From:   "Liu, Yi L" <yi.l.liu@intel.com>
To:     Matthew Rosato <mjrosato@linux.ibm.com>,
        "alex.williamson@redhat.com" <alex.williamson@redhat.com>,
        "pbonzini@redhat.com" <pbonzini@redhat.com>,
        "jgg@nvidia.com" <jgg@nvidia.com>
CC:     "cohuck@redhat.com" <cohuck@redhat.com>,
        "farman@linux.ibm.com" <farman@linux.ibm.com>,
        "pmorel@linux.ibm.com" <pmorel@linux.ibm.com>,
        "borntraeger@linux.ibm.com" <borntraeger@linux.ibm.com>,
        "frankja@linux.ibm.com" <frankja@linux.ibm.com>,
        "imbrenda@linux.ibm.com" <imbrenda@linux.ibm.com>,
        "david@redhat.com" <david@redhat.com>,
        "akrowiak@linux.ibm.com" <akrowiak@linux.ibm.com>,
        "jjherne@linux.ibm.com" <jjherne@linux.ibm.com>,
        "pasic@linux.ibm.com" <pasic@linux.ibm.com>,
        "zhenyuw@linux.intel.com" <zhenyuw@linux.intel.com>,
        "Wang, Zhi A" <zhi.a.wang@intel.com>,
        "Christopherson,, Sean" <seanjc@google.com>,
        "Tian, Kevin" <kevin.tian@intel.com>,
        "linux-s390@vger.kernel.org" <linux-s390@vger.kernel.org>,
        "kvm@vger.kernel.org" <kvm@vger.kernel.org>,
        "intel-gvt-dev@lists.freedesktop.org" 
        <intel-gvt-dev@lists.freedesktop.org>,
        "intel-gfx@lists.freedesktop.org" <intel-gfx@lists.freedesktop.org>,
        "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>
Subject: RE: [PATCH] vfio: fix deadlock between group lock and kvm lock
Thread-Topic: [PATCH] vfio: fix deadlock between group lock and kvm lock
Thread-Index: AQHZNa+RvE0EhcYo6EGu/ai3XoxXDK66Canw
Date:   Wed, 1 Feb 2023 12:42:45 +0000
Message-ID: <DS0PR11MB752972035A4CE4B2319A6C45C3D19@DS0PR11MB7529.namprd11.prod.outlook.com>
References: <20230131200635.44227-1-mjrosato@linux.ibm.com>
In-Reply-To: <20230131200635.44227-1-mjrosato@linux.ibm.com>
Accept-Language: en-US
Content-Language: en-US
x-ms-exchange-antispam-messagedata-chunkcount: 1
x-ms-exchange-antispam-messagedata-0: =?us-ascii?Q?J6BjtMuZu+XFUL75XZl70cAfovIOA/XipQ8K9UAJwcuNv3H8bn/2nlfN1FLH?=
 =?us-ascii?Q?ymbq5XISPgyxtdp3pEuhLbYB1Xh+1V/Aciiu//bC3jONi1ajeGEYL+oGltsq?=
 =?us-ascii?Q?2Ln8ieTEjI/iK4oJudgtFmetcnje1iQ2uT3WHf3nbi6/mLMkfWlUU/cLZKet?=
 =?us-ascii?Q?6tJCsFl32wUBuH5lZ5ZS2MK7QPp3TLhs367ZrzajkMfTyOr0cEJ3usJLjJ9W?=
 =?us-ascii?Q?Mtb0T/DR7k/4SGKiaPER0iOuqfl35GiCd3JnJgvPDQkseKG1DCUi+5U7RahA?=
 =?us-ascii?Q?qrXjDxGmDzYkW9mokAz8iFzICGc1RfkjMAhY8oR2PUIyrHIaVLJUhbTy6RSD?=
 =?us-ascii?Q?VxxgIgZJ+DD+zWuPfIPMFlnIMjaPxDho6Tc0RiL6zzLgmCCAp3Dw+KVkHJ7x?=
 =?us-ascii?Q?Y8h8xGmLMtRf8A4AIqLxHHqzup+xK37YsgfL6HVD3UV2MyN+cLedDcrq44RQ?=
 =?us-ascii?Q?CqqrG89ErFMPC32C55HVH3PY74OH/fdPRoW95AqCiSX+lP8zL3hcR28qYqqE?=
 =?us-ascii?Q?80zpnnwPshOxvSPwYQDmwEGH+JjHIMtD97QNINeSA1XvJrV1wl4nEVZ3IoOe?=
