From: Sebastien Boeuf <[email protected]>
This series introduces a new operation for vdpa devices. It allows them
to be resumed after they have been suspended. A new feature bit is
introduced for devices to advertise their ability to be resumed after
they have been suspended. This feature bit is different from the one
advertising the ability to be suspended, meaning a device that can be
suspended might not have the ability to be resumed.
Even if it is already possible to restore a device that has been
suspended, which is very convenient for live migrating virtual machines,
there is a major drawback as the device must be fully reset. There is no
way to resume a device that has been suspended without having to
configure the device again and without having to recreate the IOMMU
mappings. This new operation aims at filling this gap by allowing the
device to resume processing the virtqueue descriptors without having to
reset it. This is particularly useful for performing virtual machine
offline migration, also called snapshot/restore, as it allows a virtual
machine to resume to a running state after it was paused and a snapshot
of the entire system was taken.
Sebastien Boeuf (4):
vdpa: Add resume operation
vhost-vdpa: Introduce RESUME backend feature bit
vhost-vdpa: uAPI to resume the device
vdpa_sim: Implement resume vdpa op
drivers/vdpa/vdpa_sim/vdpa_sim.c | 14 +++++++++++++
drivers/vhost/vdpa.c | 34 +++++++++++++++++++++++++++++++-
include/linux/vdpa.h | 6 +++++-
include/uapi/linux/vhost.h | 8 ++++++++
include/uapi/linux/vhost_types.h | 2 ++
5 files changed, 62 insertions(+), 2 deletions(-)
--
2.34.1
---------------------------------------------------------------------
Intel Corporation SAS (French simplified joint stock company)
Registered headquarters: "Les Montalets"- 2, rue de Paris,
92196 Meudon Cedex, France
Registration Number: 302 456 199 R.C.S. NANTERRE
Capital: 5 208 026.16 Euros
This e-mail and any attachments may contain confidential material for
the sole use of the intended recipient(s). Any review or distribution
by others is strictly prohibited. If you are not the intended
recipient, please contact the sender and delete all copies.
From: Sebastien Boeuf <[email protected]>
Add a new operation to allow a vDPA device to be resumed after it has
been suspended. Trying to resume a device that wasn't suspended will
result in a no-op.
This operation is optional. If it's not implemented, the associated
backend feature bit will not be exposed. And if the feature bit is not
exposed, invoking this operation will return an error.
Acked-by: Jason Wang <[email protected]>
Signed-off-by: Sebastien Boeuf <[email protected]>
---
include/linux/vdpa.h | 6 +++++-
1 file changed, 5 insertions(+), 1 deletion(-)
diff --git a/include/linux/vdpa.h b/include/linux/vdpa.h
index 6d0f5e4e82c2..96d308cbf97b 100644
--- a/include/linux/vdpa.h
+++ b/include/linux/vdpa.h
@@ -219,7 +219,10 @@ struct vdpa_map_file {
* @reset: Reset device
* @vdev: vdpa device
* Returns integer: success (0) or error (< 0)
- * @suspend: Suspend or resume the device (optional)
+ * @suspend: Suspend the device (optional)
+ * @vdev: vdpa device
+ * Returns integer: success (0) or error (< 0)
+ * @resume: Resume the device (optional)
* @vdev: vdpa device
* Returns integer: success (0) or error (< 0)
* @get_config_size: Get the size of the configuration space includes
@@ -324,6 +327,7 @@ struct vdpa_config_ops {
void (*set_status)(struct vdpa_device *vdev, u8 status);
int (*reset)(struct vdpa_device *vdev);
int (*suspend)(struct vdpa_device *vdev);
+ int (*resume)(struct vdpa_device *vdev);
size_t (*get_config_size)(struct vdpa_device *vdev);
void (*get_config)(struct vdpa_device *vdev, unsigned int offset,
void *buf, unsigned int len);
--
2.34.1
---------------------------------------------------------------------
Intel Corporation SAS (French simplified joint stock company)
Registered headquarters: "Les Montalets"- 2, rue de Paris,
92196 Meudon Cedex, France
Registration Number: 302 456 199 R.C.S. NANTERRE
Capital: 5 208 026.16 Euros
This e-mail and any attachments may contain confidential material for
the sole use of the intended recipient(s). Any review or distribution
by others is strictly prohibited. If you are not the intended
recipient, please contact the sender and delete all copies.
