Pull VFIO updates from Alex Williamson:
- Cleanup use of extern in function prototypes (Alex Williamson)
- Simplify bus_type usage and convert to device IOMMU interfaces (Robin
Murphy)
- Check missed return value and fix comment typos (Bo Liu)
- Split migration ops from device ops and fix races in mlx5 migration
support (Yishai Hadas)
- Fix missed return value check in noiommu support (Liam Ni)
- Hardening to clear buffer pointer to avoid use-after-free (Schspa
Shi)
- Remove requirement that only the same mm can unmap a previously
mapped range (Li Zhe)
- Adjust semaphore release vs device open counter (Yi Liu)
- Remove unused arg from SPAPR support code (Deming Wang)
- Rework vfio-ccw driver to better fit new mdev framework (Eric Farman,
Michael Kawano)
- Replace DMA unmap notifier with callbacks (Jason Gunthorpe)
- Clarify SPAPR support comment relative to iommu_ops (Alexey
Kardashevskiy)
- Revise page pinning API towards compatibility with future iommufd
support (Nicolin Chen)
- Resolve issues in vfio-ccw, including use of DMA unmap callback (Eric
Farman)
* tag 'vfio-v6.0-rc1' of https://github.com/awilliam/linux-vfio: (40 commits)
vfio/pci: fix the wrong word
vfio/ccw: Check return code from subchannel quiesce
vfio/ccw: Remove FSM Close from remove handlers
vfio/ccw: Add length to DMA_UNMAP checks
vfio: Replace phys_pfn with pages for vfio_pin_pages()
vfio/ccw: Add kmap_local_page() for memcpy
vfio: Rename user_iova of vfio_dma_rw()
vfio/ccw: Change pa_pfn list to pa_iova list
vfio/ap: Change saved_pfn to saved_iova
vfio: Pass in starting IOVA to vfio_pin/unpin_pages API
vfio/ccw: Only pass in contiguous pages
vfio/ap: Pass in physical address of ind to ap_aqic()
drm/i915/gvt: Replace roundup with DIV_ROUND_UP
vfio: Make vfio_unpin_pages() return void
vfio/spapr_tce: Fix the comment
vfio: Replace the iommu notifier with a device list
vfio: Replace the DMA unmapping notifier with a callback
vfio/ccw: Move FSM open/close to MDEV open/close
vfio/ccw: Refactor vfio_ccw_mdev_reset
vfio/ccw: Create a CLOSE FSM event
...
Pull vfs lseek updates from Al Viro:
"Jason's lseek series.
Saner handling of 'lseek should fail with ESPIPE' - this gets rid of
the magical no_llseek thing and makes checks consistent.
In particular, the ad-hoc "can we do splice via internal pipe" checks
got saner (and somewhat more permissive, which is what Jason had been
after, AFAICT)"
* tag 'pull-work.lseek' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs:
fs: remove no_llseek
fs: check FMODE_LSEEK to control internal pipe splicing
vfio: do not set FMODE_LSEEK flag
dma-buf: remove useless FMODE_LSEEK flag
fs: do not compare against ->llseek
fs: clear or set FMODE_LSEEK based on llseek function
Most of the callers of vfio_pin_pages() want "struct page *" and the
low-level mm code to pin pages returns a list of "struct page *" too.
So there's no gain in converting "struct page *" to PFN in between.
Replace the output parameter "phys_pfn" list with a "pages" list, to
simplify callers. This also allows us to replace the vfio_iommu_type1
implementation with a more efficient one.
And drop the pfn_valid check in the gvt code, as there is no need to
do such a check at a page-backed struct page pointer.
For now, also update vfio_iommu_type1 to fit this new parameter too.
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Kirti Wankhede <kwankhede@nvidia.com>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Acked-by: Eric Farman <farman@linux.ibm.com>
Tested-by: Terrence Xu <terrence.xu@intel.com>
Tested-by: Eric Farman <farman@linux.ibm.com>
Signed-off-by: Nicolin Chen <nicolinc@nvidia.com>
Link: https://lore.kernel.org/r/20220723020256.30081-11-nicolinc@nvidia.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
The vfio_pin/unpin_pages() so far accepted arrays of PFNs of user IOVA.