 =?us-ascii?Q?6NOfW7UmcpIzjHGoBmSkppUAN0VOTQ47CmV1voR8yCUl4fbb4Xc3moa73W84?=
 =?us-ascii?Q?LgPvc3lqTMqzcgpHaleJ4aGwxmJDpSa4yuUmxRdafaoqE/STKqTQz5jhxHXz?=
 =?us-ascii?Q?EtGvOMjIBZf1tygIw66424hW8GpRP4c4X7EXe9QjJmTAvlA+GNC4cwrslk7L?=
 =?us-ascii?Q?LmjEjEWjt/c9HEh1WmndcL4RQi8wfdeji7X+3wt6gaR2xKtyEqeu2n7eD7jx?=
 =?us-ascii?Q?9zNX2ORbDjCM9N0NJrgEoOty1YbbOAFh9xHakBpOZf25orJO1DhLjFUjvVsA?=
 =?us-ascii?Q?KAp1Si480Y9JOgql43ySJ6Nb7Crw3MKWz6MCQwZ4DeuIieBNwsN81AecD8nv?=
 =?us-ascii?Q?QVDLVldLoP+CRtEl8pKGn2pnrm2LdtQOVD+WYKoCj6qcxs9Sx8wqDYVqAlZA?=
 =?us-ascii?Q?xtW+RlKCZX6OnvarbWAB7OIsJ1gFY3i6wHLOXYoYqlaj5Ax+9U/HJmoNfUcS?=
 =?us-ascii?Q?XPzIlTl0NrFjvqWgkx9ZkiqpiV3OhcgggQXd2QZbHhIImOhbD+8hWyfcapVR?=
 =?us-ascii?Q?HnbeccimKCRv7RlNMoDVv9fMb4EOhEJmoAk0YqHX0vL3s7YH/ZNRRFRb4qJM?=
 =?us-ascii?Q?sBEgOELwOmbszNpdv91YuDHBrXK2v/jHVtnKSBqZTX50PATSQFNT//Q5yCxL?=
 =?us-ascii?Q?n/WpsMv2giZlo8cwysVdeMXM6OLqHZtqZIoaZ0H2bvz3zkGoJzm7W8SJRnQB?=
 =?us-ascii?Q?CmJg+higu97nX9xgLuKrjMgQ/rLDLtRmfQieuE0Vd3zuwB0OMbvsf1IcDfb0?=
 =?us-ascii?Q?dvxd8F0Z++cPZqeOo9Mg/g7FVQynDuOaAVjDokIqa42XgNq7cbb7eYeMrH0m?=
 =?us-ascii?Q?v+1mLnz7mEaMuQ+g4Xfz+ZLU6Tga/BR+1CQjOpuC7+AMkJjSchhJVBSl8N/x?=
 =?us-ascii?Q?hzGQ4+QHYBDX8TDk3s2/CguU0778lNPfqdchGdhd8iMm81gdGe87KXJlwzpU?=
 =?us-ascii?Q?ZZP9yF08MFULnzaO1n1zwY959pvjecCS+cs4hKLZ?=
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
MIME-Version: 1.0
X-MS-Exchange-CrossTenant-AuthAs: Internal
X-MS-Exchange-CrossTenant-AuthSource: DS0PR11MB7529.namprd11.prod.outlook.com
X-MS-Exchange-CrossTenant-Network-Message-Id: 99180ceb-b378-4bc3-47ab-08db0451d162
X-MS-Exchange-CrossTenant-originalarrivaltime: 01 Feb 2023 12:42:45.4642
 (UTC)
X-MS-Exchange-CrossTenant-fromentityheader: Hosted
X-MS-Exchange-CrossTenant-id: 46c98d88-e344-4ed4-8496-4ed7712e255d
X-MS-Exchange-CrossTenant-mailboxtype: HOSTED
X-MS-Exchange-CrossTenant-userprincipalname: i2qt0y0kcs6iVrjUAEbWpZSbsHcaOy0ug1RJ8NSB6z12PBl+fj0/EuwlhBRSPLpdMUCwQL6fIpNOnR7Xxr6fLw==
X-MS-Exchange-Transport-CrossTenantHeadersStamped: IA1PR11MB6467
X-OriginatorOrg: intel.com
Precedence: bulk
List-ID: <linux-kernel.vger.kernel.org>
X-Mailing-List: linux-kernel@vger.kernel.org

> From: Matthew Rosato <mjrosato@linux.ibm.com>
> Sent: Wednesday, February 1, 2023 4:07 AM
>=20
> After 51cdc8bc120e, we have another deadlock scenario between the
> kvm->lock and the vfio group_lock with two different codepaths acquiring
> the locks in different order.  Specifically in vfio_open_device, vfio
> holds the vfio group_lock when issuing device->ops->open_device but
> some
> drivers (like vfio-ap) need to acquire kvm->lock during their open_device
> routine;  Meanwhile, kvm_vfio_release will acquire the kvm->lock first
> before calling vfio_file_set_kvm which will acquire the vfio group_lock.
