RE: [PATCH V10 06/11] iommu/vt-d: Add bind guest PASID support

["Tian, Kevin" <kevin.tian@intel.com>]

> From: Jacob Pan <jacob.jun.pan@linux.intel.com>
> Sent: Tuesday, March 31, 2020 4:52 AM
> 
> On Sat, 28 Mar 2020 08:02:01 +0000
> "Tian, Kevin" <kevin.tian@intel.com> wrote:
> 
> > > From: Jacob Pan <jacob.jun.pan@linux.intel.com>
> > > Sent: Saturday, March 21, 2020 7:28 AM
> > >
> > > When supporting guest SVA with emulated IOMMU, the guest PASID
> > > table is shadowed in VMM. Updates to guest vIOMMU PASID table
> > > will result in PASID cache flush which will be passed down to
> > > the host as bind guest PASID calls.
> > >
> > > For the SL page tables, it will be harvested from device's
> > > default domain (request w/o PASID), or aux domain in case of
> > > mediated device.
> > >
> > >     .-------------.  .---------------------------.
> > >     |   vIOMMU    |  | Guest process CR3, FL only|
> > >     |             |  '---------------------------'
> > >     .----------------/
> > >     | PASID Entry |--- PASID cache flush -
> > >     '-------------'                       |
> > >     |             |                       V
> > >     |             |                CR3 in GPA
> > >     '-------------'
> > > Guest
> > > ------| Shadow |--------------------------|--------
> > >       v        v                          v
> > > Host
> > >     .-------------.  .----------------------.
> > >     |   pIOMMU    |  | Bind FL for GVA-GPA  |
> > >     |             |  '----------------------'
> > >     .----------------/  |
> > >     | PASID Entry |     V (Nested xlate)
> > >     '----------------\.------------------------------.
> > >     |             |   |SL for GPA-HPA, default domain|
> > >     |             |   '------------------------------'
> > >     '-------------'
> > > Where:
> > >  - FL = First level/stage one page tables
> > >  - SL = Second level/stage two page tables
> > >
> > > Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
> > > Signed-off-by: Liu, Yi L <yi.l.liu@linux.intel.com>
> > > ---
> > >  drivers/iommu/intel-iommu.c |   4 +
> > >  drivers/iommu/intel-svm.c   | 224
> > > ++++++++++++++++++++++++++++++++++++++++++++
> > >  include/linux/intel-iommu.h |   8 +-
> > >  include/linux/intel-svm.h   |  17 ++++
> > >  4 files changed, 252 insertions(+), 1 deletion(-)
> > >
> > > diff --git a/drivers/iommu/intel-iommu.c
> > > b/drivers/iommu/intel-iommu.c index e599b2537b1c..b1477cd423dd
> > > 100644 --- a/drivers/iommu/intel-iommu.c
> > > +++ b/drivers/iommu/intel-iommu.c
> > > @@ -6203,6 +6203,10 @@ const struct iommu_ops intel_iommu_ops = {
> > >  	.dev_disable_feat	= intel_iommu_dev_disable_feat,
> > >  	.is_attach_deferred	=
> > > intel_iommu_is_attach_deferred, .pgsize_bitmap		=
> > > INTEL_IOMMU_PGSIZES, +#ifdef CONFIG_INTEL_IOMMU_SVM
> > > +	.sva_bind_gpasid	= intel_svm_bind_gpasid,
> > > +	.sva_unbind_gpasid	= intel_svm_unbind_gpasid,
> > > +#endif
> > >  };
> > >
> > >  static void quirk_iommu_igfx(struct pci_dev *dev)
> > > diff --git a/drivers/iommu/intel-svm.c b/drivers/iommu/intel-svm.c
> > > index d7f2a5358900..47c0deb5ae56 100644
> > > --- a/drivers/iommu/intel-svm.c
> > > +++ b/drivers/iommu/intel-svm.c
> > > @@ -226,6 +226,230 @@ static LIST_HEAD(global_svm_list);
> > >  	list_for_each_entry((sdev), &(svm)->devs, list)	\
> > >  		if ((d) != (sdev)->dev) {} else
> > >
> > > +int intel_svm_bind_gpasid(struct iommu_domain *domain,
> > > +			struct device *dev,
> > > +			struct iommu_gpasid_bind_data *data)
> > > +{
> > > +	struct intel_iommu *iommu = intel_svm_device_to_iommu(dev);
> > > +	struct dmar_domain *ddomain;
> >
> > what about the full name e.g. dmar_domain? though a bit longer
> > but clearer than ddomain.
> >
> Sure, I don't have preference.
> 
> > > +	struct intel_svm_dev *sdev;
> > > +	struct intel_svm *svm;
> > > +	int ret = 0;
> > > +
> > > +	if (WARN_ON(!iommu) || !data)
> > > +		return -EINVAL;
> > > +
> > > +	if (data->version != IOMMU_GPASID_BIND_VERSION_1 ||
> > > +	    data->format != IOMMU_PASID_FORMAT_INTEL_VTD)
> > > +		return -EINVAL;
> > > +
> > > +	if (dev_is_pci(dev)) {
> > > +		/* VT-d supports devices with full 20 bit PASIDs
> > > only */
> > > +		if (pci_max_pasids(to_pci_dev(dev)) != PASID_MAX)
> > > +			return -EINVAL;
> > > +	} else {
> > > +		return -ENOTSUPP;
> > > +	}
> > > +
> > > +	/*
> > > +	 * We only check host PASID range, we have no knowledge to
> > > check
> > > +	 * guest PASID range nor do we use the guest PASID.
> > > +	 */
> > > +	if (data->hpasid <= 0 || data->hpasid >= PASID_MAX)
> > > +		return -EINVAL;
> > > +
> > > +	ddomain = to_dmar_domain(domain);
> > > +
> > > +	/* Sanity check paging mode support match between host and
> > > guest */
> > > +	if (data->addr_width == ADDR_WIDTH_5LEVEL &&
> > > +	    !cap_5lp_support(iommu->cap)) {
> > > +		pr_err("Cannot support 5 level paging requested by
> > > guest!\n");
> > > +		return -EINVAL;
> > > +	}
> >
> > -ENOTSUPP?
> I was thinking from this API p.o.v, the input is invalid. Since both
> cap and addr_width are derived from input arguments.

