Re: [Intel-gfx] [PATCH v2] vfio: fix deadlock between group lock and kvm lock
On 2/1/23 11:10 PM, Tian, Kevin wrote: >> From: Alex Williamson >> Sent: Thursday, February 2, 2023 7:28 AM >>> >>> +#ifdef CONFIG_HAVE_KVM >>> +static bool vfio_kvm_get_kvm_safe(struct vfio_device *device, struct kvm >> *kvm) >> >> I'm tempted to name these vfio_device_get_kvm_safe() and only pass the >> vfio_device, where of course we can get the kvm pointer from the group >> internally. >> > > I have a different thought. In the end the cdev series also need the similar > safe get/put logic then it's better to keep it in vfio_main.c called by > the group/cdev path individually. Ah, I hadn't considered the cdev series - OK, I can move the functions back into vfio_main and externalize both via drivers/vfio/vfio.h so they can be called from group.c for this fix and then available to vfio_main.c already for cdev.
Re: [Intel-gfx] [PATCH v2] vfio: fix deadlock between group lock and kvm lock
> -Original Message-
> From: Alex Williamson
> Sent: Thursday, February 2, 2023 12:15 PM
> To: Liu, Yi L
> Cc: Matthew Rosato ; pbonz...@redhat.com;
> j...@nvidia.com; coh...@redhat.com; far...@linux.ibm.com;
> pmo...@linux.ibm.com; borntrae...@linux.ibm.com;
> fran...@linux.ibm.com; imbre...@linux.ibm.com; da...@redhat.com;
> akrow...@linux.ibm.com; jjhe...@linux.ibm.com; pa...@linux.ibm.com;
> zhen...@linux.intel.com; Wang, Zhi A ;
> Christopherson,, Sean ; Tian, Kevin
> ; linux-s...@vger.kernel.org; k...@vger.kernel.org;
> intel-gvt-...@lists.freedesktop.org; intel-gfx@lists.freedesktop.org; linux-
> ker...@vger.kernel.org
> Subject: Re: [PATCH v2] vfio: fix deadlock between group lock and kvm lock
>
> On Thu, 2 Feb 2023 03:46:59 +
> "Liu, Yi L" wrote:
>
> > > From: Alex Williamson
> > > Sent: Thursday, February 2, 2023 7:28 AM
> > >
> > > On Wed, 1 Feb 2023 14:20:10 -0500
> > > Matthew Rosato wrote:
> > >
> > > > 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.
> > > >
> > > > 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.
> > > >
> > > > Fixes: 51cdc8bc120e ("kvm/vfio: Fix potential deadlock on vfio
> group_lock")
> > > > Reported-by: Anthony Krowiak
> > > > Suggested-by: Jason Gunthorpe
> > > > Signed-off-by: Matthew Rosato
> > > > ---
> > > > Changes from v1:
> > > > * use spin_lock instead of spin_lock_irqsave (Jason)
> > > > * clear device->kvm_put as part of vfio_kvm_put_kvm (Yi)
> > > > * Re-arrange code to avoid referencing the group contents from
> within
> > > > vfio_main (Kevin) which meant moving most of the code in this patch
> > > > to group.c along with getting/dropping of the dev_set lock
> > > > ---
> > > > drivers/vfio/group.c | 90
> > > +---
> > > > drivers/vfio/vfio.h | 1 +
> > > > drivers/vfio/vfio_main.c | 11 ++---
> > > > include/linux/vfio.h | 2 +-
> > > > 4 files changed, 91 insertions(+), 13 deletions(-)
> > > >
> > > > diff --git a/drivers/vfio/group.c b/drivers/vfio/group.c
> > > > index bb24b2f0271e..52f434861294 100644
> > > > --- a/drivers/vfio/group.c
> > > > +++ b/drivers/vfio/group.c
> > > > @@ -13,6 +13,9 @@
> > > > #include
> > > > #include
> > > > #include
> > > > +#ifdef CONFIG_HAVE_KVM
> > > > +#include
> > > > +#endif
> > > > #include "vfio.h"
> > > >
> > > > static struct vfio {
> > > > @@ -154,6 +157,55 @@ static int
> vfio_group_ioctl_set_container(struct
> > > vfio_group *group,
> > > > return ret;
> > > > }
> > > >
> > > > +#ifdef CONFIG_HAVE_KVM
> > > > +static bool vfio_kvm_get_kvm_safe(struct vfio_device *device,
> struct
> > > kvm *kvm)
> > >
> > > I'm tempted to name these vfio_device_get_kvm_safe() and only pass
> the
> > > vfio_device, where of course we can get the kvm pointer from the
> group
> > > internally.
