| Message ID | 20220524170610.2255608-5-eperezma@redhat.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | Implement vdpasim stop operation | expand |
On Wed, May 25, 2022 at 1:06 AM Eugenio Pérez <eperezma@redhat.com> wrote: > > Implement stop operation for vdpa_sim devices, so vhost-vdpa will offer > that backend feature and userspace can effectively stop the device. > > This is a must before get virtqueue indexes (base) for live migration, > since the device could modify them after userland gets them. There are > individual ways to perform that action for some devices > (VHOST_NET_SET_BACKEND, VHOST_VSOCK_SET_RUNNING, ...) but there was no > way to perform it for any vhost device (and, in particular, vhost-vdpa). > > After the return of ioctl with stop != 0, the device MUST finish any > pending operations like in flight requests. It must also preserve all > the necessary state (the virtqueue vring base plus the possible device > specific states) that is required for restoring in the future. The > device must not change its configuration after that point. > > After the return of ioctl with stop == 0, the device can continue > processing buffers as long as typical conditions are met (vq is enabled, > DRIVER_OK status bit is enabled, etc). > > In the future, we will provide features similar to > VHOST_USER_GET_INFLIGHT_FD so the device can save pending operations. > > Signed-off-by: Eugenio Pérez <eperezma@redhat.com> > --- > drivers/vdpa/vdpa_sim/vdpa_sim.c | 21 +++++++++++++++++++++ > drivers/vdpa/vdpa_sim/vdpa_sim.h | 1 + > drivers/vdpa/vdpa_sim/vdpa_sim_blk.c | 3 +++ > drivers/vdpa/vdpa_sim/vdpa_sim_net.c | 3 +++ > 4 files changed, 28 insertions(+) > > diff --git a/drivers/vdpa/vdpa_sim/vdpa_sim.c b/drivers/vdpa/vdpa_sim/vdpa_sim.c > index 50d721072beb..0515cf314bed 100644 > --- a/drivers/vdpa/vdpa_sim/vdpa_sim.c > +++ b/drivers/vdpa/vdpa_sim/vdpa_sim.c > @@ -107,6 +107,7 @@ static void vdpasim_do_reset(struct vdpasim *vdpasim) > for (i = 0; i < vdpasim->dev_attr.nas; i++) > vhost_iotlb_reset(&vdpasim->iommu[i]); > > + vdpasim->running = true; > spin_unlock(&vdpasim->iommu_lock); > > vdpasim->features = 0; > @@ -505,6 +506,24 @@ static int vdpasim_reset(struct vdpa_device *vdpa) > return 0; > } > > +static int vdpasim_stop(struct vdpa_device *vdpa, bool stop) > +{ > + struct vdpasim *vdpasim = vdpa_to_sim(vdpa); > + int i; > + > + spin_lock(&vdpasim->lock); > + vdpasim->running = !stop; > + if (vdpasim->running) { > + /* Check for missed buffers */ > + for (i = 0; i < vdpasim->dev_attr.nvqs; ++i) > + vdpasim_kick_vq(vdpa, i); > + > + } > + spin_unlock(&vdpasim->lock); > + > + return 0; > +} > + > static size_t vdpasim_get_config_size(struct vdpa_device *vdpa) > { > struct vdpasim *vdpasim = vdpa_to_sim(vdpa); > @@ -694,6 +713,7 @@ static const struct vdpa_config_ops vdpasim_config_ops = { > .get_status = vdpasim_get_status, > .set_status = vdpasim_set_status, > .reset = vdpasim_reset, > + .stop = vdpasim_stop, > .get_config_size = vdpasim_get_config_size, > .get_config = vdpasim_get_config, > .set_config = vdpasim_set_config, > @@ -726,6 +746,7 @@ static const struct vdpa_config_ops vdpasim_batch_config_ops = { > .get_status = vdpasim_get_status, > .set_status = vdpasim_set_status, > .reset = vdpasim_reset, > + .stop = vdpasim_stop, > .get_config_size = vdpasim_get_config_size, > .get_config = vdpasim_get_config, > .set_config = vdpasim_set_config, > diff --git a/drivers/vdpa/vdpa_sim/vdpa_sim.h b/drivers/vdpa/vdpa_sim/vdpa_sim.h > index 622782e92239..061986f30911 100644 > --- a/drivers/vdpa/vdpa_sim/vdpa_sim.h > +++ b/drivers/vdpa/vdpa_sim/vdpa_sim.h > @@ -66,6 +66,7 @@ struct vdpasim { > u32 generation; > u64 features; > u32 groups; > + bool running; > /* spinlock to synchronize iommu table */ > spinlock_t iommu_lock; > }; > diff --git a/drivers/vdpa/vdpa_sim/vdpa_sim_blk.c b/drivers/vdpa/vdpa_sim/vdpa_sim_blk.c > index 42d401d43911..bcdb1982c378 100644 > --- a/drivers/vdpa/vdpa_sim/vdpa_sim_blk.c > +++ b/drivers/vdpa/vdpa_sim/vdpa_sim_blk.c > @@ -204,6 +204,9 @@ static void vdpasim_blk_work(struct work_struct *work) > if (!(vdpasim->status & VIRTIO_CONFIG_S_DRIVER_OK)) > goto out; > > + if (!vdpasim->running) > + goto out; > + > for (i = 0; i < VDPASIM_BLK_VQ_NUM; i++) { > struct vdpasim_virtqueue *vq = &vdpasim->vqs[i]; > > diff --git a/drivers/vdpa/vdpa_sim/vdpa_sim_net.c b/drivers/vdpa/vdpa_sim/vdpa_sim_net.c > index 5125976a4df8..886449e88502 100644 > --- a/drivers/vdpa/vdpa_sim/vdpa_sim_net.c > +++ b/drivers/vdpa/vdpa_sim/vdpa_sim_net.c > @@ -154,6 +154,9 @@ static void vdpasim_net_work(struct work_struct *work) > > spin_lock(&vdpasim->lock); > > + if (!vdpasim->running) > + goto out; > + Do we need to check vdpasim->running in vdpasim_kick_vq()? Thanks > if (!(vdpasim->status & VIRTIO_CONFIG_S_DRIVER_OK)) > goto out; > > -- > 2.27.0 >
On Tue, May 24, 2022 at 07:06:10PM +0200, Eugenio Pérez wrote: >Implement stop operation for vdpa_sim devices, so vhost-vdpa will offer >that backend feature and userspace can effectively stop the device. > >This is a must before get virtqueue indexes (base) for live migration, >since the device could modify them after userland gets them. There are >individual ways to perform that action for some devices >(VHOST_NET_SET_BACKEND, VHOST_VSOCK_SET_RUNNING, ...) but there was no >way to perform it for any vhost device (and, in particular, vhost-vdpa). > >After the return of ioctl with stop != 0, the device MUST finish any >pending operations like in flight requests. It must also preserve all >the necessary state (the virtqueue vring base plus the possible device >specific states) that is required for restoring in the future. The >device must not change its configuration after that point. > >After the return of ioctl with stop == 0, the device can continue >processing buffers as long as typical conditions are met (vq is enabled, >DRIVER_OK status bit is enabled, etc). > >In the future, we will provide features similar to >VHOST_USER_GET_INFLIGHT_FD so the device can save pending operations. > >Signed-off-by: Eugenio Pérez <eperezma@redhat.com> >--- > drivers/vdpa/vdpa_sim/vdpa_sim.c | 21 +++++++++++++++++++++ > drivers/vdpa/vdpa_sim/vdpa_sim.h | 1 + > drivers/vdpa/vdpa_sim/vdpa_sim_blk.c | 3 +++ > drivers/vdpa/vdpa_sim/vdpa_sim_net.c | 3 +++ > 4 files changed, 28 insertions(+) > >diff --git a/drivers/vdpa/vdpa_sim/vdpa_sim.c b/drivers/vdpa/vdpa_sim/vdpa_sim.c >index 50d721072beb..0515cf314bed 100644 >--- a/drivers/vdpa/vdpa_sim/vdpa_sim.c >+++ b/drivers/vdpa/vdpa_sim/vdpa_sim.c >@@ -107,6 +107,7 @@ static void vdpasim_do_reset(struct vdpasim *vdpasim) > for (i = 0; i < vdpasim->dev_attr.nas; i++) > vhost_iotlb_reset(&vdpasim->iommu[i]); > >+ vdpasim->running = true; > spin_unlock(&vdpasim->iommu_lock); > > vdpasim->features = 0; >@@ -505,6 +506,24 @@ static int vdpasim_reset(struct vdpa_device *vdpa) > return 0; > } > >+static int vdpasim_stop(struct vdpa_device *vdpa, bool stop) >+{ >+ struct vdpasim *vdpasim = vdpa_to_sim(vdpa); >+ int i; >+ >+ spin_lock(&vdpasim->lock); >+ vdpasim->running = !stop; >+ if (vdpasim->running) { >+ /* Check for missed buffers */ >+ for (i = 0; i < vdpasim->dev_attr.nvqs; ++i) >+ vdpasim_kick_vq(vdpa, i); >+ >+ } >+ spin_unlock(&vdpasim->lock); >+ >+ return 0; >+} >+ > static size_t vdpasim_get_config_size(struct vdpa_device *vdpa) > { > struct vdpasim *vdpasim = vdpa_to_sim(vdpa); >@@ -694,6 +713,7 @@ static const struct vdpa_config_ops vdpasim_config_ops = { > .get_status = vdpasim_get_status, > .set_status = vdpasim_set_status, > .reset = vdpasim_reset, >+ .stop = vdpasim_stop, > .get_config_size = vdpasim_get_config_size, > .get_config = vdpasim_get_config, > .set_config = vdpasim_set_config, >@@ -726,6 +746,7 @@ static const struct vdpa_config_ops vdpasim_batch_config_ops = { > .get_status = vdpasim_get_status, > .set_status = vdpasim_set_status, > .reset = vdpasim_reset, >+ .stop = vdpasim_stop, > .get_config_size = vdpasim_get_config_size, > .get_config = vdpasim_get_config, > .set_config = vdpasim_set_config, >diff --git a/drivers/vdpa/vdpa_sim/vdpa_sim.h b/drivers/vdpa/vdpa_sim/vdpa_sim.h >index 622782e92239..061986f30911 100644 >--- a/drivers/vdpa/vdpa_sim/vdpa_sim.h >+++ b/drivers/vdpa/vdpa_sim/vdpa_sim.h >@@ -66,6 +66,7 @@ struct vdpasim { > u32 generation; > u64 features; > u32 groups; >+ bool running; > /* spinlock to synchronize iommu table */ > spinlock_t iommu_lock; > }; >diff --git a/drivers/vdpa/vdpa_sim/vdpa_sim_blk.c b/drivers/vdpa/vdpa_sim/vdpa_sim_blk.c >index 42d401d43911..bcdb1982c378 100644 >--- a/drivers/vdpa/vdpa_sim/vdpa_sim_blk.c >+++ b/drivers/vdpa/vdpa_sim/vdpa_sim_blk.c >@@ -204,6 +204,9 @@ static void vdpasim_blk_work(struct work_struct *work) > if (!(vdpasim->status & VIRTIO_CONFIG_S_DRIVER_OK)) > goto out; > >+ if (!vdpasim->running) >+ goto out; >+ Not related to this series, but I think in vdpa_sim_blk.c we should implement something similar to what we already do in vdpa_sim_net.c and re-schedule the work after X requests handled, otherwise we risk never stopping if there are always requests to handle. Also for supporting multiple queues, that could be a problem, but for now we only support one, so there should be no problem. I have other patches to send for vdpa_sim_blk.c, so if you want I can do that in my series. Thanks, Stefano > for (i = 0; i < VDPASIM_BLK_VQ_NUM; i++) { > struct vdpasim_virtqueue *vq = &vdpasim->vqs[i]; > >diff --git a/drivers/vdpa/vdpa_sim/vdpa_sim_net.c b/drivers/vdpa/vdpa_sim/vdpa_sim_net.c >index 5125976a4df8..886449e88502 100644 >--- a/drivers/vdpa/vdpa_sim/vdpa_sim_net.c >+++ b/drivers/vdpa/vdpa_sim/vdpa_sim_net.c >@@ -154,6 +154,9 @@ static void vdpasim_net_work(struct work_struct *work) > > spin_lock(&vdpasim->lock); > >+ if (!vdpasim->running) >+ goto out; >+ > if (!(vdpasim->status & VIRTIO_CONFIG_S_DRIVER_OK)) > goto out; > >-- >2.27.0 >