ok, suppose the userspace already enumerates the capabilities before
making this call.

> 
> >
> > > +
> > > +	mutex_lock(&pasid_mutex);
> > > +	svm = ioasid_find(NULL, data->hpasid, NULL);
> > > +	if (IS_ERR(svm)) {
> > > +		ret = PTR_ERR(svm);
> > > +		goto out;
> > > +	}
> > > +
> > > +	if (svm) {
> > > +		/*
> > > +		 * If we found svm for the PASID, there must be at
> > > +		 * least one device bond, otherwise svm should be
> > > freed.
> > > +		 */
> > > +		if (WARN_ON(list_empty(&svm->devs))) {
> > > +			ret = -EINVAL;
> > > +			goto out;
> > > +		}
> > > +
> > > +		if (svm->mm == get_task_mm(current) &&
> > > +		    data->hpasid == svm->pasid &&
> > > +		    data->gpasid == svm->gpasid) {
> > > +			pr_warn("Cannot bind the same guest-host
> > > PASID for the same process\n");
> >
> > Sorry I didn’t get the rationale here. Isn't this branch is for
> > binding the same PASID to multiple devices? In that case definitely
> > it is binding the same guest-host PASID for the same process.
> > otherwise if hpasid is different then you'll hit a different
> > intel_svm, while if gpasid is different how you can use one intel_svm
> > to hold multiple gpasids?
> >
> > I feel the error condition should be the opposite. and suppose
> > SVM_FLAG_ GUEST_PASID should be verified before checking gpasid.
> >
> You are right, actually we don't need the check here. The
> scenario for multiple devices bind to the same PASID is checked in
> for_each_svm_dev()
> I will remove this code.
> 
> > > +			mmput(svm->mm);
> > > +			ret = -EINVAL;
> > > +			goto out;
> > > +		}
> > > +		mmput(current->mm);
> > > +
> > > +		for_each_svm_dev(sdev, svm, dev) {
> > > +			/* In case of multiple sub-devices of the
> > > same pdev
> > > +			 * assigned, we should allow multiple bind
> > > calls with
> > > +			 * the same PASID and pdev.
> >
> > Does sub-device mean mdev? I didn't find such notation in current
> > iommu directory.
> >
> yes it is intended for mdev.
> > and to make it clearer, "In case of multiple mdevs of the same pdev
> > assigned to the same guest process".
> >
> I am avoiding mdev on purpose since it is not a concept in iommu
> driver. sub-device is more generic.

ok, fine to me.

