linux-kernel - Re: [PATCH v3] KVM: X86: Just one leader to trigger kvmclock sync request

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <87lfoihpwt.fsf@vitty.brq.redhat.com>
Date:   Mon, 02 Mar 2020 14:01:06 +0100
From:   Vitaly Kuznetsov <vkuznets@...hat.com>
To:     Wanpeng Li <kernellwp@...il.com>
Cc:     Paolo Bonzini <pbonzini@...hat.com>,
        Sean Christopherson <sean.j.christopherson@...el.com>,
        Wanpeng Li <wanpengli@...cent.com>,
        Jim Mattson <jmattson@...gle.com>,
        Joerg Roedel <joro@...tes.org>, linux-kernel@...r.kernel.org,
        kvm@...r.kernel.org
Subject: Re: [PATCH v3] KVM: X86: Just one leader to trigger kvmclock sync request

Wanpeng Li <kernellwp@...il.com> writes:

> From: Wanpeng Li <wanpengli@...cent.com>
>
> In the progress of vCPUs creation, it queues a kvmclock sync worker to the global
> workqueue before each vCPU creation completes. The workqueue subsystem guarantees 
> not to queue the already queued work, however, we can make the logic more clear by 
> make just one leader to trigger this kvmclock sync request and save on cacheline 
> boucing due to test_and_set_bit.
>
> Signed-off-by: Wanpeng Li <wanpengli@...cent.com>
> ---
> v2 -> v3:
>  * update patch description
> v1 -> v2:
>  * check vcpu->vcpu_idx
>
>  arch/x86/kvm/x86.c | 5 +++--
>  1 file changed, 3 insertions(+), 2 deletions(-)
>
> diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
> index fb5d64e..79bc995 100644
> --- a/arch/x86/kvm/x86.c
> +++ b/arch/x86/kvm/x86.c
> @@ -9390,8 +9390,9 @@ void kvm_arch_vcpu_postcreate(struct kvm_vcpu *vcpu)
>  	if (!kvmclock_periodic_sync)
>  		return;
>  
> -	schedule_delayed_work(&kvm->arch.kvmclock_sync_work,
> -					KVMCLOCK_SYNC_PERIOD);
> +	if (vcpu->vcpu_idx == 0)
> +		schedule_delayed_work(&kvm->arch.kvmclock_sync_work,
> +						KVMCLOCK_SYNC_PERIOD);

I would've merged this new check with !kvmclock_periodic_sync above
making it more obvious when the work is scheduled

diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 5de200663f51..93550976f991 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -9389,11 +9389,9 @@ void kvm_arch_vcpu_postcreate(struct kvm_vcpu *vcpu)
 
        mutex_unlock(&vcpu->mutex);
 
-       if (!kvmclock_periodic_sync)
-               return;
-
-       schedule_delayed_work(&kvm->arch.kvmclock_sync_work,
-                                       KVMCLOCK_SYNC_PERIOD);
+       if (vcpu->vcpu_idx == 0 && kvmclock_periodic_sync)
+               schedule_delayed_work(&kvm->arch.kvmclock_sync_work,
+                                     KVMCLOCK_SYNC_PERIOD);
 }

>  }
>  
>  void kvm_arch_vcpu_destroy(struct kvm_vcpu *vcpu)

With or without the change mentioned above,
Reviewed-by: Vitaly Kuznetsov <vkuznets@...hat.com>

-- 
Vitaly