| Message ID | 20230321154228.182769-3-sgarzare@redhat.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | vdpa_sim: add support for user VA | expand |
On Tue, Mar 21, 2023 at 11:42 PM Stefano Garzarella <sgarzare@redhat.com> wrote: > > When the user call VHOST_SET_OWNER ioctl and the vDPA device > has `use_va` set to true, let's call the bind_mm callback. > In this way we can bind the device to the user address space > and directly use the user VA. > > The unbind_mm callback is called during the release after > stopping the device. > > Signed-off-by: Stefano Garzarella <sgarzare@redhat.com> > --- > > Notes: > v3: > - added `case VHOST_SET_OWNER` in vhost_vdpa_unlocked_ioctl() [Jason] > v2: > - call the new unbind_mm callback during the release [Jason] > - avoid to call bind_mm callback after the reset, since the device > is not detaching it now during the reset > > drivers/vhost/vdpa.c | 31 +++++++++++++++++++++++++++++++ > 1 file changed, 31 insertions(+) > > diff --git a/drivers/vhost/vdpa.c b/drivers/vhost/vdpa.c > index 7be9d9d8f01c..20250c3418b2 100644 > --- a/drivers/vhost/vdpa.c > +++ b/drivers/vhost/vdpa.c > @@ -219,6 +219,28 @@ static int vhost_vdpa_reset(struct vhost_vdpa *v) > return vdpa_reset(vdpa); > } > > +static long vhost_vdpa_bind_mm(struct vhost_vdpa *v) > +{ > + struct vdpa_device *vdpa = v->vdpa; > + const struct vdpa_config_ops *ops = vdpa->config; > + > + if (!vdpa->use_va || !ops->bind_mm) > + return 0; > + > + return ops->bind_mm(vdpa, v->vdev.mm); > +} > + > +static void vhost_vdpa_unbind_mm(struct vhost_vdpa *v) > +{ > + struct vdpa_device *vdpa = v->vdpa; > + const struct vdpa_config_ops *ops = vdpa->config; > + > + if (!vdpa->use_va || !ops->unbind_mm) > + return; > + > + ops->unbind_mm(vdpa); > +} > + > static long vhost_vdpa_get_device_id(struct vhost_vdpa *v, u8 __user *argp) > { > struct vdpa_device *vdpa = v->vdpa; > @@ -709,6 +731,14 @@ static long vhost_vdpa_unlocked_ioctl(struct file *filep, > case VHOST_VDPA_RESUME: > r = vhost_vdpa_resume(v); > break; > + case VHOST_SET_OWNER: > + r = vhost_dev_set_owner(d); Nit: I'd stick to the current way of passing the cmd, argp to vhost_dev_ioctl() and introduce a new switch after the vhost_dev_ioctl(). In this way, we are immune to any possible changes of dealing with VHOST_SET_OWNER in vhost core. Others look good. Thanks > + if (r) > + break; > + r = vhost_vdpa_bind_mm(v); > + if (r) > + vhost_dev_reset_owner(d, NULL); > + break; > default: > r = vhost_dev_ioctl(&v->vdev, cmd, argp); > if (r == -ENOIOCTLCMD) > @@ -1287,6 +1317,7 @@ static int vhost_vdpa_release(struct inode *inode, struct file *filep) > vhost_vdpa_clean_irq(v); > vhost_vdpa_reset(v); > vhost_dev_stop(&v->vdev); > + vhost_vdpa_unbind_mm(v); > vhost_vdpa_config_put(v); > vhost_vdpa_cleanup(v); > mutex_unlock(&d->mutex); > -- > 2.39.2 >
On Thu, Mar 23, 2023 at 11:01:39AM +0800, Jason Wang wrote: >On Tue, Mar 21, 2023 at 11:42 PM Stefano Garzarella <sgarzare@redhat.com> wrote: >> >> When the user call VHOST_SET_OWNER ioctl and the vDPA device >> has `use_va` set to true, let's call the bind_mm callback. >> In this way we can bind the device to the user address space >> and directly use the user VA. >> >> The unbind_mm callback is called during the release after >> stopping the device. >> >> Signed-off-by: Stefano Garzarella <sgarzare@redhat.com> >> --- >> >> Notes: >> v3: >> - added `case VHOST_SET_OWNER` in vhost_vdpa_unlocked_ioctl() [Jason] >> v2: >> - call the new unbind_mm callback during the release [Jason] >> - avoid to call bind_mm callback after the reset, since the device >> is not detaching it now during the reset >> >> drivers/vhost/vdpa.c | 31 +++++++++++++++++++++++++++++++ >> 1 file changed, 31 insertions(+) >> >> diff --git a/drivers/vhost/vdpa.c b/drivers/vhost/vdpa.c >> index 7be9d9d8f01c..20250c3418b2 100644 >> --- a/drivers/vhost/vdpa.c >> +++ b/drivers/vhost/vdpa.c >> @@ -219,6 +219,28 @@ static int vhost_vdpa_reset(struct vhost_vdpa *v) >> return vdpa_reset(vdpa); >> } >> >> +static long vhost_vdpa_bind_mm(struct vhost_vdpa *v) >> +{ >> + struct vdpa_device *vdpa = v->vdpa; >> + const struct vdpa_config_ops *ops = vdpa->config; >> + >> + if (!vdpa->use_va || !ops->bind_mm) >> + return 0; >> + >> + return ops->bind_mm(vdpa, v->vdev.mm); >> +} >> + >> +static void vhost_vdpa_unbind_mm(struct vhost_vdpa *v) >> +{ >> + struct vdpa_device *vdpa = v->vdpa; >> + const struct vdpa_config_ops *ops = vdpa->config; >> + >> + if (!vdpa->use_va || !ops->unbind_mm) >> + return; >> + >> + ops->unbind_mm(vdpa); >> +} >> + >> static long vhost_vdpa_get_device_id(struct vhost_vdpa *v, u8 __user *argp) >> { >> struct vdpa_device *vdpa = v->vdpa; >> @@ -709,6 +731,14 @@ static long vhost_vdpa_unlocked_ioctl(struct file *filep, >> case VHOST_VDPA_RESUME: >> r = vhost_vdpa_resume(v); >> break; >> + case VHOST_SET_OWNER: >> + r = vhost_dev_set_owner(d); > >Nit: > >I'd stick to the current way of passing the cmd, argp to >vhost_dev_ioctl() and introduce a new switch after the >vhost_dev_ioctl(). > >In this way, we are immune to any possible changes of dealing with >VHOST_SET_OWNER in vhost core. Good point, I'll change in v4. > >Others look good. Thanks, Stefano
diff --git a/drivers/vhost/vdpa.c b/drivers/vhost/vdpa.c index 7be9d9d8f01c..20250c3418b2 100644 --- a/drivers/vhost/vdpa.c +++ b/drivers/vhost/vdpa.c @@ -219,6 +219,28 @@ static int vhost_vdpa_reset(struct vhost_vdpa *v) return vdpa_reset(vdpa); } +static long vhost_vdpa_bind_mm(struct vhost_vdpa *v) +{ + struct vdpa_device *vdpa = v->vdpa; + const struct vdpa_config_ops *ops = vdpa->config; + + if (!vdpa->use_va || !ops->bind_mm) + return 0; + + return ops->bind_mm(vdpa, v->vdev.mm); +} + +static void vhost_vdpa_unbind_mm(struct vhost_vdpa *v) +{ + struct vdpa_device *vdpa = v->vdpa; + const struct vdpa_config_ops *ops = vdpa->config; + + if (!vdpa->use_va || !ops->unbind_mm) + return; + + ops->unbind_mm(vdpa); +} + static long vhost_vdpa_get_device_id(struct vhost_vdpa *v, u8 __user *argp) { struct vdpa_device *vdpa = v->vdpa; @@ -709,6 +731,14 @@ static long vhost_vdpa_unlocked_ioctl(struct file *filep, case VHOST_VDPA_RESUME: r = vhost_vdpa_resume(v); break; + case VHOST_SET_OWNER: + r = vhost_dev_set_owner(d); + if (r) + break; + r = vhost_vdpa_bind_mm(v); + if (r) + vhost_dev_reset_owner(d, NULL); + break; default: r = vhost_dev_ioctl(&v->vdev, cmd, argp); if (r == -ENOIOCTLCMD) @@ -1287,6 +1317,7 @@ static int vhost_vdpa_release(struct inode *inode, struct file *filep) vhost_vdpa_clean_irq(v); vhost_vdpa_reset(v); vhost_dev_stop(&v->vdev); + vhost_vdpa_unbind_mm(v); vhost_vdpa_config_put(v); vhost_vdpa_cleanup(v); mutex_unlock(&d->mutex);
When the user call VHOST_SET_OWNER ioctl and the vDPA device has `use_va` set to true, let's call the bind_mm callback. In this way we can bind the device to the user address space and directly use the user VA. The unbind_mm callback is called during the release after stopping the device. Signed-off-by: Stefano Garzarella <sgarzare@redhat.com> --- Notes: v3: - added `case VHOST_SET_OWNER` in vhost_vdpa_unlocked_ioctl() [Jason] v2: - call the new unbind_mm callback during the release [Jason] - avoid to call bind_mm callback after the reset, since the device is not detaching it now during the reset drivers/vhost/vdpa.c | 31 +++++++++++++++++++++++++++++++ 1 file changed, 31 insertions(+)