> 
> > > +			 */
> > > +			sdev->users++;
> > > +			goto out;
> > > +		}
> > > +	} else {
> > > +		/* We come here when PASID has never been bond to a
> > > device. */
> > > +		svm = kzalloc(sizeof(*svm), GFP_KERNEL);
> > > +		if (!svm) {
> > > +			ret = -ENOMEM;
> > > +			goto out;
> > > +		}
> > > +		/* REVISIT: upper layer/VFIO can track host
> > > process that bind the PASID.
> > > +		 * ioasid_set = mm might be sufficient for vfio to
> > > check pasid VMM
> > > +		 * ownership.
> > > +		 */
> >
> > Above message is unclear about what should be revisited. Does it
> > describe the current implementation or the expected revision in the
> > future?
> >
> What I meant was if VFIO can check PASID-mm ownership by itself, then
> we don;t have to store svm->mm here. Will drop the line below.
> I will add this comment to clarify.
> 
> > > +		svm->mm = get_task_mm(current);
> > > +		svm->pasid = data->hpasid;
> > > +		if (data->flags & IOMMU_SVA_GPASID_VAL) {
> > > +			svm->gpasid = data->gpasid;
> > > +			svm->flags |= SVM_FLAG_GUEST_PASID;
> > > +		}
> > > +		ioasid_set_data(data->hpasid, svm);
> > > +		INIT_LIST_HEAD_RCU(&svm->devs);
> > > +		mmput(svm->mm);
> > > +	}
> > > +	sdev = kzalloc(sizeof(*sdev), GFP_KERNEL);
> > > +	if (!sdev) {
> > > +		if (list_empty(&svm->devs)) {
> > > +			ioasid_set_data(data->hpasid, NULL);
> > > +			kfree(svm);
> > > +		}
> > > +		ret = -ENOMEM;
> > > +		goto out;
> > > +	}
> > > +	sdev->dev = dev;
> > > +	sdev->users = 1;
> > > +
> > > +	/* Set up device context entry for PASID if not enabled
> > > already */
> > > +	ret = intel_iommu_enable_pasid(iommu, sdev->dev);
> > > +	if (ret) {
> > > +		dev_err(dev, "Failed to enable PASID
> > > capability\n");
> > > +		kfree(sdev);
> > > +		/*
> > > +		 * If this this a new PASID that never bond to a
> > > device, then
> > > +		 * the device list must be empty which indicates
> > > struct svm
> > > +		 * was allocated in this function.
> > > +		 */
> >
> > the comment better move to the 1st occurrence when sdev allocation
> > fails. or even better put it in out label...
> >
> Sounds good.
> 
> > > +		if (list_empty(&svm->devs)) {
> > > +			ioasid_set_data(data->hpasid, NULL);
> > > +			kfree(svm);
> > > +		}
> > > +		goto out;
> > > +	}
> > > +
> > > +	/*
> > > +	 * For guest bind, we need to set up PASID table entry as
> > > follows:
> > > +	 * - FLPM matches guest paging mode
> > > +	 * - turn on nested mode
> > > +	 * - SL guest address width matching
> > > +	 */
> >
> > looks above just explains the internal detail of
> > intel_pasid_setup_nested, which is not necessary to be here.
> >
> Right, will remove the comments.
> 
> > > +	ret = intel_pasid_setup_nested(iommu,
> > > +				       dev,
> > > +				       (pgd_t *)data->gpgd,
> > > +				       data->hpasid,
> > > +				       &data->vtd,
> > > +				       ddomain,
> > > +				       data->addr_width);
> >
> > It's worthy of an explanation here that setup_nested is required for
> > every device (even when they are sharing same intel_svm) because
> > we allocate pasid table per device. Otherwise I made a mistake to
> > think that only the 1st device bound to a new hpasid requires this
> > step. 😊
> >