> > >
> > > > +{
> > > > + void (*pfn)(struct kvm *kvm);
> > > > + bool (*fn)(struct kvm *kvm);
> > > > + bool ret;
> > > > +
> > >
> > > We should assert_lockdep_held(>dev_set->lock) in both of
> these
> > > since that seems to be what's protecting device->kvm and
> > > device->put_kvm.
> > >
> > > If we change as above to get the kvm pointer from the group within this
> > > function, we can also move the kvm_ref_lock here, which seems to
> > > simplify the caller quite a bit.
> > >
> > > > + pfn = symbol_get(kvm_put_kvm);
> > > > + if (WARN_ON(!pfn))
> > > > + return false;
> > > > +
> > > > + fn = symbol_get(kvm_get_kvm_safe);
> > > > + if (WARN_ON(!fn)) {
> > > > + symbol_put(kvm_put_kvm);
> > > > + return false;
> > > > + }
> > > > +
> > > > + ret = fn(kvm);
> > > > + if (ret)
> > > > + device->put_kvm = 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;
> > >
> > > It simplifies the
Re: [Intel-gfx] [PATCH v2] vfio: fix deadlock between group lock and kvm lock
On Thu, 2 Feb 2023 03:46:59 +
"Liu, Yi L" wrote:
> > From: Alex Williamson
> > Sent: Thursday, February 2, 2023 7:28 AM
> >
> > On Wed, 1 Feb 2023 14:20:10 -0500
> > Matthew Rosato wrote:
> >
> > > 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.
> > >
> > > 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.
> > >
> > > Fixes: 51cdc8bc120e ("kvm/vfio: Fix potential deadlock on vfio
> > > group_lock")
> > > Reported-by: Anthony Krowiak
> > > Suggested-by: Jason Gunthorpe
> > > Signed-off-by: Matthew Rosato
> > > ---
> > > Changes from v1:
> > > * use spin_lock instead of spin_lock_irqsave (Jason)
> > > * clear device->kvm_put as part of vfio_kvm_put_kvm (Yi)
> > > * Re-arrange code to avoid referencing the group contents from within
> > > vfio_main (Kevin) which meant moving most of the code in this patch
> > > to group.c along with getting/dropping of the dev_set lock
> > > ---
> > > drivers/vfio/group.c | 90
> > +---
> > > drivers/vfio/vfio.h | 1 +
> > > drivers/vfio/vfio_main.c | 11 ++---
> > > include/linux/vfio.h | 2 +-
> > > 4 files changed, 91 insertions(+), 13 deletions(-)
> > >
> > > diff --git a/drivers/vfio/group.c b/drivers/vfio/group.c
> > > index bb24b2f0271e..52f434861294 100644
> > > --- a/drivers/vfio/group.c
> > > +++ b/drivers/vfio/group.c
> > > @@ -13,6 +13,9 @@
> > > #include
> > > #include
> > > #include
> > > +#ifdef CONFIG_HAVE_KVM
> > > +#include
> > > +#endif
> > > #include "vfio.h"
> > >
> > > static struct vfio {
> > > @@ -154,6 +157,55 @@ static int vfio_group_ioctl_set_container(struct
> > vfio_group *group,
> > > return ret;
> > > }
> > >
> > > +#ifdef CONFIG_HAVE_KVM
> > > +static bool vfio_kvm_get_kvm_safe(struct vfio_device *device, struct
> > kvm *kvm)
> >
> > I'm tempted to name these vfio_device_get_kvm_safe() and only pass the
> > vfio_device, where of course we can get the kvm pointer from the group
> > internally.
> >
> > > +{
> > > + void (*pfn)(struct kvm *kvm);
> > > + bool (*fn)(struct kvm *kvm);
> > > + bool ret;
> > > +
> >
> > We should assert_lockdep_held(>dev_set->lock) in both of these
> > since that seems to be what's protecting device->kvm and
> > device->put_kvm.
> >
> > If we change as above to get the kvm pointer from the group within this
> > function, we can also move the kvm_ref_lock here, which seems to
> > simplify the caller quite a bit.