On Wed, May 25, 2022 at 4:54 AM Jason Wang <jasowang@redhat.com> wrote: > > On Wed, May 25, 2022 at 1:06 AM Eugenio Pérez <eperezma@redhat.com> wrote: > > > > Implement stop operation for vdpa_sim devices, so vhost-vdpa will offer > > that backend feature and userspace can effectively stop the device. > > > > This is a must before get virtqueue indexes (base) for live migration, > > since the device could modify them after userland gets them. There are > > individual ways to perform that action for some devices > > (VHOST_NET_SET_BACKEND, VHOST_VSOCK_SET_RUNNING, ...) but there was no > > way to perform it for any vhost device (and, in particular, vhost-vdpa). > > > > After the return of ioctl with stop != 0, the device MUST finish any > > pending operations like in flight requests. It must also preserve all > > the necessary state (the virtqueue vring base plus the possible device > > specific states) that is required for restoring in the future. The > > device must not change its configuration after that point. > > > > After the return of ioctl with stop == 0, the device can continue > > processing buffers as long as typical conditions are met (vq is enabled, > > DRIVER_OK status bit is enabled, etc). > > > > In the future, we will provide features similar to > > VHOST_USER_GET_INFLIGHT_FD so the device can save pending operations. > > > > Signed-off-by: Eugenio Pérez <eperezma@redhat.com> > > --- > > drivers/vdpa/vdpa_sim/vdpa_sim.c | 21 +++++++++++++++++++++ > > drivers/vdpa/vdpa_sim/vdpa_sim.h | 1 + > > drivers/vdpa/vdpa_sim/vdpa_sim_blk.c | 3 +++ > > drivers/vdpa/vdpa_sim/vdpa_sim_net.c | 3 +++ > > 4 files changed, 28 insertions(+) > > > > diff --git a/drivers/vdpa/vdpa_sim/vdpa_sim.c b/drivers/vdpa/vdpa_sim/vdpa_sim.c > > index 50d721072beb..0515cf314bed 100644 > > --- a/drivers/vdpa/vdpa_sim/vdpa_sim.c > > +++ b/drivers/vdpa/vdpa_sim/vdpa_sim.c > > @@ -107,6 +107,7 @@ static void vdpasim_do_reset(struct vdpasim *vdpasim) > > for (i = 0; i < vdpasim->dev_attr.nas; i++) > > vhost_iotlb_reset(&vdpasim->iommu[i]); > > > > + vdpasim->running = true; > > spin_unlock(&vdpasim->iommu_lock); > > > > vdpasim->features = 0; > > @@ -505,6 +506,24 @@ static int vdpasim_reset(struct vdpa_device *vdpa) > > return 0; > > } > > > > +static int vdpasim_stop(struct vdpa_device *vdpa, bool stop) > > +{ > > + struct vdpasim *vdpasim = vdpa_to_sim(vdpa); > > + int i; > > + > > + spin_lock(&vdpasim->lock); > > + vdpasim->running = !stop; > > + if (vdpasim->running) { > > + /* Check for missed buffers */ > > + for (i = 0; i < vdpasim->dev_attr.nvqs; ++i) > > + vdpasim_kick_vq(vdpa, i); > > + > > + } > > + spin_unlock(&vdpasim->lock); > > + > > + return 0; > > +} > > + > > static size_t vdpasim_get_config_size(struct vdpa_device *vdpa) > > { > > struct vdpasim *vdpasim = vdpa_to_sim(vdpa); > > @@ -694,6 +713,7 @@ static const struct vdpa_config_ops vdpasim_config_ops = { > > .get_status = vdpasim_get_status, > > .set_status = vdpasim_set_status, > > .reset = vdpasim_reset, > > + .stop = vdpasim_stop, > > .get_config_size = vdpasim_get_config_size, > > .get_config = vdpasim_get_config, > > .set_config = vdpasim_set_config, > > @@ -726,6 +746,7 @@ static const struct vdpa_config_ops vdpasim_batch_config_ops = { > > .get_status = vdpasim_get_status, > > .set_status = vdpasim_set_status, > > .reset = vdpasim_reset, > > + .stop = vdpasim_stop, > > .get_config_size = vdpasim_get_config_size, > > .get_config = vdpasim_get_config, > > .set_config = vdpasim_set_config, > > diff --git a/drivers/vdpa/vdpa_sim/vdpa_sim.h b/drivers/vdpa/vdpa_sim/vdpa_sim.h > > index 622782e92239..061986f30911 100644 > > --- a/drivers/vdpa/vdpa_sim/vdpa_sim.h > > +++ b/drivers/vdpa/vdpa_sim/vdpa_sim.h > > @@ -66,6 +66,7 @@ struct vdpasim { > > u32 generation; > > u64 features; > > u32 groups; > > + bool running; > > /* spinlock to synchronize iommu table */ > > spinlock_t iommu_lock; > > }; > > diff --git a/drivers/vdpa/vdpa_sim/vdpa_sim_blk.c b/drivers/vdpa/vdpa_sim/vdpa_sim_blk.c > > index 42d401d43911..bcdb1982c378 100644 > > --- a/drivers/vdpa/vdpa_sim/vdpa_sim_blk.c > > +++ b/drivers/vdpa/vdpa_sim/vdpa_sim_blk.c > > @@ -204,6 +204,9 @@ static void vdpasim_blk_work(struct work_struct *work) > > if (!(vdpasim->status & VIRTIO_CONFIG_S_DRIVER_OK)) > > goto out; > > > > + if (!vdpasim->running) > > + goto out; > > + > > for (i = 0; i < VDPASIM_BLK_VQ_NUM; i++) { > > struct vdpasim_virtqueue *vq = &vdpasim->vqs[i]; > > > > diff --git a/drivers/vdpa/vdpa_sim/vdpa_sim_net.c b/drivers/vdpa/vdpa_sim/vdpa_sim_net.c > > index 5125976a4df8..886449e88502 100644 > > --- a/drivers/vdpa/vdpa_sim/vdpa_sim_net.c > > +++ b/drivers/vdpa/vdpa_sim/vdpa_sim_net.c > > @@ -154,6 +154,9 @@ static void vdpasim_net_work(struct work_struct *work) > > > > spin_lock(&vdpasim->lock); > > > > + if (!vdpasim->running) > > + goto out; > > + > > Do we need to check vdpasim->running in vdpasim_kick_vq()? > I'd say that not really: The important part is that we don't process more buffers, and that is more accurate here. To check it here will always avoid it although we queue work. Maybe we can see it as an optimization: either to check before queuing the work as you propose or simply stop polling kick file descriptors? Thanks! > Thanks > > > if (!(vdpasim->status & VIRTIO_CONFIG_S_DRIVER_OK)) > > goto out; > > > > -- > > 2.27.0 > > >
diff --git a/drivers/vdpa/vdpa_sim/vdpa_sim.c b/drivers/vdpa/vdpa_sim/vdpa_sim.c index 50d721072beb..0515cf314bed 100644 --- a/drivers/vdpa/vdpa_sim/vdpa_sim.c +++ b/drivers/vdpa/vdpa_sim/vdpa_sim.c @@ -107,6 +107,7 @@ static void vdpasim_do_reset(struct vdpasim *vdpasim) for (i = 0; i < vdpasim->dev_attr.nas; i++) vhost_iotlb_reset(&vdpasim->iommu[i]); + vdpasim->running = true; spin_unlock(&vdpasim->iommu_lock); vdpasim->features = 0; @@ -505,6 +506,24 @@ static int vdpasim_reset(struct vdpa_device *vdpa) return 0; } +static int vdpasim_stop(struct vdpa_device *vdpa, bool stop) +{ + struct vdpasim *vdpasim = vdpa_to_sim(vdpa); + int i; + + spin_lock(&vdpasim->lock); + vdpasim->running = !stop; + if (vdpasim->running) { + /* Check for missed buffers */ + for (i = 0; i < vdpasim->dev_attr.nvqs; ++i) + vdpasim_kick_vq(vdpa, i); + + } + spin_unlock(&vdpasim->lock); + + return 0; +} + static size_t vdpasim_get_config_size(struct vdpa_device *vdpa) { struct vdpasim *vdpasim = vdpa_to_sim(vdpa); @@ -694,6 +713,7 @@ static