From: Sebastien Boeuf <[email protected]>
This new ioctl adds support for resuming the device from userspace.
This is required when trying to restore the device in a functioning
state after it's been suspended. It is already possible to reset a
suspended device, but that means the device must be reconfigured and
all the IOMMU/IOTLB mappings must be recreated. This new operation
allows the device to be resumed without going through a full reset.
This is particularly useful when trying to perform offline migration of
a virtual machine (also known as snapshot/restore) as it allows the VMM
to resume the virtual machine back to a running state after the snapshot
is performed.
Acked-by: Jason Wang <[email protected]>
Signed-off-by: Sebastien Boeuf <[email protected]>
---
drivers/vhost/vdpa.c | 18 ++++++++++++++++++
include/uapi/linux/vhost.h | 8 ++++++++
2 files changed, 26 insertions(+)
diff --git a/drivers/vhost/vdpa.c b/drivers/vhost/vdpa.c
index 833617d00ef6..1db7bd39fb63 100644
--- a/drivers/vhost/vdpa.c
+++ b/drivers/vhost/vdpa.c
@@ -502,6 +502,21 @@ static long vhost_vdpa_suspend(struct vhost_vdpa *v)
return ops->suspend(vdpa);
}
+/* After a successful return of this ioctl the device resumes processing
+ * virtqueue descriptors. The device becomes fully operational the same way it
+ * was before it was suspended.
+ */
+static long vhost_vdpa_resume(struct vhost_vdpa *v)
+{
+ struct vdpa_device *vdpa = v->vdpa;
+ const struct vdpa_config_ops *ops = vdpa->config;
+
+ if (!ops->resume)
+ return -EOPNOTSUPP;
+
+ return ops->resume(vdpa);
+}
+
static long vhost_vdpa_vring_ioctl(struct vhost_vdpa *v, unsigned int cmd,
void __user *argp)
{
@@ -687,6 +702,9 @@ static long vhost_vdpa_unlocked_ioctl(struct file *filep,
case VHOST_VDPA_SUSPEND:
r = vhost_vdpa_suspend(v);
break;
+ case VHOST_VDPA_RESUME:
+ r = vhost_vdpa_resume(v);
+ break;
default:
r = vhost_dev_ioctl(&v->vdev, cmd, argp);
if (r == -ENOIOCTLCMD)
diff --git a/include/uapi/linux/vhost.h b/include/uapi/linux/vhost.h
index f9f115a7c75b..92e1b700b51c 100644
--- a/include/uapi/linux/vhost.h
+++ b/include/uapi/linux/vhost.h
@@ -180,4 +180,12 @@
*/
#define VHOST_VDPA_SUSPEND _IO(VHOST_VIRTIO, 0x7D)
+/* Resume a device so it can resume processing virtqueue requests
+ *
+ * After the return of this ioctl the device will have restored all the
+ * necessary states and it is fully operational to continue processing the
+ * virtqueue descriptors.
+ */
+#define VHOST_VDPA_RESUME _IO(VHOST_VIRTIO, 0x7E)
+
#endif
--
2.34.1
---------------------------------------------------------------------
Intel Corporation SAS (French simplified joint stock company)
Registered headquarters: "Les Montalets"- 2, rue de Paris,
92196 Meudon Cedex, France
Registration Number: 302 456 199 R.C.S. NANTERRE
Capital: 5 208 026.16 Euros
This e-mail and any attachments may contain confidential material for
the sole use of the intended recipient(s). Any review or distribution
by others is strictly prohibited. If you are not the intended
recipient, please contact the sender and delete all copies.
From: Sebastien Boeuf <[email protected]>
Implement resume operation for vdpa_sim devices, so vhost-vdpa will
offer that backend feature and userspace can effectively resume the
device.