> >
> > > + pfn = symbol_get(kvm_put_kvm);
> > > + if (WARN_ON(!pfn))
> > > + return false;
> > > +
> > > + fn = symbol_get(kvm_get_kvm_safe);
> > > + if (WARN_ON(!fn)) {
> > > + symbol_put(kvm_put_kvm);
> > > + return false;
> > > + }
> > > +
> > > + ret = fn(kvm);
> > > + if (ret)
> > > + device->put_kvm = 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;
> >
> > It simplifies the caller if we can use this even in the !device->kvm
> > case.
> >
> > > +
> > > + device->put_kvm(device->kvm);
> > > + device->put_kvm = NULL;
> > > + 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_group_open(struct vfio_device *device)
> > > {
> > > int ret;
> > > @@ -164,14 +216,32 @@ static int vfio_device_group_open(struct
> > vfio_device *device)
> > > goto out_unlock;
> > > }
> > >
> > > + mutex_lock(>dev_set->lock);
> > > +
> > > /*
> > > - * 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.
> > > + * Before the first device open, get the KVM pointer currently
> > > + * associated with the group (if
Re: [Intel-gfx] [PATCH v2] vfio: fix deadlock between group lock and kvm lock
> From: Alex Williamson > Sent: Thursday, February 2, 2023 7:28 AM > > > > +#ifdef CONFIG_HAVE_KVM > > +static bool vfio_kvm_get_kvm_safe(struct vfio_device *device, struct kvm > *kvm) > > I'm tempted to name these vfio_device_get_kvm_safe() and only pass the > vfio_device, where of course we can get the kvm pointer from the group > internally. > I have a different thought. In the end the cdev series also need the similar safe get/put logic then it's better to keep it in vfio_main.c called by the group/cdev path individually.
Re: [Intel-gfx] [PATCH v2] vfio: fix deadlock between group lock and kvm lock
> From: Liu, Yi L > Sent: Thursday, February 2, 2023 11:47 AM > > > ret = vfio_device_open(device, device->group->iommufd, > > > device->group->kvm); > > > > We're using device->group->kvm outside of kvm_ref_lock here, it should > > be using device->kvm. > > Existing code set device->kvm in the vfio_device_first_open() which is > called by vfio_device_open(). After above change, seems not necessary > to pass kvm pointer into the call chain. Isn't it? > Looks so.
Re: [Intel-gfx] [PATCH v2] vfio: fix deadlock between group lock and kvm lock
> From: Alex Williamson
> Sent: Thursday, February 2, 2023 7:28 AM
>
> On Wed, 1 Feb 2023 14:20:10 -0500
> Matthew Rosato wrote:
>
> > 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.
> >
> > 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.
> >
> > Fixes: 51cdc8bc120e ("kvm/vfio: Fix potential deadlock on vfio group_lock")
> > Reported-by: Anthony Krowiak
> > Suggested-by: Jason Gunthorpe
> > Signed-off-by: Matthew Rosato
> > ---
> > Changes from v1:
> > * use spin_lock instead of spin_lock_irqsave (Jason)
> > * clear device->kvm_put as part of vfio_kvm_put_kvm (Yi)
> > * Re-arrange code to avoid referencing the group contents from within
> > vfio_main (Kevin) which meant moving most of the code in this patch
> > to group.c along with getting/dropping of the dev_set lock
> > ---
> > drivers/vfio/group.c | 90
> +---
> > drivers/vfio/vfio.h | 1 +
> > drivers/vfio/vfio_main.c | 11 ++---
> > include/linux/vfio.h | 2 +-
> > 4 files changed, 91 insertions(+), 13 deletions(-)
> >
> > diff --git a/drivers/vfio/group.c b/drivers/vfio/group.c
> > index bb24b2f0271e..52f434861294 100644
> > --- a/drivers/vfio/group.c
> > +++ b/drivers/vfio/group.c
> > @@ -13,6 +13,9 @@
> > #include
> > #include
> > #include
> > +#ifdef CONFIG_HAVE_KVM
> > +#include
> > +#endif
> > #include "vfio.h"
> >
> > static struct vfio {
> > @@ -154,6 +157,55 @@ static int vfio_group_ioctl_set_container(struct
> vfio_group *group,
> > return ret;
> > }
> >
> > +#ifdef CONFIG_HAVE_KVM
> > +static bool vfio_kvm_get_kvm_safe(struct vfio_device *device, struct
> kvm *kvm)
>
> I'm tempted to name these vfio_device_get_kvm_safe() and only pass the
> vfio_device, where of course we can get the kvm pointer from the group
> internally.