Among all three callers, there was only one caller possibly passing in
a non-contiguous PFN list, which is now ensured to have contiguous PFN
inputs too.
Pass in the starting address with "iova" alone to simplify things, so
callers no longer need to maintain a PFN list or to pin/unpin one page
at a time. This also allows VFIO to use more efficient implementations
of pin/unpin_pages.
For now, also update vfio_iommu_type1 to fit this new parameter too,
while keeping its input intact (being user_iova) since we don't want
to spend too much effort swapping its parameters and local variables
at that level.
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Kirti Wankhede <kwankhede@nvidia.com>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Tony Krowiak <akrowiak@linux.ibm.com>
Acked-by: Eric Farman <farman@linux.ibm.com>
Tested-by: Terrence Xu <terrence.xu@intel.com>
Tested-by: Eric Farman <farman@linux.ibm.com>
Signed-off-by: Nicolin Chen <nicolinc@nvidia.com>
Link: https://lore.kernel.org/r/20220723020256.30081-6-nicolinc@nvidia.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
There's only one caller that checks its return value with a WARN_ON_ONCE,
while all other callers don't check the return value at all. Above that,
an undo function should not fail. So, simplify the API to return void by
embedding similar WARN_ONs.
Also for users to pinpoint which condition fails, separate WARN_ON lines,
yet remove the "driver->ops->unpin_pages" check, since it's unreasonable
for callers to unpin on something totally random that wasn't even pinned.
And remove NULL pointer checks for they would trigger oops vs. warnings.
Note that npage is already validated in the vfio core, thus drop the same
check in the type1 code.
Suggested-by: Christoph Hellwig <hch@infradead.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Reviewed-by: Kirti Wankhede <kwankhede@nvidia.com>
Tested-by: Terrence Xu <terrence.xu@intel.com>
Signed-off-by: Nicolin Chen <nicolinc@nvidia.com>
Link: https://lore.kernel.org/r/20220723020256.30081-2-nicolinc@nvidia.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Instead of bouncing the function call to the driver op through a blocking
notifier just have the iommu layer call it directly.
Register each device that is being attached to the iommu with the lower
driver which then threads them on a linked list and calls the appropriate
driver op at the right time.
Currently the only use is if dma_unmap() is defined.
Also, fully lock all the debugging tests on the pinning path that a
dma_unmap is registered.
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Link: https://lore.kernel.org/r/2-v4-681e038e30fd+78-vfio_unmap_notif_jgg@nvidia.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
The test isn't going to work if a group doesn't exist. Normally this isn't
a problem since VFIO isn't going to create a device if there is no group,
but the special CONFIG_VFIO_NOIOMMU behavior allows bypassing this
prevention. The new cap test effectively forces a group and breaks this
config option.
Move the cap test to vfio_group_find_or_alloc() which is the earliest time
we know we have a group available and thus are not running in noiommu mode.
Fixes: e8ae0e140c ("vfio: Require that devices support DMA cache coherence")
Reported-by: Xiang Chen <chenxiang66@hisilicon.com>
Tested-by: Xiang Chen <chenxiang66@hisilicon.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Link: https://lore.kernel.org/r/0-v1-e8934b490f36+f4-vfio_cap_fix_jgg@nvidia.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
We do not protect the vfio_device::open_count with group_rwsem elsewhere (see
vfio_device_fops_release as a comparison, where we already drop group_rwsem
before open_count--). So move the group_rwsem unlock prior to open_count--.
This change now also drops group_rswem before setting device->kvm = NULL,
but that's also OK (again, just like vfio_device_fops_release). The setting
of device->kvm before open_device is technically done while holding the
group_rwsem, this is done to protect the group kvm value we are copying from,
and we should not be relying on that to protect the contents of device->kvm;
instead we assume this value will not change until after the device is closed
and while under the dev_set->lock.
Cc: Matthew Rosato <mjrosato@linux.ibm.com>
Cc: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Yi Liu <yi.l.liu@intel.com>
Reviewed-by: Matthew Rosato <mjrosato@linux.ibm.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Link: https://lore.kernel.org/r/20220627074119.523274-1-yi.l.liu@intel.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
vfio core checks whether the driver sets some migration op (e.g.
set_state/get_state) and accordingly calls its op.
