Re: [PATCH v2 1/3] docs: IOMMU user API

From: Jacob Pan
Date: Fri Jun 12 2020 - 18:47:15 EST


Hi Jon,

On Thu, 11 Jun 2020 10:30:32 +0100
Jonathan Cameron <Jonathan.Cameron@xxxxxxxxxx> wrote:

> On Wed, 10 Jun 2020 21:12:13 -0700
> Jacob Pan <jacob.jun.pan@xxxxxxxxxxxxxxx> wrote:
>
> > IOMMU UAPI is newly introduced to support communications between
> > guest virtual IOMMU and host IOMMU. There has been lots of
> > discussions on how it should work with VFIO UAPI and userspace in
> > general.
> >
> > This document is indended to clarify the UAPI design and usage. The
> > mechenics of how future extensions should be achieved are also
> > covered
>
> mechanics
>
will fix,

> > in this documentation.
> >
> > Signed-off-by: Liu Yi L <yi.l.liu@xxxxxxxxx>
> > Signed-off-by: Jacob Pan <jacob.jun.pan@xxxxxxxxxxxxxxx>
> Mostly seems sensible. A few comments / queries inline.
>
> Jonathan
>
> > ---
> > Documentation/userspace-api/iommu.rst | 210
> > ++++++++++++++++++++++++++++++++++ 1 file changed, 210 insertions(+)
> > create mode 100644 Documentation/userspace-api/iommu.rst
> >
> > diff --git a/Documentation/userspace-api/iommu.rst
> > b/Documentation/userspace-api/iommu.rst new file mode 100644
> > index 000000000000..e95dc5a04a41
> > --- /dev/null
> > +++ b/Documentation/userspace-api/iommu.rst
> > @@ -0,0 +1,210 @@
> > +.. SPDX-License-Identifier: GPL-2.0
> > +.. iommu:
> > +
> > +=====================================
> > +IOMMU Userspace API
> > +=====================================
> > +
> > +IOMMU UAPI is used for virtualization cases where communications
> > are +needed between physical and virtual IOMMU drivers. For native
> > +usage, IOMMU is a system device which does not need to communicate
> > +with user space directly.
> > +
> > +The primary use cases are guest Shared Virtual Address (SVA) and
> > +guest IO virtual address (IOVA), wherein virtual IOMMU (vIOMMU)
> > is
>
> wherein _a_ virtual IOMMU
right,

>
> > +required to communicate with the physical IOMMU in the host.
> > +
> > +.. contents:: :local:
> > +
> > +Functionalities
> > +====================================================
> > +Communications of user and kernel involve both directions. The
> > +supported user-kernel APIs are as follows:
> > +
> > +1. Alloc/Free PASID
> > +2. Bind/unbind guest PASID (e.g. Intel VT-d)
> > +3. Bind/unbind guest PASID table (e.g. ARM sMMU)
> > +4. Invalidate IOMMU caches
> > +5. Service page request
> > +
> > +Requirements
> > +====================================================
> > +The IOMMU UAPIs are generic and extensible to meet the following
> > +requirements:
> > +
> > +1. Emulated and para-virtualised vIOMMUs
> > +2. Multiple vendors (Intel VT-d, ARM sMMU, etc.)
> > +3. Extensions to the UAPI shall not break existing user space
> > +
> > +Interfaces
> > +====================================================
> > +Although the data structures defined in IOMMU UAPI are
> > self-contained, +there is no user API functions introduced.
> > Instead, IOMMU UAPI is +designed to work with existing user driver
> > frameworks such as VFIO. +
> > +Extension Rules & Precautions
> > +-----------------------------
> > +When IOMMU UAPI gets extended, the data structures can *only* be
> > +modified in two ways:
> > +
> > +1. Adding new fields by re-purposing the padding[] field. No size
> > change. +2. Adding new union members at the end. May increase in
> > size. +
> > +No new fields can be added *after* the variable size union in that
> > it +will break backward compatibility when offset moves. In both
> > cases, a +new flag must be accompanied with a new field such that
> > the IOMMU +driver can process the data based on the new flag.
> > Version field is +only reserved for the unlikely event of UAPI
> > upgrade at its entirety. +
> > +It's *always* the caller's responsibility to indicate the size of
> > the +structure passed by setting argsz appropriately.
> > +
> > +When IOMMU UAPI extension results in size increase, user such as
> > VFIO +has to handle the following scenarios:
> > +
> > +1. User and kernel has exact size match
> > +2. An older user with older kernel header (smaller UAPI size)
> > running on a
> > + newer kernel (larger UAPI size)
> > +3. A newer user with newer kernel header (larger UAPI size) running
> > + on a older kernel.
> > +4. A malicious/misbehaving user pass illegal/invalid size but
> > within
> > + range. The data may contain garbage.
> > +
> > +
> > +Feature Checking
> > +----------------
> > +While launching a guest with vIOMMU, it is important to ensure
> > that host +can support the UAPI data structures to be used for
> > vIOMMU-pIOMMU +communications. Without the upfront compatibility
> > checking, future +faults are difficult to report even in normal
> > conditions. For example, +TLB invalidations should always succeed
> > from vIOMMU's +perspective.
>
> This statement has me concerned. If a TLB invalidation fails, but
> is reported to the guest as successful do we have possible breaking
> of iommu isolation guarantees?
>
Good point. we should never report success if TLB invalidation fails.
Perhaps reword as:
"For example, TLB invalidations should always succeed. There is no
architectural way to report back to the vIOMMU if the UAPI data is
incompatible. If that happens, in order to protect IOMMU iosolation
guarantee, we have to resort to not giving completion status. This
may result in VM hang."


