Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
MIME-Version: 1.0
References: <20210819154910.1064090-1-pgonda@google.com> <20210819154910.1064090-2-pgonda@google.com>
 <CAA03e5Gh0kJYHP1R3F7uh6x83LBFPp=af2xt7q3epgg+8XW53g@mail.gmail.com>
 <CAMkAt6oJcW3MHP3fod9RnRHCEYp-whdEtBTyfuqgFgATKa=3Hg@mail.gmail.com>
 <YR7iD6kdTUpWwwRn@google.com> <CAA03e5F62WEcs3PN6M9qGzW+wuufp+BjwDHcTt18yaB42RDYkA@mail.gmail.com>
In-Reply-To: <CAA03e5F62WEcs3PN6M9qGzW+wuufp+BjwDHcTt18yaB42RDYkA@mail.gmail.com>
From:   Peter Gonda <pgonda@google.com>
Date:   Mon, 23 Aug 2021 10:39:40 -0600
Message-ID: <CAMkAt6rJTTojQ7Ymbywmdns-8AWNBVEde4D6cZkceL=czMaEig@mail.gmail.com>
Subject: Re: [PATCH 1/2 V4] KVM, SEV: Add support for SEV intra host migration
To:     Marc Orr <marcorr@google.com>
Cc:     Sean Christopherson <seanjc@google.com>,
        kvm list <kvm@vger.kernel.org>,
        Paolo Bonzini <pbonzini@redhat.com>,
        David Rientjes <rientjes@google.com>,
        "Dr . David Alan Gilbert" <dgilbert@redhat.com>,
        Brijesh Singh <brijesh.singh@amd.com>,
        Vitaly Kuznetsov <vkuznets@redhat.com>,
        Wanpeng Li <wanpengli@tencent.com>,
        Jim Mattson <jmattson@google.com>,
        Joerg Roedel <joro@8bytes.org>,
        Thomas Gleixner <tglx@linutronix.de>,
        Ingo Molnar <mingo@redhat.com>, Borislav Petkov <bp@alien8.de>,
        "H. Peter Anvin" <hpa@zytor.com>, linux-kernel@vger.kernel.org
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

On Fri, Aug 20, 2021 at 2:53 PM Marc Orr <marcorr@google.com> wrote:
>
> On Thu, Aug 19, 2021 at 3:58 PM Sean Christopherson <seanjc@google.com> wrote:
> >
> > On Thu, Aug 19, 2021, Peter Gonda wrote:
> > > > >
> > > > > +static int svm_sev_lock_for_migration(struct kvm *kvm)
> > > > > +{
> > > > > +       struct kvm_sev_info *sev = &to_kvm_svm(kvm)->sev_info;
> > > > > +       int ret;
> > > > > +
> > > > > +       /*
> > > > > +        * Bail if this VM is already involved in a migration to avoid deadlock
> > > > > +        * between two VMs trying to migrate to/from each other.
> > > > > +        */
> > > > > +       spin_lock(&sev->migration_lock);
> > > > > +       if (sev->migration_in_progress)
> > > > > +               ret = -EBUSY;
> > > > > +       else {
> > > > > +               /*
> > > > > +                * Otherwise indicate VM is migrating and take the KVM lock.
> > > > > +                */
> > > > > +               sev->migration_in_progress = true;
> > > > > +               mutex_lock(&kvm->lock);
> >
> > Deadlock aside, mutex_lock() can sleep, which is not allowed while holding a
> > spinlock, i.e. this patch does not work.  That's my suggestion did the crazy
> > dance of "acquiring" a flag.

Ah, makes sense.

> >
> > What I don't know is why on earth I suggested a global spinlock, a simple atomic
> > should work, e.g.
> >
> >                 if (atomic_cmpxchg_acquire(&sev->migration_in_progress, 0, 1))
> >                         return -EBUSY;
> >
> >                 mutex_lock(&kvm->lock);
> >
> > and on the backend...
> >
> >                 mutex_unlock(&kvm->lock);
> >
> >                 atomic_set_release(&sev->migration_in_progress, 0);
>
> +1 to replacing the spin lock with an atomic flag. Correctness issues
> aside, I think it's also cleaner. Also, I'd suggest adding a comment
> to source code to explain that the `migration_in_progress` flag is to
> prevent deadlock due to the "double migration" discussed previously.

Thanks! I've updated these locks to use the atomic. It looks much cleaner now.