However, currently mlx5 driver sets the above ops without regards to its
migration caps.
This might lead to unexpected usage/Oops if user space may call to the
above ops even if the driver doesn't support migration. As for example,
the migration state_mutex is not initialized in that case.
The cleanest way to manage that seems to split the migration ops from
the main device ops, this will let the driver setting them separately
from the main ops when it's applicable.
As part of that, validate ops construction on registration and include a
check for VFIO_MIGRATION_STOP_COPY since the uAPI claims it must be set
in migration_flags.
HISI driver was changed as well to match this scheme.
This scheme may enable down the road to come with some extra group of
ops (e.g. DMA log) that can be set without regards to the other options
based on driver caps.
Fixes: 6fadb02126 ("vfio/mlx5: Implement vfio_pci driver for mlx5 devices")
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Signed-off-by: Yishai Hadas <yishaih@nvidia.com>
Link: https://lore.kernel.org/r/20220628155910.171454-3-yishaih@nvidia.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Pull vfio updates from Alex Williamson:
- Improvements to mlx5 vfio-pci variant driver, including support for
parallel migration per PF (Yishai Hadas)
- Remove redundant iommu_present() check (Robin Murphy)
- Ongoing refactoring to consolidate the VFIO driver facing API to use
vfio_device (Jason Gunthorpe)
- Use drvdata to store vfio_device among all vfio-pci and variant
drivers (Jason Gunthorpe)
- Remove redundant code now that IOMMU core manages group DMA ownership
(Jason Gunthorpe)
- Remove vfio_group from external API handling struct file ownership
(Jason Gunthorpe)
- Correct typo in uapi comments (Thomas Huth)
- Fix coccicheck detected deadlock (Wan Jiabing)
- Use rwsem to remove races and simplify code around container and kvm
association to groups (Jason Gunthorpe)
- Harden access to devices in low power states and use runtime PM to
enable d3cold support for unused devices (Abhishek Sahu)
- Fix dma_owner handling of fake IOMMU groups (Jason Gunthorpe)
- Set driver_managed_dma on vfio-pci variant drivers (Jason Gunthorpe)
- Pass KVM pointer directly rather than via notifier (Matthew Rosato)
* tag 'vfio-v5.19-rc1' of https://github.com/awilliam/linux-vfio: (38 commits)
vfio: remove VFIO_GROUP_NOTIFY_SET_KVM
vfio/pci: Add driver_managed_dma to the new vfio_pci drivers
vfio: Do not manipulate iommu dma_owner for fake iommu groups
vfio/pci: Move the unused device into low power state with runtime PM
vfio/pci: Virtualize PME related registers bits and initialize to zero
vfio/pci: Change the PF power state to D0 before enabling VFs
vfio/pci: Invalidate mmaps and block the access in D3hot power state
vfio: Change struct vfio_group::container_users to a non-atomic int
vfio: Simplify the life cycle of the group FD
vfio: Fully lock struct vfio_group::container
vfio: Split up vfio_group_get_device_fd()
vfio: Change struct vfio_group::opened from an atomic to bool
vfio: Add missing locking for struct vfio_group::kvm
kvm/vfio: Fix potential deadlock problem in vfio
include/uapi/linux/vfio.h: Fix trivial typo - _IORW should be _IOWR instead
vfio/pci: Use the struct file as the handle not the vfio_group
kvm/vfio: Remove vfio_group from kvm
vfio: Change vfio_group_set_kvm() to vfio_file_set_kvm()
vfio: Change vfio_external_check_extension() to vfio_file_enforced_coherent()
vfio: Remove vfio_external_group_match_file()
...
Since asserting dma ownership now causes the group to have its DMA blocked
the iommu layer requires a working iommu. This means the dma_owner APIs
cannot be used on the fake groups that VFIO creates. Test for this and
avoid calling them.
Otherwise asserting dma ownership will fail for VFIO mdev devices as a
BLOCKING iommu_domain cannot be allocated due to the NULL iommu ops.