>=20
> To resolve this, let's remove the need for the vfio group_lock from the
> kvm_vfio_release codepath.  This is done by introducing a new spinlock to
> protect modifications to the vfio group kvm pointer, and acquiring a kvm
> ref from within vfio while holding this spinlock, with the reference held
> until the last close for the device in question.
>=20
> Fixes: 51cdc8bc120e ("kvm/vfio: Fix potential deadlock on vfio group_lock=
")
> Reported-by: Anthony Krowiak <akrowiak@linux.ibm.com>
> Suggested-by: Jason Gunthorpe <jgg@nvidia.com>
> Signed-off-by: Matthew Rosato <mjrosato@linux.ibm.com>
> ---
>  drivers/vfio/group.c     | 14 +++----
>  drivers/vfio/vfio.h      |  5 ++-
>  drivers/vfio/vfio_main.c | 88
> ++++++++++++++++++++++++++++++++++++----
>  include/linux/vfio.h     |  2 +-
>  4 files changed, 88 insertions(+), 21 deletions(-)
>=20
> diff --git a/drivers/vfio/group.c b/drivers/vfio/group.c
> index bb24b2f0271e..8f67d33e0e8d 100644
> --- a/drivers/vfio/group.c
> +++ b/drivers/vfio/group.c
> @@ -164,13 +164,7 @@ static int vfio_device_group_open(struct
> vfio_device *device)
>  		goto out_unlock;
>  	}
>=20
> -	/*
> -	 * Here we pass the KVM pointer with the group under the lock.  If
> the
> -	 * device driver will use it, it must obtain a reference and release it
> -	 * during close_device.
> -	 */
> -	ret =3D vfio_device_open(device, device->group->iommufd,
> -			       device->group->kvm);
> +	ret =3D vfio_device_open(device, device->group);
>=20
>  out_unlock:
>  	mutex_unlock(&device->group->group_lock);
> @@ -450,6 +444,7 @@ static struct vfio_group *vfio_group_alloc(struct
> iommu_group *iommu_group,
>=20
>  	refcount_set(&group->drivers, 1);
>  	mutex_init(&group->group_lock);
> +	spin_lock_init(&group->kvm_ref_lock);
>  	INIT_LIST_HEAD(&group->device_list);
>  	mutex_init(&group->device_lock);
>  	group->iommu_group =3D iommu_group;
> @@ -799,13 +794,14 @@
> EXPORT_SYMBOL_GPL(vfio_file_enforced_coherent);
>  void vfio_file_set_kvm(struct file *file, struct kvm *kvm)
>  {
>  	struct vfio_group *group =3D file->private_data;
> +	unsigned long flags;
>=20
>  	if (!vfio_file_is_group(file))
>  		return;
>=20
> -	mutex_lock(&group->group_lock);
> +	spin_lock_irqsave(&group->kvm_ref_lock, flags);
>  	group->kvm =3D kvm;
> -	mutex_unlock(&group->group_lock);
> +	spin_unlock_irqrestore(&group->kvm_ref_lock, flags);
>  }
>  EXPORT_SYMBOL_GPL(vfio_file_set_kvm);
>=20
> diff --git a/drivers/vfio/vfio.h b/drivers/vfio/vfio.h
> index f8219a438bfb..d153ad35de24 100644
> --- a/drivers/vfio/vfio.h
> +++ b/drivers/vfio/vfio.h
> @@ -15,11 +15,11 @@ struct iommufd_ctx;
>  struct iommu_group;
>  struct vfio_device;
>  struct vfio_container;
> +struct vfio_group;
>=20
>  void vfio_device_put_registration(struct vfio_device *device);
>  bool vfio_device_try_get_registration(struct vfio_device *device);
> -int vfio_device_open(struct vfio_device *device,
> -		     struct iommufd_ctx *iommufd, struct kvm *kvm);
> +int vfio_device_open(struct vfio_device *device, struct vfio_group
> *group);
>  void vfio_device_close(struct vfio_device *device,
>  		       struct iommufd_ctx *iommufd);
>=20