Signed-off-by: Sebastien Boeuf <[email protected]>
---
drivers/vdpa/vdpa_sim/vdpa_sim.c | 14 ++++++++++++++
1 file changed, 14 insertions(+)
diff --git a/drivers/vdpa/vdpa_sim/vdpa_sim.c b/drivers/vdpa/vdpa_sim/vdpa_sim.c
index b071f0d842fb..05e3802fb746 100644
--- a/drivers/vdpa/vdpa_sim/vdpa_sim.c
+++ b/drivers/vdpa/vdpa_sim/vdpa_sim.c
@@ -527,6 +527,18 @@ static int vdpasim_suspend(struct vdpa_device *vdpa)
return 0;
}
+static int vdpasim_resume(struct vdpa_device *vdpa)
+{
+ struct vdpasim *vdpasim = vdpa_to_sim(vdpa);
+ int i;
+
+ spin_lock(&vdpasim->lock);
+ vdpasim->running = true;
+ spin_unlock(&vdpasim->lock);
+
+ return 0;
+}
+
static size_t vdpasim_get_config_size(struct vdpa_device *vdpa)
{
struct vdpasim *vdpasim = vdpa_to_sim(vdpa);
@@ -717,6 +729,7 @@ static const struct vdpa_config_ops vdpasim_config_ops = {
.set_status = vdpasim_set_status,
.reset = vdpasim_reset,
.suspend = vdpasim_suspend,
+ .resume = vdpasim_resume,
.get_config_size = vdpasim_get_config_size,
.get_config = vdpasim_get_config,
.set_config = vdpasim_set_config,
@@ -750,6 +763,7 @@ static const struct vdpa_config_ops vdpasim_batch_config_ops = {
.set_status = vdpasim_set_status,
.reset = vdpasim_reset,
.suspend = vdpasim_suspend,
+ .resume = vdpasim_resume,
.get_config_size = vdpasim_get_config_size,
.get_config = vdpasim_get_config,
.set_config = vdpasim_set_config,
--
2.34.1
---------------------------------------------------------------------
Intel Corporation SAS (French simplified joint stock company)
Registered headquarters: "Les Montalets"- 2, rue de Paris,
92196 Meudon Cedex, France
Registration Number: 302 456 199 R.C.S. NANTERRE
Capital: 5 208 026.16 Euros
This e-mail and any attachments may contain confidential material for
the sole use of the intended recipient(s). Any review or distribution
by others is strictly prohibited. If you are not the intended
recipient, please contact the sender and delete all copies.
From: Sebastien Boeuf <[email protected]>
Userspace knows if the device can be resumed or not by checking this
feature bit.
It's only exposed if the vdpa driver backend implements the resume()
operation callback. Userspace trying to negotiate this feature when it
hasn't been exposed will result in an error.
Acked-by: Jason Wang <[email protected]>
Signed-off-by: Sebastien Boeuf <[email protected]>
---
drivers/vhost/vdpa.c | 16 +++++++++++++++-
include/uapi/linux/vhost_types.h | 2 ++
2 files changed, 17 insertions(+), 1 deletion(-)
diff --git a/drivers/vhost/vdpa.c b/drivers/vhost/vdpa.c
index 166044642fd5..833617d00ef6 100644
--- a/drivers/vhost/vdpa.c
+++ b/drivers/vhost/vdpa.c
@@ -355,6 +355,14 @@ static bool vhost_vdpa_can_suspend(const struct vhost_vdpa *v)
return ops->suspend;
}
+static bool vhost_vdpa_can_resume(const struct vhost_vdpa *v)
+{
+ struct vdpa_device *vdpa = v->vdpa;
+ const struct vdpa_config_ops *ops = vdpa->config;
+
+ return ops->resume;
+}
+
static long vhost_vdpa_get_features(struct vhost_vdpa *v, u64 __user *featurep)
{
struct vdpa_device *vdpa = v->vdpa;
@@ -602,11 +610,15 @@ static long vhost_vdpa_unlocked_ioctl(struct file *filep,
if (copy_from_user(&features, featurep, sizeof(features)))
return -EFAULT;
if (features & ~(VHOST_VDPA_BACKEND_FEATURES |