Fixes: 0286300e60 ("iommu: iommu_group_claim_dma_owner() must always assign a domain")
Reported-by: Eric Farman <farman@linux.ibm.com>
Tested-by: Eric Farman <farman@linux.ibm.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/0-v1-9cfc47edbcd4+13546-vfio_dma_owner_fix_jgg@nvidia.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Once userspace opens a group FD it is prevented from opening another
instance of that same group FD until all the prior group FDs and users of
the container are done.
The first is done trivially by checking the group->opened during group FD
open.
However, things get a little weird if userspace creates a device FD and
then closes the group FD. The group FD still cannot be re-opened, but this
time it is because the group->container is still set and container_users
is elevated by the device FD.
Due to this mismatched lifecycle we have the
vfio_group_try_dissolve_container() which tries to auto-free a container
after the group FD is closed but the device FD remains open.
Instead have the device FD hold onto a reference to the single group
FD. This directly prevents vfio_group_fops_release() from being called
when any device FD exists and makes the lifecycle model more
understandable.
vfio_group_try_dissolve_container() is removed as the only place a
container is auto-deleted is during vfio_group_fops_release(). At this
point the container_users is either 1 or 0 since all device FDs must be
closed.
Change group->opened to group->opened_file which points to the single
struct file * that is open for the group. If the group->open_file is
NULL then group->container == NULL.
If all device FDs have closed then the group's notifier list must be
empty.
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Tested-by: Nicolin Chen <nicolinc@nvidia.com>
Tested-by: Matthew Rosato <mjrosato@linux.ibm.com>
Link: https://lore.kernel.org/r/5-v2-d035a1842d81+1bf-vfio_group_locking_jgg@nvidia.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
This is necessary to avoid various user triggerable races, for instance
racing SET_CONTAINER/UNSET_CONTAINER:
ioctl(VFIO_GROUP_SET_CONTAINER)
ioctl(VFIO_GROUP_UNSET_CONTAINER)
vfio_group_unset_container
int users = atomic_cmpxchg(&group->container_users, 1, 0);
// users == 1 container_users == 0
__vfio_group_unset_container(group);
container = group->container;
vfio_group_set_container()
if (!atomic_read(&group->container_users))
down_write(&container->group_lock);
group->container = container;
up_write(&container->group_lock);
down_write(&container->group_lock);
group->container = NULL;
up_write(&container->group_lock);
vfio_container_put(container);
/* woops we lost/leaked the new container */
This can then go on to NULL pointer deref since container == 0 and
container_users == 1.
Wrap all touches of container, except those on a performance path with a
known open device, with the group_rwsem.
The only user of vfio_group_add_container_user() holds the user count for
a simple operation, change it to just hold the group_lock over the
operation and delete vfio_group_add_container_user(). Containers now only
gain a user when a device FD is opened.
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Tested-by: Nicolin Chen <nicolinc@nvidia.com>
Tested-by: Matthew Rosato <mjrosato@linux.ibm.com>
Link: https://lore.kernel.org/r/4-v2-d035a1842d81+1bf-vfio_group_locking_jgg@nvidia.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Without locking userspace can trigger a UAF by racing
KVM_DEV_VFIO_GROUP_DEL with VFIO_GROUP_GET_DEVICE_FD:
CPU1 CPU2
ioctl(KVM_DEV_VFIO_GROUP_DEL)
ioctl(VFIO_GROUP_GET_DEVICE_FD)
vfio_group_get_device_fd
open_device()
intel_vgpu_open_device()
vfio_register_notifier()
vfio_register_group_notifier()
blocking_notifier_call_chain(&group->notifier,
VFIO_GROUP_NOTIFY_SET_KVM, group->kvm);
set_kvm()
group->kvm = NULL
close()
kfree(kvm)
intel_vgpu_group_notifier()
vdev->kvm = data
[..]
kvm_get_kvm(vgpu->kvm);
// UAF!
Add a simple rwsem in the group to protect the kvm while the notifier is
using it.
Note this doesn't fix the race internal to i915 where userspace can
trigger two VFIO_GROUP_NOTIFY_SET_KVM's before we reach a consumer of
vgpu->kvm and trigger this same UAF, it just makes the notifier
self-consistent.
