Pull iommufd updates from Jason Gunthorpe:
"This broadly brings the assigned HW command queue support to iommufd.
This feature is used to improve SVA performance in VMs by avoiding
paravirtualization traps during SVA invalidations.
Along the way I think some of the core logic is in a much better state
to support future driver backed features.
Summary:
- IOMMU HW now has features to directly assign HW command queues to a
guest VM. In this mode the command queue operates on a limited set
of invalidation commands that are suitable for improving guest
invalidation performance and easy for the HW to virtualize.
This brings the generic infrastructure to allow IOMMU drivers to
expose such command queues through the iommufd uAPI, mmap the
doorbell pages, and get the guest physical range for the command
queue ring itself.
- An implementation for the NVIDIA SMMUv3 extension "cmdqv" is built
on the new iommufd command queue features. It works with the
existing SMMU driver support for cmdqv in guest VMs.
- Many precursor cleanups and improvements to support the above
cleanly, changes to the general ioctl and object helpers, driver
support for VDEVICE, and mmap pgoff cookie infrastructure.
- Sequence VDEVICE destruction to always happen before VFIO device
destruction. When using the above type features, and also in future
confidential compute, the internal virtual device representation
becomes linked to HW or CC TSM configuration and objects. If a VFIO
device is removed from iommufd those HW objects should also be
cleaned up to prevent a sort of UAF. This became important now that
we have HW backing the VDEVICE.
- Fix one syzkaller found error related to math overflows during iova
allocation"
* tag 'for-linus-iommufd' of git://git.kernel.org/pub/scm/linux/kernel/git/jgg/iommufd: (57 commits)
iommu/arm-smmu-v3: Replace vsmmu_size/type with get_viommu_size
iommu/arm-smmu-v3: Do not bother impl_ops if IOMMU_VIOMMU_TYPE_ARM_SMMUV3
iommufd: Rename some shortterm-related identifiers
iommufd/selftest: Add coverage for vdevice tombstone
iommufd/selftest: Explicitly skip tests for inapplicable variant
iommufd/vdevice: Remove struct device reference from struct vdevice
iommufd: Destroy vdevice on idevice destroy
iommufd: Add a pre_destroy() op for objects
iommufd: Add iommufd_object_tombstone_user() helper
iommufd/viommu: Roll back to use iommufd_object_alloc() for vdevice
iommufd/selftest: Test reserved regions near ULONG_MAX
iommufd: Prevent ALIGN() overflow
iommu/tegra241-cmdqv: import IOMMUFD module namespace
iommufd: Do not allow _iommufd_object_alloc_ucmd if abort op is set
iommu/tegra241-cmdqv: Add IOMMU_VEVENTQ_TYPE_TEGRA241_CMDQV support
iommu/tegra241-cmdqv: Add user-space use support
iommu/tegra241-cmdqv: Do not statically map LVCMDQs
iommu/tegra241-cmdqv: Simplify deinit flow in tegra241_cmdqv_remove_vintf()
iommu/tegra241-cmdqv: Use request_threaded_irq
iommu/arm-smmu-v3-iommufd: Add hw_info to impl_ops
...
Commit 869c788909 ("selftests: harness: Stop using setjmp()/longjmp()")
changed the harness structure. For some unknown reason, two build warnings
occur to the iommufd selftest:
iommufd.c: In function ‘wrapper_iommufd_mock_domain_all_aligns’:
iommufd.c:1807:17: warning: ‘mfd’ may be used uninitialized in this function
1807 | close(mfd);
| ^~~~~~~~~~
iommufd.c:1767:13: note: ‘mfd’ was declared here
1767 | int mfd;
| ^~~
iommufd.c: In function ‘wrapper_iommufd_mock_domain_all_aligns_copy’:
iommufd.c:1870:17: warning: ‘mfd’ may be used uninitialized in this function
1870 | close(mfd);
| ^~~~~~~~~~
iommufd.c:1819:13: note: ‘mfd’ was declared here
1819 | int mfd;
| ^~~
All the mfd have been used in the variant->file path only, so it's likely
a false alarm.
FWIW, the commit mentioned above does not cause this, yet it might affect
gcc in a certain way that resulted in the warnings. It is also found that
ading a dummy setjmp (which doesn't make sense) could mute the warnings:
https://lore.kernel.org/all/aEi8DV+ReF3v3Rlf@nvidia.com/
The job of this selftest is to catch kernel bug, while such warnings will
unlikely disrupt its role. Mute the warning by force initializing the mfd
and add an ASSERT_GT().
