From mboxrd@z Thu Jan  1 00:00:00 1970
From: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Subject: Re: [PATCH v2 1/2] KVM: MMU: Fix infinite loop when there is no
 available mmu page
Date: Wed, 6 Dec 2017 06:12:37 -0500
Message-ID: <20171206111237.GB28047@char.us.oracle.com>
References: <1512552510-13858-1-git-send-email-wanpeng.li@hotmail.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
Cc: linux-kernel@vger.kernel.org, kvm@vger.kernel.org,
        Paolo Bonzini <pbonzini@redhat.com>,
        Radim =?utf-8?B?S3LEjW3DocWZ?= <rkrcmar@redhat.com>,
        Wanpeng Li <wanpeng.li@hotmail.com>
To: Wanpeng Li <kernellwp@gmail.com>
Return-path: <linux-kernel-owner@vger.kernel.org>
Content-Disposition: inline
In-Reply-To: <1512552510-13858-1-git-send-email-wanpeng.li@hotmail.com>
Sender: linux-kernel-owner@vger.kernel.org
List-Id: kvm.vger.kernel.org

On Wed, Dec 06, 2017 at 01:28:30AM -0800, Wanpeng Li wrote:
> From: Wanpeng Li <wanpeng.li@hotmail.com>
>=20
> The below test case can cause infinite loop in kvm when ept=3D0.
>=20
>     #include <unistd.h>
>     #include <sys/syscall.h>
>     #include <string.h>
>     #include <stdint.h>
>     #include <linux/kvm.h>
>     #include <fcntl.h>
>     #include <sys/ioctl.h>
>=20
>     long r[5];
>     int main()
>     {
>     	r[2] =3D open("/dev/kvm", O_RDONLY);
>     	r[3] =3D ioctl(r[2], KVM_CREATE_VM, 0);
>     	r[4] =3D ioctl(r[3], KVM_CREATE_VCPU, 7);
>     	ioctl(r[4], KVM_RUN, 0);
>     }
>=20
> It doesn't setup the memory regions, mmu_alloc_shadow/direct_roots() in
> kvm return 1 when kvm fails to allocate root page table which can resul=
t
> in beblow infinite loop:
>=20
> 	vcpu_run() {
> 		for (;;) {
> 			r =3D vcpu_enter_guest()::kvm_mmu_reload() returns 1
> 			if (r <=3D 0)
> 				break;
> 			if (need_resched())
> 				cond_resched();
> 		}
> 	}
>=20
> This patch fixes it by returning -ENOSPC when there is no available kvm=
 mmu

No space left?

Why not ENXIO ?

> page for root page table.
>=20
> Cc: Paolo Bonzini <pbonzini@redhat.com>
> Cc: Radim Kr=C4=8Dm=C3=A1=C5=99 <rkrcmar@redhat.com>
> Fixes: 26eeb53cf0f (KVM: MMU: Bail out immediately if there is no avail=
able mmu page)
> Signed-off-by: Wanpeng Li <wanpeng.li@hotmail.com>
> ---
> v1 -> v2:
>  * cleanup patch description
>=20
>  arch/x86/kvm/mmu.c | 8 ++++----
>  1 file changed, 4 insertions(+), 4 deletions(-)
>=20
> diff --git a/arch/x86/kvm/mmu.c b/arch/x86/kvm/mmu.c
> index c9aaa18..89da688 100644
> --- a/arch/x86/kvm/mmu.c
> +++ b/arch/x86/kvm/mmu.c
> @@ -3395,7 +3395,7 @@ static int mmu_alloc_direct_roots(struct kvm_vcpu=
 *vcpu)
>  		spin_lock(&vcpu->kvm->mmu_lock);
>  		if(make_mmu_pages_available(vcpu) < 0) {
>  			spin_unlock(&vcpu->kvm->mmu_lock);
> -			return 1;
> +			return -ENOSPC;
>  		}
>  		sp =3D kvm_mmu_get_page(vcpu, 0, 0,
>  				vcpu->arch.mmu.shadow_root_level, 1, ACC_ALL);
> @@ -3410,7 +3410,7 @@ static int mmu_alloc_direct_roots(struct kvm_vcpu=
 *vcpu)
>  			spin_lock(&vcpu->kvm->mmu_lock);
>  			if (make_mmu_pages_available(vcpu) < 0) {
>  				spin_unlock(&vcpu->kvm->mmu_lock);
> -				return 1;
> +				return -ENOSPC;
>  			}
>  			sp =3D kvm_mmu_get_page(vcpu, i << (30 - PAGE_SHIFT),
>  					i << 30, PT32_ROOT_LEVEL, 1, ACC_ALL);
> @@ -3450,7 +3450,7 @@ static int mmu_alloc_shadow_roots(struct kvm_vcpu=
 *vcpu)
>  		spin_lock(&vcpu->kvm->mmu_lock);
>  		if (make_mmu_pages_available(vcpu) < 0) {
>  			spin_unlock(&vcpu->kvm->mmu_lock);
> -			return 1;
> +			return -ENOSPC;
>  		}
>  		sp =3D kvm_mmu_get_page(vcpu, root_gfn, 0,
>  				vcpu->arch.mmu.shadow_root_level, 0, ACC_ALL);
> @@ -3487,7 +3487,7 @@ static int mmu_alloc_shadow_roots(struct kvm_vcpu=
 *vcpu)
>  		spin_lock(&vcpu->kvm->mmu_lock);
>  		if (make_mmu_pages_available(vcpu) < 0) {
>  			spin_unlock(&vcpu->kvm->mmu_lock);
> -			return 1;
> +			return -ENOSPC;
>  		}
>  		sp =3D kvm_mmu_get_page(vcpu, root_gfn, i << 30, PT32_ROOT_LEVEL,
>  				      0, ACC_ALL);
> --=20
> 2.7.4
>=20