Fixes: ccd46dbae7 ("vfio: support notifier chain in vfio_group")
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Tested-by: Nicolin Chen <nicolinc@nvidia.com>
Tested-by: Matthew Rosato <mjrosato@linux.ibm.com>
Link: https://lore.kernel.org/r/1-v2-d035a1842d81+1bf-vfio_group_locking_jgg@nvidia.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
VFIO PCI does a security check as part of hot reset to prove that the user
has permission to manipulate all the devices that will be impacted by the
reset.
Use a new API vfio_file_has_dev() to perform this security check against
the struct file directly and remove the vfio_group from VFIO PCI.
Since VFIO PCI was the last user of vfio_group_get_external_user() and
vfio_group_put_external_user() remove it as well.
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Link: https://lore.kernel.org/r/8-v3-f7729924a7ea+25e33-vfio_kvm_no_group_jgg@nvidia.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
vfio_group_fops_open() ensures there is only ever one struct file open for
any struct vfio_group at any time:
/* Do we need multiple instances of the group open? Seems not. */
opened = atomic_cmpxchg(&group->opened, 0, 1);
if (opened) {
vfio_group_put(group);
return -EBUSY;
Therefor the struct file * can be used directly to search the list of VFIO
groups that KVM keeps instead of using the
vfio_external_group_match_file() callback to try to figure out if the
passed in FD matches the list or not.
Delete vfio_external_group_match_file().
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Yi Liu <yi.l.liu@intel.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Link: https://lore.kernel.org/r/4-v3-f7729924a7ea+25e33-vfio_kvm_no_group_jgg@nvidia.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
The only caller wants to get a pointer to the struct iommu_group
associated with the VFIO group file. Instead of returning the group ID
then searching sysfs for that string to get the struct iommu_group just
directly return the iommu_group pointer already held by the vfio_group
struct.
It already has a safe lifetime due to the struct file kref, the vfio_group
and thus the iommu_group cannot be destroyed while the group file is open.
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Yi Liu <yi.l.liu@intel.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Link: https://lore.kernel.org/r/3-v3-f7729924a7ea+25e33-vfio_kvm_no_group_jgg@nvidia.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
The last user of this function is in PCI callbacks that want to convert
their struct pci_dev to a vfio_device. Instead of searching use the
vfio_device available trivially through the drvdata.
When a callback in the device_driver is called, the caller must hold the
device_lock() on dev. The purpose of the device_lock is to prevent
remove() from being called (see __device_release_driver), and allow the
driver to safely interact with its drvdata without races.
The PCI core correctly follows this and holds the device_lock() when
calling error_detected (see report_error_detected) and
sriov_configure (see sriov_numvfs_store).
Further, since the drvdata holds a positive refcount on the vfio_device
any access of the drvdata, under the device_lock(), from a driver callback
needs no further protection or refcounting.
Thus the remark in the vfio_device_get_from_dev() comment does not apply
here, VFIO PCI drivers all call vfio_unregister_group_dev() from their
remove callbacks under the device_lock() and cannot race with the
remaining callers.
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Shameer Kolothum <shameerali.kolothum.thodi@huawei.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Link: https://lore.kernel.org/r/2-v4-c841817a0349+8f-vfio_get_from_dev_jgg@nvidia.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
When the open_device() op is called the container_users is incremented and
held incremented until close_device(). Thus, so long as drivers call
functions within their open_device()/close_device() region they do not
need to worry about the container_users.
These functions can all only be called between open_device() and
close_device():
vfio_pin_pages()
vfio_unpin_pages()
vfio_dma_rw()
vfio_register_notifier()
vfio_unregister_notifier()
Eliminate the calls to vfio_group_add_container_user() and add
vfio_assert_device_open() to detect driver mis-use. This causes the
close_device() op to check device->open_count so always leave it elevated
while calling the op.
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Link: https://lore.kernel.org/r/7-v4-8045e76bf00b+13d-vfio_mdev_no_group_jgg@nvidia.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Every caller has a readily available vfio_device pointer, use that instead
of passing in a generic struct device. Change vfio_dma_rw() to take in the
struct vfio_device and move the container users that would have been held
by vfio_group_get_external_user_from_dev() to vfio_dma_rw() directly, like
vfio_pin/unpin_pages().