> @@ -74,6 +74,7 @@ struct vfio_group {
>  	struct file			*opened_file;
>  	struct blocking_notifier_head	notifier;
>  	struct iommufd_ctx		*iommufd;
> +	spinlock_t			kvm_ref_lock;
>  };
>=20
>  int vfio_device_set_group(struct vfio_device *device,
> diff --git a/drivers/vfio/vfio_main.c b/drivers/vfio/vfio_main.c
> index 5177bb061b17..b0defb0d0d87 100644
> --- a/drivers/vfio/vfio_main.c
> +++ b/drivers/vfio/vfio_main.c
> @@ -16,6 +16,9 @@
>  #include <linux/fs.h>
>  #include <linux/idr.h>
>  #include <linux/iommu.h>
> +#ifdef CONFIG_HAVE_KVM
> +#include <linux/kvm_host.h>
> +#endif
>  #include <linux/list.h>
>  #include <linux/miscdevice.h>
>  #include <linux/module.h>
> @@ -344,9 +347,59 @@ static bool vfio_assert_device_open(struct
> vfio_device *device)
>  	return !WARN_ON_ONCE(!READ_ONCE(device->open_count));
>  }
>=20
> +#ifdef CONFIG_HAVE_KVM
> +static bool vfio_kvm_get_kvm_safe(struct vfio_device *device, struct kvm
> *kvm)
> +{
> +	void (*pfn)(struct kvm *kvm);
> +	bool (*fn)(struct kvm *kvm);
> +	bool ret;
> +
> +	pfn =3D symbol_get(kvm_put_kvm);
> +	if (WARN_ON(!pfn))
> +		return false;
> +
> +	fn =3D symbol_get(kvm_get_kvm_safe);
> +	if (WARN_ON(!fn)) {
> +		symbol_put(kvm_put_kvm);
> +		return false;
> +	}
> +
> +	ret =3D fn(kvm);
> +	if (ret)
> +		device->put_kvm =3D pfn;
> +	else
> +		symbol_put(kvm_put_kvm);
> +
> +	symbol_put(kvm_get_kvm_safe);
> +
> +	return ret;
> +}
> +
> +static void vfio_kvm_put_kvm(struct vfio_device *device)
> +{
> +	if (WARN_ON(!device->kvm || !device->put_kvm))
> +		return;
> +
> +	device->put_kvm(device->kvm);
> +	symbol_put(kvm_put_kvm);
> +}
> +
> +#else
> +static bool vfio_kvm_get_kvm_safe(struct vfio_device *device, struct kvm
> *kvm)
> +{
> +	return false;
> +}
> +
> +static void vfio_kvm_put_kvm(struct vfio_device *device)
> +{
> +}
> +#endif
> +
>  static int vfio_device_first_open(struct vfio_device *device,
> -				  struct iommufd_ctx *iommufd, struct kvm
> *kvm)
> +				  struct vfio_group *group)
>  {
> +	struct iommufd_ctx *iommufd =3D group->iommufd;
> +	unsigned long flags;
>  	int ret;
>=20
>  	lockdep_assert_held(&device->dev_set->lock);
> @@ -361,16 +414,30 @@ static int vfio_device_first_open(struct
> vfio_device *device,
>  	if (ret)
>  		goto err_module_put;
>=20
> -	device->kvm =3D kvm;
> +	/*
> +	 * Get the KVM pointer currently associated with the group, if there
> +	 * is one, and obtain a reference now that will be held until the
> +	 * open_count reaches 0.  Save the pointer in the device for use by
> +	 * drivers.
> +	 */
> +	spin_lock_irqsave(&group->kvm_ref_lock, flags);
> +	if (group->kvm && vfio_kvm_get_kvm_safe(device, group->kvm))
> +		device->kvm =3D group->kvm;
> +	spin_unlock_irqrestore(&group->kvm_ref_lock, flags);
> +
>  	if (device->ops->open_device) {
>  		ret =3D device->ops->open_device(device);
>  		if (ret)
> -			goto err_unuse_iommu;
> +			goto err_put_kvm;
>  	}
>  	return 0;
>=20
> -err_unuse_iommu:
> -	device->kvm =3D NULL;
> +err_put_kvm:
> +	if (device->kvm) {
> +		vfio_kvm_put_kvm(device);
> +		device->put_kvm =3D NULL;

Can "device->put_kvm =3D NULL;" be part of vfio_kvm_put_kvm()?

Regards,
Yi Liu