- BIT_ULL(VHOST_BACKEND_F_SUSPEND)))
+ BIT_ULL(VHOST_BACKEND_F_SUSPEND) |
+ BIT_ULL(VHOST_BACKEND_F_RESUME)))
return -EOPNOTSUPP;
if ((features & BIT_ULL(VHOST_BACKEND_F_SUSPEND)) &&
!vhost_vdpa_can_suspend(v))
return -EOPNOTSUPP;
+ if ((features & BIT_ULL(VHOST_BACKEND_F_RESUME)) &&
+ !vhost_vdpa_can_resume(v))
+ return -EOPNOTSUPP;
vhost_set_backend_features(&v->vdev, features);
return 0;
}
@@ -658,6 +670,8 @@ static long vhost_vdpa_unlocked_ioctl(struct file *filep,
features = VHOST_VDPA_BACKEND_FEATURES;
if (vhost_vdpa_can_suspend(v))
features |= BIT_ULL(VHOST_BACKEND_F_SUSPEND);
+ if (vhost_vdpa_can_resume(v))
+ features |= BIT_ULL(VHOST_BACKEND_F_RESUME);
if (copy_to_user(featurep, &features, sizeof(features)))
r = -EFAULT;
break;
diff --git a/include/uapi/linux/vhost_types.h b/include/uapi/linux/vhost_types.h
index 53601ce2c20a..c5690a8992d8 100644
--- a/include/uapi/linux/vhost_types.h
+++ b/include/uapi/linux/vhost_types.h
@@ -163,5 +163,7 @@ struct vhost_vdpa_iova_range {
#define VHOST_BACKEND_F_IOTLB_ASID 0x3
/* Device can be suspended */
#define VHOST_BACKEND_F_SUSPEND 0x4
+/* Device can be resumed */
+#define VHOST_BACKEND_F_RESUME 0x5
#endif
--
2.34.1
---------------------------------------------------------------------
Intel Corporation SAS (French simplified joint stock company)
Registered headquarters: "Les Montalets"- 2, rue de Paris,
92196 Meudon Cedex, France
Registration Number: 302 456 199 R.C.S. NANTERRE
Capital: 5 208 026.16 Euros
This e-mail and any attachments may contain confidential material for
the sole use of the intended recipient(s). Any review or distribution
by others is strictly prohibited. If you are not the intended
recipient, please contact the sender and delete all copies.
Hi,
Thank you for the patch! Perhaps something to improve:
[auto build test WARNING on mst-vhost/linux-next]
[also build test WARNING on linus/master v6.1-rc1 next-20221018]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch#_base_tree_information]
url: https://github.com/intel-lab-lkp/linux/commits/sebastien-boeuf-intel-com/vdpa-Add-resume-operation/20221018-163858
base: https://git.kernel.org/pub/scm/linux/kernel/git/mst/vhost.git linux-next
patch link: https://lore.kernel.org/r/56c045ac70e44e7d80f3f9e901deae3d7485b2a1.1666082013.git.sebastien.boeuf%40intel.com
patch subject: [PATCH v4 4/4] vdpa_sim: Implement resume vdpa op
config: x86_64-allyesconfig
compiler: gcc-11 (Debian 11.3.0-8) 11.3.0
reproduce (this is a W=1 build):
# https://github.com/intel-lab-lkp/linux/commit/63349eb63715840620ea16e0098110ad5883b901
git remote add linux-review https://github.com/intel-lab-lkp/linux
git fetch --no-tags linux-review sebastien-boeuf-intel-com/vdpa-Add-resume-operation/20221018-163858
git checkout 63349eb63715840620ea16e0098110ad5883b901
# save the config file
mkdir build_dir && cp config build_dir/.config
make W=1 O=build_dir ARCH=x86_64 SHELL=/bin/bash drivers/vdpa/
If you fix the issue, kindly add following tag where applicable
| Reported-by: kernel test robot <[email protected]>
All warnings (new ones prefixed by >>):
drivers/vdpa/vdpa_sim/vdpa_sim.c: In function 'vdpasim_resume':
>> drivers/vdpa/vdpa_sim/vdpa_sim.c:533:13: warning: unused variable 'i' [-Wunused-variable]