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Link: https://lore.kernel.org/r/4-v4-8045e76bf00b+13d-vfio_mdev_no_group_jgg@nvidia.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
IOMMU_CACHE means that normal DMAs do not require any additional coherency
mechanism and is the basic uAPI that VFIO exposes to userspace. For
instance VFIO applications like DPDK will not work if additional coherency
operations are required.
Therefore check IOMMU_CAP_CACHE_COHERENCY like vdpa & usnic do before
allowing an IOMMU backed VFIO device to be created.
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Acked-by: Alex Williamson <alex.williamson@redhat.com>
Acked-by: Robin Murphy <robin.murphy@arm.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Link: https://lore.kernel.org/r/4-v3-2cf356649677+a32-intel_no_snoop_jgg@nvidia.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
commit 60720a0fc6 ("vfio: Add device tracking during unbind") added the
unbound list to plug a problem with KVM where KVM_DEV_VFIO_GROUP_DEL
relied on vfio_group_get_external_user() succeeding to return the
vfio_group from a group file descriptor. The unbound list allowed
vfio_group_get_external_user() to continue to succeed in edge cases.
However commit 5d6dee80a1 ("vfio: New external user group/file match")
deleted the call to vfio_group_get_external_user() during
KVM_DEV_VFIO_GROUP_DEL. Instead vfio_external_group_match_file() is used
to directly match the file descriptor to the group pointer.
This in turn avoids the call down to vfio_dev_viable() during
KVM_DEV_VFIO_GROUP_DEL and also avoids the trouble the first commit was
trying to fix.
There are no other users of vfio_dev_viable() that care about the time
after vfio_unregister_group_dev() returns, so simply delete the
unbound_list entirely.
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
Acked-by: Alex Williamson <alex.williamson@redhat.com>
Link: https://lore.kernel.org/r/20220418005000.897664-10-baolu.lu@linux.intel.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
As DMA ownership is claimed for the iommu group when a VFIO group is
added to a VFIO container, the VFIO group viability is guaranteed as long
as group->container_users > 0. Remove those unnecessary group viability
checks which are only hit when group->container_users is not zero.
The only remaining reference is in GROUP_GET_STATUS, which could be called
at any time when group fd is valid. Here we just replace the
vfio_group_viable() by directly calling IOMMU core to get viability status.
Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Acked-by: Alex Williamson <alex.williamson@redhat.com>
Link: https://lore.kernel.org/r/20220418005000.897664-9-baolu.lu@linux.intel.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
Claim group dma ownership when an IOMMU group is set to a container,
and release the dma ownership once the iommu group is unset from the
container.
This change disallows some unsafe bridge drivers to bind to non-ACS
bridges while devices under them are assigned to user space. This is an
intentional enhancement and possibly breaks some existing
configurations. The recommendation to such an affected user would be
that the previously allowed host bridge driver was unsafe for this use
case and to continue to enable assignment of devices within that group,
the driver should be unbound from the bridge device or replaced with the
pci-stub driver.
For any bridge driver, we consider it unsafe if it satisfies any of the
following conditions:
1) The bridge driver uses DMA. Calling pci_set_master() or calling any
kernel DMA API (dma_map_*() and etc.) is an indicate that the
driver is doing DMA.
2) If the bridge driver uses MMIO, it should be tolerant to hostile
userspace also touching the same MMIO registers via P2P DMA
attacks.
If the bridge driver turns out to be a safe one, it could be used as
before by setting the driver's .driver_managed_dma field, just like what
we have done in the pcieport driver.
Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Acked-by: Alex Williamson <alex.williamson@redhat.com>
Link: https://lore.kernel.org/r/20220418005000.897664-8-baolu.lu@linux.intel.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
The RUNNING_P2P state is designed to support multiple devices in the same
VM that are doing P2P transactions between themselves. When in RUNNING_P2P
the device must be able to accept incoming P2P transactions but should not
generate outgoing P2P transactions.
As an optional extension to the mandatory states it is defined as
in between STOP and RUNNING:
STOP -> RUNNING_P2P -> RUNNING -> RUNNING_P2P -> STOP
For drivers that are unable to support RUNNING_P2P the core code
silently merges RUNNING_P2P and RUNNING together. Unless driver support
is present, the new state cannot be used in SET_STATE.
