzuki.poulose@arm•com>
Subject: Re: [PATCH 5/5] KVM: Convert the kvm->vcpus array to a xarray
Date: Fri, 5 Nov 2021 20:21:36 +0000	[thread overview]
Message-ID: <YYWSUJ1qzhfqjQow@google.com> (raw)
In-Reply-To: <20211105192101.3862492-6-maz@kernel.org>

On Fri, Nov 05, 2021, Marc Zyngier wrote:
> At least on arm64 and x86, the vcpus array is pretty huge (512 entries),
> and is mostly empty in most cases (running 512 vcpu VMs is not that
> common). This mean that we end-up with a 4kB block of unused memory
> in the middle of the kvm structure.

Heh, x86 is now up to 1024 entries.
 
> Instead of wasting away this memory, let's use an xarray instead,
> which gives us almost the same flexibility as a normal array, but
> with a reduced memory usage with smaller VMs.
> 
> Signed-off-by: Marc Zyngier <maz@kernel•org>
> ---
> @@ -693,7 +694,7 @@ static inline struct kvm_vcpu *kvm_get_vcpu(struct kvm *kvm, int i)
>  
>  	/* Pairs with smp_wmb() in kvm_vm_ioctl_create_vcpu.  */
>  	smp_rmb();
> -	return kvm->vcpus[i];
> +	return xa_load(&kvm->vcpu_array, i);
>  }

It'd be nice for this series to convert kvm_for_each_vcpu() to use xa_for_each()
as well.  Maybe as a patch on top so that potential explosions from that are
isolated from the initiali conversion?

Or maybe even use xa_for_each_range() to cap at online_vcpus?  That's technically
a functional change, but IMO it's easier to reason about iterating over a snapshot
of vCPUs as opposed to being able to iterate over vCPUs as their being added.  In
practice I doubt it matters.

#define kvm_for_each_vcpu(idx, vcpup, kvm) \
	xa_for_each_range(&kvm->vcpu_array, idx, vcpup, 0, atomic_read(&kvm->online_vcpus))

  reply	other threads:[~2021-11-05 20:22 UTC|newest]

Thread overview: 24+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-11-05 19:20 [PATCH 0/5] KVM: Turn the vcpu array into an xarray Marc Zyngier
2021-11-05 19:20 ` [PATCH 1/5] KVM: Move wiping of the kvm->vcpus array to common code Marc Zyngier
2021-11-05 20:12   ` Sean Christopherson
2021-11-06 11:17     ` Marc Zyngier
2021-11-16 13:49       ` Paolo Bonzini
2021-11-08 12:12   ` Claudio Imbrenda
2021-11-05 19:20 ` [PATCH 2/5] KVM: mips: Use kvm_get_vcpu() instead of open-coded access Marc Zyngier
2021-11-06 15:56   ` Philippe Mathieu-Daudé
2021-11-05 19:20 ` [PATCH 3/5] KVM: s390: " Marc Zyngier
2021-11-08 12:13   ` Claudio Imbrenda
2021-11-05 19:21 ` [PATCH 4/5] KVM: x86: " Marc Zyngier
2021-11-05 20:03   ` Sean Christopherson
2021-11-16 14:04     ` Paolo Bonzini
2021-11-16 16:07       ` Sean Christopherson
2021-11-16 16:48         ` Paolo Bonzini
2021-11-05 19:21 ` [PATCH 5/5] KVM: Convert the kvm->vcpus array to a xarray Marc Zyngier
2021-11-05 20:21   ` Sean Christopherson [this message]
2021-11-06 11:48     ` Marc Zyngier
2021-11-08  8:23       ` Marc Zyngier
2021-11-16 14:13 ` [PATCH 0/5] KVM: Turn the vcpu array into an xarray Juergen Gross
2021-11-16 14:21   ` Paolo Bonzini
2021-11-16 14:54     ` Juergen Gross
2021-11-16 15:03 ` Paolo Bonzini
2021-11-16 15:40   ` Marc Zyngier

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=YYWSUJ1qzhfqjQow@google.com \
    --to=seanjc@google$(echo .)com \
    --cc=aleksandar.qemu.devel@gmail$(echo .)com \
    --cc=alexandru.elisei@arm$(echo .)com \
    --cc=anup.patel@wdc$(echo .)com \
    --cc=atish.patra@wdc$(echo .)com \
    --cc=borntraeger@de$(echo .)ibm.com \
    --cc=chenhuacai@kernel$(echo .)org \
    --cc=david@redhat$(echo .)com \
    --cc=frankja@linux$(echo .)ibm.com \
    --cc=imbrenda@linux$(echo .)ibm.com \
    --cc=james.morse@arm$(echo .)com \
    --cc=jgross@suse$(echo .)com \
    --cc=kernel-team@android$(echo .)com \
    --cc=kvm@vger$(echo .)kernel.org \
    --cc=kvmarm@lists$(echo .)cs.columbia.edu \
    --cc=linux-mips@vger$(echo .)kernel.org \
    --cc=linuxppc-dev@lists$(echo .)ozlabs.org \
    --cc=maz@kernel$(echo .)org \
    --cc=npiggin@gmail$(echo .)com \
    --cc=paulus@samba$(echo .)org \
    --cc=pbonzini@redhat$(echo .)com \
    --cc=suzuki.poulose@arm$(echo .)com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html
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