>
> > +{
> > + void (*pfn)(struct kvm *kvm);
> > + bool (*fn)(struct kvm *kvm);
> > + bool ret;
> > +
>
> We should assert_lockdep_held(>dev_set->lock) in both of these
> since that seems to be what's protecting device->kvm and
> device->put_kvm.
>
> If we change as above to get the kvm pointer from the group within this
> function, we can also move the kvm_ref_lock here, which seems to
> simplify the caller quite a bit.
>
> > + pfn = symbol_get(kvm_put_kvm);
> > + if (WARN_ON(!pfn))
> > + return false;
> > +
> > + fn = symbol_get(kvm_get_kvm_safe);
> > + if (WARN_ON(!fn)) {
> > + symbol_put(kvm_put_kvm);
> > + return false;
> > + }
> > +
> > + ret = fn(kvm);
> > + if (ret)
> > + device->put_kvm = 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;
>
> It simplifies the caller if we can use this even in the !device->kvm
> case.
>
> > +
> > + device->put_kvm(device->kvm);
> > + device->put_kvm = NULL;
> > + 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_group_open(struct vfio_device *device)
> > {
> > int ret;
> > @@ -164,14 +216,32 @@ static int vfio_device_group_open(struct
> vfio_device *device)
> > goto out_unlock;
> > }
> >
> > + mutex_lock(>dev_set->lock);
> > +
> > /*
> > -* 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.
> > +* Before the first device open, 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 again. Save
> > +* the pointer in the device for use by drivers.
> > */
> > + if (device->open_count == 0) {
> > + spin_lock(>group->kvm_ref_lock);
> > + if (device->group->kvm &&
> > +
Re: [Intel-gfx] [PATCH v2] vfio: fix deadlock between group lock and kvm lock
On Wed, 1 Feb 2023 14:20:10 -0500
Matthew Rosato wrote:
> 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.
>
> 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.
>
> Fixes: 51cdc8bc120e ("kvm/vfio: Fix potential deadlock on vfio group_lock")
> Reported-by: Anthony Krowiak
> Suggested-by: Jason Gunthorpe
> Signed-off-by: Matthew Rosato
> ---
> Changes from v1:
> * use spin_lock instead of spin_lock_irqsave (Jason)
> * clear device->kvm_put as part of vfio_kvm_put_kvm (Yi)
> * Re-arrange code to avoid referencing the group contents from within
> vfio_main (Kevin) which meant moving most of the code in this patch
> to group.c along with getting/dropping of the dev_set lock
> ---
> drivers/vfio/group.c | 90 +---
> drivers/vfio/vfio.h | 1 +
> drivers/vfio/vfio_main.c | 11 ++---
> include/linux/vfio.h | 2 +-
> 4 files changed, 91 insertions(+), 13 deletions(-)
>
> diff --git a/drivers/vfio/group.c b/drivers/vfio/group.c
> index bb24b2f0271e..52f434861294 100644
> --- a/drivers/vfio/group.c
> +++ b/drivers/vfio/group.c
> @@ -13,6 +13,9 @@
> #include
> #include
> #include
> +#ifdef CONFIG_HAVE_KVM
> +#include
> +#endif
> #include "vfio.h"
>
> static struct vfio {
> @@ -154,6 +157,55 @@ static int vfio_group_ioctl_set_container(struct
> vfio_group *group,
> return ret;
> }
>
> +#ifdef CONFIG_HAVE_KVM
> +static bool vfio_kvm_get_kvm_safe(struct vfio_device *device, struct kvm
> *kvm)
I'm tempted to name these vfio_device_get_kvm_safe() and only pass the
vfio_device, where of course we can get the kvm pointer from the group
internally.
> +{
> + void (*pfn)(struct kvm *kvm);
> + bool (*fn)(struct kvm *kvm);
> + bool ret;
> +
We should assert_lockdep_held(>dev_set->lock) in both of these
since that seems to be what's protecting device->kvm and
device->put_kvm.
If we change as above to get the kvm pointer from the group within this
function, we can also move the kvm_ref_lock here, which seems to
simplify the caller quite a bit.
> + pfn = symbol_get(kvm_put_kvm);
> + if (WARN_ON(!pfn))
> + return false;
> +
> + fn = symbol_get(kvm_get_kvm_safe);
> + if (WARN_ON(!fn)) {
> + symbol_put(kvm_put_kvm);
> + return false;
> + }
> +
> + ret = fn(kvm);
> + if (ret)
> + device->put_kvm = 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;
It simplifies the caller if we can use this even in the !device->kvm
case.