Drivers that support this will be required to implement 4 FSM arcs
beyond the basic FSM. 2 of the basic FSM arcs become combination
transitions.
Compared to the v1 clarification, NDMA is redefined into FSM states and is
described in terms of the desired P2P quiescent behavior, noting that
halting all DMA is an acceptable implementation.
Link: https://lore.kernel.org/all/20220224142024.147653-11-yishaih@nvidia.com
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Tested-by: Shameer Kolothum <shameerali.kolothum.thodi@huawei.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Yishai Hadas <yishaih@nvidia.com>
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Replace the existing region based migration protocol with an ioctl based
protocol. The two protocols have the same general semantic behaviors, but
the way the data is transported is changed.
This is the STOP_COPY portion of the new protocol, it defines the 5 states
for basic stop and copy migration and the protocol to move the migration
data in/out of the kernel.
Compared to the clarification of the v1 protocol Alex proposed:
https://lore.kernel.org/r/163909282574.728533.7460416142511440919.stgit@omen
This has a few deliberate functional differences:
- ERROR arcs allow the device function to remain unchanged.
- The protocol is not required to return to the original state on
transition failure. Instead userspace can execute an unwind back to
the original state, reset, or do something else without needing kernel
support. This simplifies the kernel design and should userspace choose
a policy like always reset, avoids doing useless work in the kernel
on error handling paths.
- PRE_COPY is made optional, userspace must discover it before using it.
This reflects the fact that the majority of drivers we are aware of
right now will not implement PRE_COPY.
- segmentation is not part of the data stream protocol, the receiver
does not have to reproduce the framing boundaries.
The hybrid FSM for the device_state is described as a Mealy machine by
documenting each of the arcs the driver is required to implement. Defining
the remaining set of old/new device_state transitions as 'combination
transitions' which are naturally defined as taking multiple FSM arcs along
the shortest path within the FSM's digraph allows a complete matrix of
transitions.
A new VFIO_DEVICE_FEATURE of VFIO_DEVICE_FEATURE_MIG_DEVICE_STATE is
defined to replace writing to the device_state field in the region. This
allows returning a brand new FD whenever the requested transition opens
a data transfer session.
The VFIO core code implements the new feature and provides a helper
function to the driver. Using the helper the driver only has to
implement 6 of the FSM arcs and the other combination transitions are
elaborated consistently from those arcs.
A new VFIO_DEVICE_FEATURE of VFIO_DEVICE_FEATURE_MIGRATION is defined to
report the capability for migration and indicate which set of states and
arcs are supported by the device. The FSM provides a lot of flexibility to
make backwards compatible extensions but the VFIO_DEVICE_FEATURE also
allows for future breaking extensions for scenarios that cannot support
even the basic STOP_COPY requirements.
The VFIO_DEVICE_FEATURE_MIG_DEVICE_STATE with the GET option (i.e.
VFIO_DEVICE_FEATURE_GET) can be used to read the current migration state
of the VFIO device.
Data transfer sessions are now carried over a file descriptor, instead of
the region. The FD functions for the lifetime of the data transfer
session. read() and write() transfer the data with normal Linux stream FD
semantics. This design allows future expansion to support poll(),
io_uring, and other performance optimizations.
The complicated mmap mode for data transfer is discarded as current qemu
doesn't take meaningful advantage of it, and the new qemu implementation
avoids substantially all the performance penalty of using a read() on the
region.
Link: https://lore.kernel.org/all/20220224142024.147653-10-yishaih@nvidia.com
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Tested-by: Shameer Kolothum <shameerali.kolothum.thodi@huawei.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Alex Williamson <alex.williamson@redhat.com>
Reviewed-by: Cornelia Huck <cohuck@redhat.com>
Signed-off-by: Yishai Hadas <yishaih@nvidia.com>
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
vfio.c abuses (misuses) "/**", which indicates the beginning of
kernel-doc notation in the kernel tree. This causes a bunch of
kernel-doc complaints about this source file, so quieten all of
them by changing all "/**" to "/*".