> If you get a TLB invalidation not happening, for some reason, that's
> a critical fault, isolate the device using the IOMMU or kill the VM.
>
> I'd reword it as "TLB invalidations should always succeed."
>
yes.

> As you mention, we should never get to this state anyway!
>
> > There is no architectural way to report back to the vIOMMU
> > +if the UAPI data is incompatible. For this reason the following
> > IOMMU +UAPIs cannot fail:
> > +
> > +1. Free PASID
> > +2. Unbind guest PASID
> > +3. Unbind guest PASID table (SMMU)
> > +4. Cache invalidate
> > +5. Page response
>
> Of these, page response is a bit different. Shouldn't be a problem to
> occasionally say a page response was handled when it wasn't (or it
> was but has gone away again before the response reached the
> hardware). In high load environments that happens anyway sometimes.
>
> The others are all cases where any failure at all is fatal to the
> guest continuing to use the hardware.
>
You are right, PRS fail is expected in that guest could be scheduled
out for a long time. host IOMMU driver has tracking and timeout which
will respond back to the device to avoid device hang.

I will remove Page Response from the list.

Here from UAPI compatibility checking perspective, I meant if PRS data
is not compatible, all page responses could fail, difficult to handle.


> > +
> > +User applications such as QEMU is expected to import kernel UAPI
> > +headers. Only backward compatibility is supported. For example, an
> > +older QEMU (with older kernel header) can run on newer kernel.
> > Newer +QEMU (with new kernel header) may fail on older kernel.
>
> I'd define fail a bit tighter here. I presume refuse to initialize?
>
agreed. also with Alex's input, we should still support new header on
old kernel as long as only old features are used. So this could be
"refuse to initialize if new fields/flags are used on the older kernel"

> > +
> > +IOMMU vendor driver should report the below features to IOMMU UAPI
> > +consumers (e.g. via VFIO).
> > +
> > +1. IOMMU_NESTING_FEAT_SYSWIDE_PASID
> > +2. IOMMU_NESTING_FEAT_BIND_PGTBL
> > +3. IOMMU_NESTING_FEAT_BIND_PASID_TABLE
> > +4. IOMMU_NESTING_FEAT_CACHE_INVLD
> > +5. IOMMU_NESTING_FEAT_PAGE_REQUEST
> > +
> > +Take VFIO as example, upon request from VFIO user space (e.g.
> > QEMU), +VFIO kernel code shall query IOMMU vendor driver for the
> > support of +the above features. Query result can then be reported
> > back to the +user-space caller. Details can be found in
> > +Documentation/driver-api/vfio.rst.
> > +
> > +
> > +Data Passing Example with VFIO
> > +------------------------------
> > +As the ubiquitous userspace driver framework, VFIO is already IOMMU
> > +aware and share many key concepts such as device model, group, and
> > +protection domain. Other user driver frameworks can also be
> > extended +to support IOMMU UAPI but it is outside the scope of this
> > document. +
> > +In this tight-knit VFIO-IOMMU interface, the ultimate consumer of
> > the +IOMMU UAPI data is the host IOMMU driver. VFIO facilitates
> > user-kernel +transport, capability checking, security, and life
> > cycle management of +process address space ID (PASID).
> > +
> > +Unlike normal user data passed via VFIO UAPI IOTCL, IOMMU driver
> > is the +ultimate consumer of its UAPI data. At VFIO layer, the
> > IOMMU UAPI data +is wrapped in a VFIO UAPI data for sanity
> > checking. It follows the +pattern below:
> > +
> > +::
> > +
> > + struct {
> > + __u32 argsz;
> > + __u32 flags;
> > + __u8 data[];
> > + }
>
> That final bracket needs to be indented one more space.
> Make sure you check the output of this file as there are a few of
> these.
>
will do. thanks

