Date: Thu, 21 Apr 2016 16:30:07 +0200
From: Greg Kurz <gkurz@linux.vnet.ibm.com>
To: David Hildenbrand <dahi@linux.vnet.ibm.com>
Cc: Paolo Bonzini <pbonzini@redhat.com>, james.hogan@imgtec.com,
        mingo@redhat.com, linux-mips@linux-mips.org, kvm@vger.kernel.org,
        rkrcmar@redhat.com, linux-kernel@vger.kernel.org, qemu-ppc@nongnu.org,
        Cornelia Huck <cornelia.huck@de.ibm.com>,
        Paul Mackerras <paulus@samba.org>,
        David Gibson <david@gibson.dropbear.id.au>
Subject: Re: [PATCH v4 1/2] KVM: remove NULL return path for vcpu ids >=
 KVM_MAX_VCPUS
Message-ID: <20160421163007.6302a1bc@bahia.huguette.org>
In-Reply-To: <20160421161729.2f430008@thinkpad-w530>
References: <146124809455.32509.15232948272580716135.stgit@bahia.huguette.org>
	<146124810201.32509.2946887043729554992.stgit@bahia.huguette.org>
	<20160421161729.2f430008@thinkpad-w530>
Organization: IBM
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1802
Lines: 59

On Thu, 21 Apr 2016 16:17:29 +0200
David Hildenbrand <dahi@linux.vnet.ibm.com> wrote:

> > Commit c896939f7cff ("KVM: use heuristic for fast VCPU lookup by id") added
> > a return path that prevents vcpu ids to exceed KVM_MAX_VCPUS. This is a
> > problem for powerpc where vcpu ids can grow up to 8*KVM_MAX_VCPUS.
> > 
> > This patch simply reverses the logic so that we only try fast path if the
> > vcpu id can be tried as an index in kvm->vcpus[]. The slow path is not
> > affected by the change.
> > 
> > Signed-off-by: Greg Kurz <gkurz@linux.vnet.ibm.com>
> > ---
> >  include/linux/kvm_host.h |    7 ++++---
> >  1 file changed, 4 insertions(+), 3 deletions(-)
> > 
> > diff --git a/include/linux/kvm_host.h b/include/linux/kvm_host.h
> > index 5276fe0916fc..23bfe1bd159c 100644
> > --- a/include/linux/kvm_host.h
> > +++ b/include/linux/kvm_host.h
> > @@ -447,12 +447,13 @@ static inline struct kvm_vcpu *kvm_get_vcpu(struct kvm *kvm, int i)
> > 
> >  static inline struct kvm_vcpu *kvm_get_vcpu_by_id(struct kvm *kvm, int id)
> >  {
> > -	struct kvm_vcpu *vcpu;
> > +	struct kvm_vcpu *vcpu = NULL;
> >  	int i;
> > 
> > -	if (id < 0 || id >= KVM_MAX_VCPUS)
> > +	if (id < 0)
> >  		return NULL;
> > -	vcpu = kvm_get_vcpu(kvm, id);
> > +	if (id < KVM_MAX_VCPUS)
> > +		vcpu = kvm_get_vcpu(kvm, id);  
> 
> Maybe this check even should go into kvm_get_vcpu()
> 

Yeah possibly, but there are 19 users for kvm_get_vcpu() and I'm not sure if
none of them is on a hot path where this extra check could hurt... maybe this
can be done in a cleanup patch afterwards ?

> >  	if (vcpu && vcpu->vcpu_id == id)
> >  		return vcpu;
> >  	kvm_for_each_vcpu(i, vcpu, kvm)
> >   
> 
> Anyhow,
> 
> Reviewed-by: David Hildenbrand <dahi@linux.vnet.ibm.com>
> 
> David

Thanks for the review !

Cheers.

--
Greg