> Good suggestion, I will add the comments as:
> /*
>  * PASID table is per device for better security. Therefore, for
>  * each bind of a new device even with an existing PASID, we need to
>  * call the nested mode setup function here.
>  */
> 
> > > +	if (ret) {
> > > +		dev_err(dev, "Failed to set up PASID %llu in
> > > nested mode, Err %d\n",
> > > +			data->hpasid, ret);
> > > +		/*
> > > +		 * PASID entry should be in cleared state if
> > > nested mode
> > > +		 * set up failed. So we only need to clear IOASID
> > > tracking
> > > +		 * data such that free call will succeed.
> > > +		 */
> > > +		kfree(sdev);
> > > +		if (list_empty(&svm->devs)) {
> > > +			ioasid_set_data(data->hpasid, NULL);
> > > +			kfree(svm);
> > > +		}
> > > +		goto out;
> > > +	}
> > > +	svm->flags |= SVM_FLAG_GUEST_MODE;
> > > +
> > > +	init_rcu_head(&sdev->rcu);
> > > +	list_add_rcu(&sdev->list, &svm->devs);
> > > + out:
> > > +	mutex_unlock(&pasid_mutex);
> > > +	return ret;
> > > +}
> > > +
> > > +int intel_svm_unbind_gpasid(struct device *dev, int pasid)
> > > +{
> > > +	struct intel_iommu *iommu = intel_svm_device_to_iommu(dev);
> > > +	struct intel_svm_dev *sdev;
> > > +	struct intel_svm *svm;
> > > +	int ret = -EINVAL;
> > > +
> > > +	if (WARN_ON(!iommu))
> > > +		return -EINVAL;
> > > +
> > > +	mutex_lock(&pasid_mutex);
> > > +	svm = ioasid_find(NULL, pasid, NULL);
> > > +	if (!svm) {
> > > +		ret = -EINVAL;
> > > +		goto out;
> > > +	}
> > > +
> > > +	if (IS_ERR(svm)) {
> > > +		ret = PTR_ERR(svm);
> > > +		goto out;
> > > +	}
> > > +
> > > +	for_each_svm_dev(sdev, svm, dev) {
> > > +		ret = 0;
> > > +		sdev->users--;
> > > +		if (!sdev->users) {
> > > +			list_del_rcu(&sdev->list);
> > > +			intel_pasid_tear_down_entry(iommu, dev,
> > > svm-
> > > >pasid);
> > > +			/* TODO: Drain in flight PRQ for the PASID
> > > since it
> > > +			 * may get reused soon, we don't want to
> > > +			 * confuse with its previous life.
> > > +			 * intel_svm_drain_prq(dev, pasid);
> > > +			 */
> > > +			kfree_rcu(sdev, rcu);
> > > +
> > > +			if (list_empty(&svm->devs)) {
> > > +				/*
> > > +				 * We do not free PASID here until
> > > explicit call
> > > +				 * from VFIO to free. The PASID
> > > life cycle
> > > +				 * management is largely tied to
> > > VFIO management
> > > +				 * of assigned device life cycles.
> > > In case of
> > > +				 * guest exit without a explicit
> > > free PASID call,
> > > +				 * the responsibility lies in VFIO
> > > layer to free
> > > +				 * the PASIDs allocated for the
> > > guest.
> > > +				 * For security reasons, VFIO has
> > > to track the
> > > +				 * PASID ownership per guest
> > > anyway to ensure
> > > +				 * that PASID allocated by one
> > > guest cannot be
> > > +				 * used by another.
> >
> > As commented in other patches, VFIO is only one example user of this
> > API...
> >
> Right, how about this:
> 	/*
> 	 * We do not free the IOASID here in that
> 	 * IOMMU driver did not allocate it.
> 	 * Unlike native SVM, IOASID for guest use was
> 	 * allocated prior to the bind call.
> 	 * In any case, if the free call comes before
> 	 * the unbind, IOMMU driver will get notified
> 	 * and perform cleanup.
> 	 */