533 | int i;
| ^
vim +/i +533 drivers/vdpa/vdpa_sim/vdpa_sim.c
529
530 static int vdpasim_resume(struct vdpa_device *vdpa)
531 {
532 struct vdpasim *vdpasim = vdpa_to_sim(vdpa);
> 533 int i;
534
535 spin_lock(&vdpasim->lock);
536 vdpasim->running = true;
537 spin_unlock(&vdpasim->lock);
538
539 return 0;
540 }
541
--
0-DAY CI Kernel Test Service
https://01.org/lkp
On Wed, 2022-10-19 at 11:31 +0200, Eugenio Perez Martin wrote:
> On Tue, Oct 18, 2022 at 10:38 AM <[email protected]> wrote:
> >
> > From: Sebastien Boeuf <[email protected]>
> >
> > Implement resume operation for vdpa_sim devices, so vhost-vdpa will
> > offer that backend feature and userspace can effectively resume the
> > device.
> >
> > Signed-off-by: Sebastien Boeuf <[email protected]>
> > ---
> > drivers/vdpa/vdpa_sim/vdpa_sim.c | 14 ++++++++++++++
> > 1 file changed, 14 insertions(+)
> >
> > diff --git a/drivers/vdpa/vdpa_sim/vdpa_sim.c
> > b/drivers/vdpa/vdpa_sim/vdpa_sim.c
> > index b071f0d842fb..05e3802fb746 100644
> > --- a/drivers/vdpa/vdpa_sim/vdpa_sim.c
> > +++ b/drivers/vdpa/vdpa_sim/vdpa_sim.c
> > @@ -527,6 +527,18 @@ static int vdpasim_suspend(struct vdpa_device
> > *vdpa)
> > return 0;
> > }
> >
> > +static int vdpasim_resume(struct vdpa_device *vdpa)
> > +{
> > + struct vdpasim *vdpasim = vdpa_to_sim(vdpa);
> > + int i;
> > +
> > + spin_lock(&vdpasim->lock);
> > + vdpasim->running = true;
> > + spin_unlock(&vdpasim->lock);
> > +
> > + return 0;
> > +}
> > +
>
> To never kick at resuming is not the right thing to do :).
>
> Maybe to store in the vdpasim_virtqueue if it was kicked during the
> suspend window?
>
> Thanks!
Let's hear what Michael think about this approach?
I just want to make sure we're all on the same page before I send the
next version :)
Thanks,
Sebastien
>
>
>
> > static size_t vdpasim_get_config_size(struct vdpa_device *vdpa)
> > {
> > struct vdpasim *vdpasim = vdpa_to_sim(vdpa);
> > @@ -717,6 +729,7 @@ static const struct vdpa_config_ops
> > vdpasim_config_ops = {
> > .set_status = vdpasim_set_status,
> > .reset = vdpasim_reset,
> > .suspend = vdpasim_suspend,
> > + .resume = vdpasim_resume,
> > .get_config_size = vdpasim_get_config_size,
> > .get_config = vdpasim_get_config,
> > .set_config = vdpasim_set_config,
> > @@ -750,6 +763,7 @@ static const struct vdpa_config_ops
> > vdpasim_batch_config_ops = {
> > .set_status = vdpasim_set_status,
> > .reset = vdpasim_reset,
> > .suspend = vdpasim_suspend,
> > + .resume = vdpasim_resume,
> > .get_config_size = vdpasim_get_config_size,
> > .get_config = vdpasim_get_config,
> > .set_config = vdpasim_set_config,
> > --
> > 2.34.1
> >
> > -------------------------------------------------------------------
> > --
> > Intel Corporation SAS (French simplified joint stock company)
> > Registered headquarters: "Les Montalets"- 2, rue de Paris,
> > 92196 Meudon Cedex, France
> > Registration Number: 302 456 199 R.C.S. NANTERRE
> > Capital: 5 208 026.16 Euros
> >
> > This e-mail and any attachments may contain confidential material
> > for
> > the sole use of the intended recipient(s). Any review or
> > distribution
> > by others is strictly prohibited. If you are not the intended
> > recipient, please contact the sender and delete all copies.