Link: https://patch.msgid.link/r/6951d85d5cd34cbf22abab7714542654e63ecc44.1750787928.git.nicolinc@nvidia.com
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Nicolin Chen <nicolinc@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
The hugepage test cases of iommufd_dirty_tracking have the 64MB and 128MB
coverages. Both of them are smaller than the default hugepage size 512MB,
when CONFIG_PAGE_SIZE_64KB=y. However, these test cases have a variant of
using huge pages, which would mmap(MAP_HUGETLB) using these smaller sizes
than the system hugepag size. This results in the kernel aligning up the
smaller size to 512MB. If a memory was located between the upper 64/128MB
size boundary and the hugepage 512MB boundary, it would get wiped out:
https://lore.kernel.org/all/aEoUhPYIAizTLADq@nvidia.com/
Given that this aligning up behavior is well documented, we have no choice
but to allocate a hugepage aligned size to avoid this unintended wipe out.
Instead of relying on the kernel's internal force alignment, pass the same
size to posix_memalign() and map().
Also, fix the FIXTURE_TEARDOWN() misusing munmap() to free the memory from
posix_memalign(), as munmap() doesn't destroy the allocator meta data. So,
call free() instead.
Fixes: a9af47e382 ("iommufd/selftest: Test IOMMU_HWPT_GET_DIRTY_BITMAP")
Link: https://patch.msgid.link/r/1ea8609ae6d523fdd4d8efb179ddee79c8582cb6.1750787928.git.nicolinc@nvidia.com
Cc: stable@vger.kernel.org
Suggested-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Nicolin Chen <nicolinc@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
With a vIOMMU object, use space can flush any IOMMU related cache that can
be directed via a vIOMMU object. It is similar to the IOMMU_HWPT_INVALIDATE
uAPI, but can cover a wider range than IOTLB, e.g. device/desciprtor cache.
Allow hwpt_id of the iommu_hwpt_invalidate structure to carry a viommu_id,
and reuse the IOMMU_HWPT_INVALIDATE uAPI for vIOMMU invalidations. Drivers
can define different structures for vIOMMU invalidations v.s. HWPT ones.
Since both the HWPT-based and vIOMMU-based invalidation pathways check own
cache invalidation op, remove the WARN_ON_ONCE in the allocator.
Update the uAPI, kdoc, and selftest case accordingly.
Link: https://patch.msgid.link/r/b411e2245e303b8a964f39f49453a5dff280968f.1730836308.git.nicolinc@nvidia.com
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Signed-off-by: Nicolin Chen <nicolinc@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
This results in passing 0 or just IOMMU_CACHE to iommu_map(). Most of
the page table formats don't like this:
amdv1 - -EINVAL
armv7s - returns 0, doesn't update mapped
arm-lpae - returns 0 doesn't update mapped
dart - returns 0, doesn't update mapped
VT-D - returns -EINVAL
Unfortunately the three formats that return 0 cause serious problems:
- Returning ret = but not uppdating mapped from domain->map_pages()
causes an infinite loop in __iommu_map()
- Not writing ioptes means that VFIO/iommufd have no way to recover them
and we will have memory leaks and worse during unmap
Since almost nothing can support this, and it is a useless thing to do,
block it early in iommufd.
Cc: stable@kernel.org
Fixes: aad37e71d5 ("iommufd: IOCTLs for the io_pagetable")
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Reviewed-by: Nicolin Chen <nicolinc@nvidia.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Link: https://lore.kernel.org/r/1-v1-1211e1294c27+4b1-iommu_no_prot_jgg@nvidia.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
Lu Baolu says:
====================
This series implements the functionality of delivering IO page faults to
user space through the IOMMUFD framework. One feasible use case is the
nested translation. Nested translation is a hardware feature that supports
two-stage translation tables for IOMMU. The second-stage translation table
is managed by the host VMM, while the first-stage translation table is
owned by user space. This allows user space to control the IOMMU mappings
for its devices.
When an IO page fault occurs on the first-stage translation table, the
IOMMU hardware can deliver the page fault to user space through the
IOMMUFD framework. User space can then handle the page fault and respond
to the device top-down through the IOMMUFD. This allows user space to
implement its own IO page fault handling policies.
User space application that is capable of handling IO page faults should
allocate a fault object, and bind the fault object to any domain that it
is willing to handle the fault generatd for them. On a successful return
of fault object allocation, the user can retrieve and respond to page
faults by reading or writing to the file descriptor (FD) returned.
The iommu selftest framework has been updated to test the IO page fault
delivery and response functionality.
====================
* iommufd_pri:
iommufd/selftest: Add coverage for IOPF test
iommufd/selftest: Add IOPF support for mock device
iommufd: Associate fault object with iommufd_hw_pgtable
iommufd: Fault-capable hwpt attach/detach/replace
iommufd: Add iommufd fault object
iommufd: Add fault and response message definitions
iommu: Extend domain attach group with handle support
iommu: Add attach handle to struct iopf_group
iommu: Remove sva handle list
iommu: Introduce domain attachment handle
Link: https://lore.kernel.org/all/20240702063444.105814-1-baolu.lu@linux.intel.com
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Extend the selftest tool to add coverage of testing IOPF handling. This
would include the following tests:
- Allocating and destroying an iommufd fault object.