looks good.

> 
> > > +				 */
> > > +				ioasid_set_data(pasid, NULL);
> > > +				kfree(svm);
> > > +			}
> > > +		}
> > > +		break;
> > > +	}
> >
> > what about no dev match? an -EINVAL is also required then.
> >
> Yes, ret is initialized as -EINVAL
> 
> > > +out:
> > > +	mutex_unlock(&pasid_mutex);
> > > +
> > > +	return ret;
> > > +}
> > > +
> > >  int intel_svm_bind_mm(struct device *dev, int *pasid, int flags,
> > > struct svm_dev_ops *ops)
> > >  {
> > >  	struct intel_iommu *iommu = intel_svm_device_to_iommu(dev);
> > > diff --git a/include/linux/intel-iommu.h
> > > b/include/linux/intel-iommu.h index eda1d6687144..85b05120940e
> > > 100644 --- a/include/linux/intel-iommu.h
> > > +++ b/include/linux/intel-iommu.h
> > > @@ -681,7 +681,9 @@ struct dmar_domain *find_domain(struct device
> > > *dev);
> > >  extern void intel_svm_check(struct intel_iommu *iommu);
> > >  extern int intel_svm_enable_prq(struct intel_iommu *iommu);
> > >  extern int intel_svm_finish_prq(struct intel_iommu *iommu);
> > > -
> > > +extern int intel_svm_bind_gpasid(struct iommu_domain *domain,
> > > +		struct device *dev, struct iommu_gpasid_bind_data
> > > *data); +extern int intel_svm_unbind_gpasid(struct device *dev, int
> > > pasid); struct svm_dev_ops;
> > >
> > >  struct intel_svm_dev {
> > > @@ -698,9 +700,13 @@ struct intel_svm_dev {
> > >  struct intel_svm {
> > >  	struct mmu_notifier notifier;
> > >  	struct mm_struct *mm;
> > > +
> > >  	struct intel_iommu *iommu;
> > >  	int flags;
> > >  	int pasid;
> > > +	int gpasid; /* Guest PASID in case of vSVA bind with
> > > non-identity host
> > > +		     * to guest PASID mapping.
> > > +		     */
> >
> > we don't need to highlight identity or non-identity thing, since
> > either way shares the same infrastructure here and it is not the
> > knowledge that the kernel driver should assume
> >
> Sorry, I don't get your point.
> 
> What I meant was that this field "gpasid" is only used for non-identity
> case. For identity case, we don't have SVM_FLAG_GUEST_PASID.

what's the problem if a guest tries to set gpasid even in identity
case? do you want to add check to reject it? Also I remember we
discussed before that we want to provide a consistent interface 
to other consumer e.g. KVM to setup VMCS PASID translation table.
In that case, regardless of identity or non-identity, we need provide
such mapping info.

> 
> > >  	struct list_head devs;
> > >  	struct list_head list;
> > >  };
> > > diff --git a/include/linux/intel-svm.h b/include/linux/intel-svm.h
> > > index d7c403d0dd27..c19690937540 100644
> > > --- a/include/linux/intel-svm.h
> > > +++ b/include/linux/intel-svm.h
> > > @@ -44,6 +44,23 @@ struct svm_dev_ops {
> > >   * do such IOTLB flushes automatically.
> > >   */
> > >  #define SVM_FLAG_SUPERVISOR_MODE	(1<<1)
> > > +/*
> > > + * The SVM_FLAG_GUEST_MODE flag is used when a guest process bind
> > > to a device.
> > > + * In this case the mm_struct is in the guest kernel or userspace,
> > > its life
> > > + * cycle is managed by VMM and VFIO layer. For IOMMU driver, this
> > > API provides
> > > + * means to bind/unbind guest CR3 with PASIDs allocated for a
> > > device.
> > > + */
> > > +#define SVM_FLAG_GUEST_MODE	(1<<2)
> > > +/*
> > > + * The SVM_FLAG_GUEST_PASID flag is used when a guest has its own
> > > PASID space,
> > > + * which requires guest and host PASID translation at both
> > > directions. We keep
> > > + * track of guest PASID in order to provide lookup service to
> > > device drivers.
> > > + * One such example is a physical function (PF) driver that
> > > supports mediated
> > > + * device (mdev) assignment. Guest programming of mdev
> > > configuration space can
> > > + * only be done with guest PASID, therefore PF driver needs to
> > > find the matching
> > > + * host PASID to program the real hardware.
> > > + */
> > > +#define SVM_FLAG_GUEST_PASID	(1<<3)
> > >
> > >  #ifdef CONFIG_INTEL_IOMMU_SVM
> > >
> > > --
> > > 2.7.4
> >
> 
> [Jacob Pan]