> +
> + device->put_kvm(device->kvm);
> + device->put_kvm = NULL;
> + 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_group_open(struct vfio_device *device)
> {
> int ret;
> @@ -164,14 +216,32 @@ static int vfio_device_group_open(struct vfio_device
> *device)
> goto out_unlock;
> }
>
> + mutex_lock(>dev_set->lock);
> +
> /*
> - * 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.
> + * Before the first device open, 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 again. Save
> + * the pointer in the device for use by drivers.
>*/
> + if (device->open_count == 0) {
> + spin_lock(>group->kvm_ref_lock);
> + if (device->group->kvm &&
> + vfio_kvm_get_kvm_safe(device, device->group->kvm))
> + device->kvm = device->group->kvm;
> + spin_unlock(>group->kvm_ref_lock);
> + }
> +
> ret = vfio_device_open(device, device->group->iommufd,
>
[Intel-gfx] [PATCH v2] vfio: fix deadlock between group lock and kvm lock
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.
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.
Fixes: 51cdc8bc120e ("kvm/vfio: Fix potential deadlock on vfio group_lock")
Reported-by: Anthony Krowiak
Suggested-by: Jason Gunthorpe
Signed-off-by: Matthew Rosato
---
Changes from v1:
* use spin_lock instead of spin_lock_irqsave (Jason)
* clear device->kvm_put as part of vfio_kvm_put_kvm (Yi)
* Re-arrange code to avoid referencing the group contents from within
vfio_main (Kevin) which meant moving most of the code in this patch
to group.c along with getting/dropping of the dev_set lock
---
drivers/vfio/group.c | 90 +---
drivers/vfio/vfio.h | 1 +
drivers/vfio/vfio_main.c | 11 ++---
include/linux/vfio.h | 2 +-
4 files changed, 91 insertions(+), 13 deletions(-)
diff --git a/drivers/vfio/group.c b/drivers/vfio/group.c
index bb24b2f0271e..52f434861294 100644
--- a/drivers/vfio/group.c
+++ b/drivers/vfio/group.c
@@ -13,6 +13,9 @@
#include
#include
#include
+#ifdef CONFIG_HAVE_KVM
+#include
+#endif
#include "vfio.h"
static struct vfio {
@@ -154,6 +157,55 @@ static int vfio_group_ioctl_set_container(struct
vfio_group *group,
return ret;
}
+#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 = symbol_get(kvm_put_kvm);
+ if (WARN_ON(!pfn))
+ return false;
+
+ fn = symbol_get(kvm_get_kvm_safe);
+ if (WARN_ON(!fn)) {
+ symbol_put(kvm_put_kvm);
+ return false;
+ }
+
+ ret = fn(kvm);
+ if (ret)
+ device->put_kvm = 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);
+ device->put_kvm = NULL;
+ 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_group_open(struct vfio_device *device)
{
int ret;
@@ -164,14 +216,32 @@ static int vfio_device_group_open(struct vfio_device
*device)
goto out_unlock;
}
+ mutex_lock(>dev_set->lock);
+
/*
-* 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.
+* Before the first device open, 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 again. Save
+* the pointer in the device for use by drivers.
*/
+ if (device->open_count == 0) {
+ spin_lock(>group->kvm_ref_lock);
+ if (device->group->kvm &&
+ vfio_kvm_get_kvm_safe(device, device->group->kvm))
+ device->kvm = device->group->kvm;
+ spin_unlock(>group->kvm_ref_lock);
+ }
+
ret = vfio_device_open(device, device->group->iommufd,
device->group->kvm);
+ if (ret && device->kvm && device->open_count == 0) {
+ vfio_kvm_put_kvm(device);
+ device->kvm = NULL;
+ }
+
+ mutex_unlock(>dev_set->lock);
+
out_unlock:
mutex_unlock(>group->group_lock);
return ret;
@@ -180,7 +250,16 @@ static int vfio_device_group_open(struct vfio_device
*device)
void vfio_device_group_close(struct vfio_device *device)
{
mutex_lock(>group->group_lock);
+ mutex_lock(>dev_set->lock);
+
vfio_device_close(device, device->group->iommufd);
+
+ if (device->kvm && device->open_count == 0) {
+ vfio_kvm_put_kvm(device);
+ device->kvm = NULL;
+ }
+
+ mutex_unlock(>dev_set->lock);