> >
>
---------------------------------------------------------------------
Intel Corporation SAS (French simplified joint stock company)
Registered headquarters: "Les Montalets"- 2, rue de Paris,
92196 Meudon Cedex, France
Registration Number: 302 456 199 R.C.S. NANTERRE
Capital: 5 208 026.16 Euros
This e-mail and any attachments may contain confidential material for
the sole use of the intended recipient(s). Any review or distribution
by others is strictly prohibited. If you are not the intended
recipient, please contact the sender and delete all copies.
On Tue, Oct 18, 2022 at 10:38 AM <[email protected]> wrote:
>
> From: Sebastien Boeuf <[email protected]>
>
> Implement resume operation for vdpa_sim devices, so vhost-vdpa will
> offer that backend feature and userspace can effectively resume the
> device.
>
> Signed-off-by: Sebastien Boeuf <[email protected]>
> ---
> drivers/vdpa/vdpa_sim/vdpa_sim.c | 14 ++++++++++++++
> 1 file changed, 14 insertions(+)
>
> diff --git a/drivers/vdpa/vdpa_sim/vdpa_sim.c b/drivers/vdpa/vdpa_sim/vdpa_sim.c
> index b071f0d842fb..05e3802fb746 100644
> --- a/drivers/vdpa/vdpa_sim/vdpa_sim.c
> +++ b/drivers/vdpa/vdpa_sim/vdpa_sim.c
> @@ -527,6 +527,18 @@ static int vdpasim_suspend(struct vdpa_device *vdpa)
> return 0;
> }
>
> +static int vdpasim_resume(struct vdpa_device *vdpa)
> +{
> + struct vdpasim *vdpasim = vdpa_to_sim(vdpa);
> + int i;
> +
> + spin_lock(&vdpasim->lock);
> + vdpasim->running = true;
> + spin_unlock(&vdpasim->lock);
> +
> + return 0;
> +}
> +
To never kick at resuming is not the right thing to do :).
Maybe to store in the vdpasim_virtqueue if it was kicked during the
suspend window?
Thanks!
> static size_t vdpasim_get_config_size(struct vdpa_device *vdpa)
> {
> struct vdpasim *vdpasim = vdpa_to_sim(vdpa);
> @@ -717,6 +729,7 @@ static const struct vdpa_config_ops vdpasim_config_ops = {
> .set_status = vdpasim_set_status,
> .reset = vdpasim_reset,
> .suspend = vdpasim_suspend,
> + .resume = vdpasim_resume,
> .get_config_size = vdpasim_get_config_size,
> .get_config = vdpasim_get_config,
> .set_config = vdpasim_set_config,
> @@ -750,6 +763,7 @@ static const struct vdpa_config_ops vdpasim_batch_config_ops = {
> .set_status = vdpasim_set_status,
> .reset = vdpasim_reset,
> .suspend = vdpasim_suspend,
> + .resume = vdpasim_resume,
> .get_config_size = vdpasim_get_config_size,
> .get_config = vdpasim_get_config,
> .set_config = vdpasim_set_config,
> --
> 2.34.1
>
> ---------------------------------------------------------------------
> Intel Corporation SAS (French simplified joint stock company)
> Registered headquarters: "Les Montalets"- 2, rue de Paris,
> 92196 Meudon Cedex, France
> Registration Number: 302 456 199 R.C.S. NANTERRE
> Capital: 5 208 026.16 Euros
>
> This e-mail and any attachments may contain confidential material for
> the sole use of the intended recipient(s). Any review or distribution
> by others is strictly prohibited. If you are not the intended
> recipient, please contact the sender and delete all copies.
>