Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Subject: Re: [PATCH v2 2/3] KVM: VMX: Do not change PID.NDST when loading a
 blocked vCPU
To:     Wanpeng Li <kernellwp@gmail.com>
Cc:     kvm <kvm@vger.kernel.org>, LKML <linux-kernel@vger.kernel.org>,
        Paolo Bonzini <pbonzini@redhat.com>,
        =?UTF-8?B?UmFkaW0gS3LEjW3DocWZ?= <rkrcmar@redhat.com>,
        Sean Christopherson <sean.j.christopherson@intel.com>,
        Vitaly Kuznetsov <vkuznets@redhat.com>,
        Wanpeng Li <wanpengli@tencent.com>,
        Jim Mattson <jmattson@google.com>,
        Joerg Roedel <joro@8bytes.org>,
        Liran Alon <liran.alon@oracle.com>,
        Jag Raman <jag.raman@oracle.com>
References: <20191111172012.28356-1-joao.m.martins@oracle.com>
 <20191111172012.28356-3-joao.m.martins@oracle.com>
 <CANRm+CyrbiJ068zLRH8ZMttqjnEG38qb1W1SMND+H-D=8N8tVw@mail.gmail.com>
From:   Joao Martins <joao.m.martins@oracle.com>
Message-ID: <6b6acdee-f080-241f-8a6e-89708f746944@oracle.com>
Date:   Tue, 19 Nov 2019 14:44:27 +0000
MIME-Version: 1.0
In-Reply-To: <CANRm+CyrbiJ068zLRH8ZMttqjnEG38qb1W1SMND+H-D=8N8tVw@mail.gmail.com>
Content-Type: text/plain; charset=utf-8
Content-Language: en-US
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On 11/19/19 11:36 AM, Wanpeng Li wrote:
> On Tue, 12 Nov 2019 at 01:23, Joao Martins <joao.m.martins@oracle.com> wrote:
>>
>> While vCPU is blocked (in kvm_vcpu_block()), it may be preempted which
>> will cause vmx_vcpu_pi_put() to set PID.SN.  If later the vCPU will be
> 
> How can this happen? See the prepare_to_swait_exlusive() in
> kvm_vcpu_block(), the task will be set in TASK_INTERRUPTIBLE state,
> kvm_sched_out will set vcpu->preempted to true iff current->state ==
> TASK_RUNNING.

You're right.

But preemption (thus setting PID.SN) can still happen before vcpu_block(), or
otherwise scheduled out if we are already in vcpu_block() (with task in
TASK_INTERRUPTIBLE). The right term we should have written in that sentence
above would have been 'scheduled out' and drop the vcpu_block mention and it
would encompass both cases. A better sentence would perhaps be:

"While vCPU is blocked (or about to block) it may be scheduled out which will
cause vmx_vcpu_pi_put() to be called."

But setting or not preempted/PID.SN doesn't change the rest and was mentioned
for completeness.

	Joao