const struct vdpa_config_ops vdpasim_config_ops = { .get_status = vdpasim_get_status, .set_status = vdpasim_set_status, .reset = vdpasim_reset, + .stop = vdpasim_stop, .get_config_size = vdpasim_get_config_size, .get_config = vdpasim_get_config, .set_config = vdpasim_set_config, @@ -726,6 +746,7 @@ static const struct vdpa_config_ops vdpasim_batch_config_ops = { .get_status = vdpasim_get_status, .set_status = vdpasim_set_status, .reset = vdpasim_reset, + .stop = vdpasim_stop, .get_config_size = vdpasim_get_config_size, .get_config = vdpasim_get_config, .set_config = vdpasim_set_config, diff --git a/drivers/vdpa/vdpa_sim/vdpa_sim.h b/drivers/vdpa/vdpa_sim/vdpa_sim.h index 622782e92239..061986f30911 100644 --- a/drivers/vdpa/vdpa_sim/vdpa_sim.h +++ b/drivers/vdpa/vdpa_sim/vdpa_sim.h @@ -66,6 +66,7 @@ struct vdpasim { u32 generation; u64 features; u32 groups; + bool running; /* spinlock to synchronize iommu table */ spinlock_t iommu_lock; }; diff --git a/drivers/vdpa/vdpa_sim/vdpa_sim_blk.c b/drivers/vdpa/vdpa_sim/vdpa_sim_blk.c index 42d401d43911..bcdb1982c378 100644 --- a/drivers/vdpa/vdpa_sim/vdpa_sim_blk.c +++ b/drivers/vdpa/vdpa_sim/vdpa_sim_blk.c @@ -204,6 +204,9 @@ static void vdpasim_blk_work(struct work_struct *work) if (!(vdpasim->status & VIRTIO_CONFIG_S_DRIVER_OK)) goto out; + if (!vdpasim->running) + goto out; + for (i = 0; i < VDPASIM_BLK_VQ_NUM; i++) { struct vdpasim_virtqueue *vq = &vdpasim->vqs[i]; diff --git a/drivers/vdpa/vdpa_sim/vdpa_sim_net.c b/drivers/vdpa/vdpa_sim/vdpa_sim_net.c index 5125976a4df8..886449e88502 100644 --- a/drivers/vdpa/vdpa_sim/vdpa_sim_net.c +++ b/drivers/vdpa/vdpa_sim/vdpa_sim_net.c @@ -154,6 +154,9 @@ static void vdpasim_net_work(struct work_struct *work) spin_lock(&vdpasim->lock); + if (!vdpasim->running) + goto out; + if (!(vdpasim->status & VIRTIO_CONFIG_S_DRIVER_OK)) goto out;
Implement stop operation for vdpa_sim devices, so vhost-vdpa will offer that backend feature and userspace can effectively stop the device. This is a must before get virtqueue indexes (base) for live migration, since the device could modify them after userland gets them. There are individual ways to perform that action for some devices (VHOST_NET_SET_BACKEND, VHOST_VSOCK_SET_RUNNING, ...) but there was no way to perform it for any vhost device (and, in particular, vhost-vdpa). After the return of ioctl with stop != 0, the device MUST finish any pending operations like in flight requests. It must also preserve all the necessary state (the virtqueue vring base plus the possible device specific states) that is required for restoring in the future. The device must not change its configuration after that point. After the return of ioctl with stop == 0, the device can continue processing buffers as long as typical conditions are met (vq is enabled, DRIVER_OK status bit is enabled, etc). In the future, we will provide features similar to VHOST_USER_GET_INFLIGHT_FD so the device can save pending operations. Signed-off-by: Eugenio Pérez <eperezma@redhat.com> --- drivers/vdpa/vdpa_sim/vdpa_sim.c | 21 +++++++++++++++++++++ drivers/vdpa/vdpa_sim/vdpa_sim.h | 1 + drivers/vdpa/vdpa_sim/vdpa_sim_blk.c | 3 +++ drivers/vdpa/vdpa_sim/vdpa_sim_net.c | 3 +++ 4 files changed, 28 insertions(+)