> > +
> > +Here data[] contains the IOMMU UAPI data structures.
> > +
> > +In order to determine the size and feature set of the user data,
> > argsz +and flags are also embedded in the IOMMU UAPI data
> > structures. +A "__u32 argsz" field is *always* at the beginning of
> > each structure. +
> > +For example:
> > +::
> > +
> > + struct iommu_gpasid_bind_data {
> > + __u32 argsz;
> > + __u32 version;
> > + #define IOMMU_PASID_FORMAT_INTEL_VTD 1
> > + __u32 format;
> > + #define IOMMU_SVA_GPASID_VAL (1 << 0)
> > + __u64 flags;
> > + __u64 gpgd;
> > + __u64 hpasid;
> > + __u64 gpasid;
> > + __u32 addr_width;
> > + __u8 padding[12];
> > + /* Vendor specific data */
> > + union {
> > + struct iommu_gpasid_bind_data_vtd vtd;
> > + };
> > + };
> > +
> > +Use bind guest PASID as an example, VFIO code shall process IOMMU
> > UAPI +request as follows:
> > +
> > +::
> > +
> > + 1 /* Minsz must include IOMMU UAPI "argsz" of __u32 */
> > + 2 minsz = offsetofend(struct vfio_iommu_type1_bind, flags)
> > +
> > + sizeof(u32);
> > + 3 copy_from_user(&vfio_bind, (void __user *)arg, minsz);
> > + 4
> > + 5 /* Check VFIO argsz */
> > + 6 if (vfio_bind.argsz < minsz)
> > + 7 return -EINVAL;
> > + 8
> > + 9 /* VFIO flags must be included in minsz */
>
> Nice to keep indentation across change in line number length.
>
will do

> > + 10 switch (vfio_bind.flags) {
> > + 11 case VFIO_IOMMU_BIND_GUEST_PGTBL:
> > + 12 /*
> > + 13 * Get the current IOMMU bind GPASID data size,
> > + 14 * which accounted for the largest union member.
> > + 15 */
> > + 16 data_size = sizeof(struct
> > iommu_gpasid_bind_data);
> > + 17 iommu_argsz = vfio_bind.argsz - minsz;
> > + 18 if (iommu_argsz > data_size) {
> > + 19 /* User data > current kernel */
> > + 20 return -E2BIG;
> > + 21 }
> > + 22 copy_from_user(&iommu_bind, (void __user *)
> > + 23 vfio_bind.data, iommu_argsz);
> > + 24 /*
> > + 25 * Deal with trailing bytes that is bigger than
> > user
> > + 26 * provided UAPI size but smaller than the
> > current
> > + 27 * kernel data size. Zero fill the trailing
> > bytes.
> > + 28 */
> > + 29 memset(iommu_bind + iommu_argsz, 0, data_size -
> > + 30 iommu_argsz;
> > + 31
> > + 32 iommu_sva_bind_gpasid(domain, dev,
> > iommu_bind_data);
> > + 33 break;
> > +
> > +
> > +Case #1 & 2 are supported per backward compatibility rule.
> > +
> > +Case #3 will fail with -E2BIG at line #20. Case
>
> Is that always the case? As we have multiple structures in a union
> it's possible that a given architecture might expand without changing
> the union size. So this might not detect a UAPI change.
>
not always, if expand without size change, there must be a new flag.
the flag will guide the processing of the new union member.

> > +
> > +Case #4 may result in other error processed by IOMMU vendor
> > driver. However, +the damage shall not exceed the scope of the
> > offending user.
>
>

[Jacob Pan]