vfio.c:236: warning: This comment starts with '/**', but isn't a kernel-doc comment. Refer Documentation/doc-guide/kernel-doc.rst
* IOMMU driver registration
vfio.c:236: warning: missing initial short description on line:
* IOMMU driver registration
vfio.c:295: warning: expecting prototype for Container objects(). Prototype was for vfio_container_get() instead
vfio.c:317: warning: expecting prototype for Group objects(). Prototype was for __vfio_group_get_from_iommu() instead
vfio.c:496: warning: Function parameter or member 'device' not described in 'vfio_device_put'
vfio.c:496: warning: expecting prototype for Device objects(). Prototype was for vfio_device_put() instead
vfio.c:599: warning: This comment starts with '/**', but isn't a kernel-doc comment. Refer Documentation/doc-guide/kernel-doc.rst
* Async device support
vfio.c:599: warning: missing initial short description on line:
* Async device support
vfio.c:693: warning: This comment starts with '/**', but isn't a kernel-doc comment. Refer Documentation/doc-guide/kernel-doc.rst
* VFIO driver API
vfio.c:693: warning: missing initial short description on line:
* VFIO driver API
vfio.c:835: warning: This comment starts with '/**', but isn't a kernel-doc comment. Refer Documentation/doc-guide/kernel-doc.rst
* Get a reference to the vfio_device for a device. Even if the
vfio.c:835: warning: missing initial short description on line:
* Get a reference to the vfio_device for a device. Even if the
vfio.c:969: warning: This comment starts with '/**', but isn't a kernel-doc comment. Refer Documentation/doc-guide/kernel-doc.rst
* VFIO base fd, /dev/vfio/vfio
vfio.c:969: warning: missing initial short description on line:
* VFIO base fd, /dev/vfio/vfio
vfio.c:1187: warning: This comment starts with '/**', but isn't a kernel-doc comment. Refer Documentation/doc-guide/kernel-doc.rst
* VFIO Group fd, /dev/vfio/$GROUP
vfio.c:1187: warning: missing initial short description on line:
* VFIO Group fd, /dev/vfio/$GROUP
vfio.c:1540: warning: This comment starts with '/**', but isn't a kernel-doc comment. Refer Documentation/doc-guide/kernel-doc.rst
* VFIO Device fd
vfio.c:1540: warning: missing initial short description on line:
* VFIO Device fd
vfio.c:1615: warning: This comment starts with '/**', but isn't a kernel-doc comment. Refer Documentation/doc-guide/kernel-doc.rst
* External user API, exported by symbols to be linked dynamically.
vfio.c:1615: warning: missing initial short description on line:
* External user API, exported by symbols to be linked dynamically.
vfio.c:1663: warning: This comment starts with '/**', but isn't a kernel-doc comment. Refer Documentation/doc-guide/kernel-doc.rst
* External user API, exported by symbols to be linked dynamically.
vfio.c:1663: warning: missing initial short description on line:
* External user API, exported by symbols to be linked dynamically.
vfio.c:1742: warning: Function parameter or member 'caps' not described in 'vfio_info_cap_add'
vfio.c:1742: warning: Function parameter or member 'size' not described in 'vfio_info_cap_add'
vfio.c:1742: warning: Function parameter or member 'id' not described in 'vfio_info_cap_add'
vfio.c:1742: warning: Function parameter or member 'version' not described in 'vfio_info_cap_add'
vfio.c:1742: warning: expecting prototype for Sub(). Prototype was for vfio_info_cap_add() instead
vfio.c:2276: warning: This comment starts with '/**', but isn't a kernel-doc comment. Refer Documentation/doc-guide/kernel-doc.rst
* Module/class support
vfio.c:2276: warning: missing initial short description on line:
* Module/class support
Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
Reported-by: kernel test robot <lkp@intel.com>
Cc: Jason Gunthorpe <jgg@nvidia.com>
Cc: Alex Williamson <alex.williamson@redhat.com>
Cc: Eric Auger <eric.auger@redhat.com>
Cc: Cornelia Huck <cohuck@redhat.com>
Cc: kvm@vger.kernel.org
Link: https://lore.kernel.org/r/38a9cb92-a473-40bf-b8f9-85cc5cfc2da4@infradead.org
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Reviewed-by: Cornelia Huck <cohuck@redhat.com>
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