- Allocating and destroying an IOPF-capable HWPT.
- Attaching/detaching/replacing an IOPF-capable HWPT on a device.
- Triggering an IOPF on the mock device.
- Retrieving and responding to the IOPF through the file interface.
Link: https://lore.kernel.org/r/20240702063444.105814-11-baolu.lu@linux.intel.com
Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
With 64k base pages, the first 128k iova length test requires less than a
byte for a bitmap, exposing a bug in the tests that assume that bitmaps are
at least a byte.
Rather than dealing with bytes, have _test_mock_dirty_bitmaps() pass the
number of bits. The caller functions are adjusted to also use bits as well,
and converting to bytes when clearing, allocating and freeing the bitmap.
Link: https://lore.kernel.org/r/20240627110105.62325-2-joao.m.martins@oracle.com
Reported-by: Matt Ochs <mochs@nvidia.com>
Fixes: a9af47e382 ("iommufd/selftest: Test IOMMU_HWPT_GET_DIRTY_BITMAP")
Signed-off-by: Joao Martins <joao.m.martins@oracle.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Tested-by: Matt Ochs <mochs@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Leverage previously added MOCK_FLAGS_DEVICE_HUGE_IOVA flag to create an
IOMMU domain with more than MOCK_IO_PAGE_SIZE supported.
Plumb the hugetlb backing memory for buffer allocation and change the
expected page size to MOCK_HUGE_PAGE_SIZE (1M) when hugepage variant test
cases are used. These so far are limited to 128M and 256M IOVA range tests
cases which is when 1M hugepages can be used.
Link: https://lore.kernel.org/r/20240202133415.23819-9-joao.m.martins@oracle.com
Signed-off-by: Joao Martins <joao.m.martins@oracle.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Rework the functions that test and set the bitmaps to receive a new
parameter (the pte_page_size) that reflects the expected PTE size in the
page tables. The same scheme is still used i.e. even bits are dirty and
odd page indexes aren't dirty. Here it just refactors to consider the size
of the PTE rather than hardcoded to IOMMU mock base page assumptions.
While at it, refactor dirty bitmap tests to use the idev_id created by the
fixture instead of creating a new one.
This is in preparation for doing tests with IOMMU hugepages where multiple
bits set as part of recording a whole hugepage as dirty and thus the
pte_page_size will vary depending on io hugepages or io base pages.
Link: https://lore.kernel.org/r/20240202133415.23819-6-joao.m.martins@oracle.com
Signed-off-by: Joao Martins <joao.m.martins@oracle.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Exercise the dirty tracking bitmaps with byte unaligned addresses in
addition to the PAGE_SIZE unaligned bitmaps, using a address towards the
end of the page boundary.
In doing so, increase the tailroom we allocate for the bitmap from
MOCK_PAGE_SIZE(2K) into PAGE_SIZE(4K), such that we can test end of bitmap
boundary.
Link: https://lore.kernel.org/r/20240202133415.23819-4-joao.m.martins@oracle.com
Signed-off-by: Joao Martins <joao.m.martins@oracle.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Change test_mock_dirty_bitmaps() to pass a flag where it specifies the flag
under test. The test does the same thing as the GET_DIRTY_BITMAP regular
test. Except that it tests whether the dirtied bits are fetched all the
same a second time, as opposed to observing them cleared.
Link: https://lore.kernel.org/r/20231024135109.73787-19-joao.m.martins@oracle.com
Signed-off-by: Joao Martins <joao.m.martins@oracle.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Add a new test ioctl for simulating the dirty IOVAs in the mock domain, and
implement the mock iommu domain ops that get the dirty tracking supported.
The selftest exercises the usual main workflow of:
1) Setting dirty tracking from the iommu domain
2) Read and clear dirty IOPTEs
Different fixtures will test different IOVA range sizes, that exercise
corner cases of the bitmaps.
Link: https://lore.kernel.org/r/20231024135109.73787-17-joao.m.martins@oracle.com
Signed-off-by: Joao Martins <joao.m.martins@oracle.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
TEST_LENGTH passing ".size = sizeof(struct _struct) - 1" expects -EINVAL
from "if (ucmd.user_size < op->min_size)" check in iommufd_fops_ioctl().
This has been working when min_size is exactly the size of the structure.
However, if the size of the structure becomes larger than min_size, i.e.
the passing size above is larger than min_size, that min_size sanity no
longer works.
Since the first test in TEST_LENGTH() was to test that min_size sanity
routine, rework it to support a min_size calculation, rather than using
the full size of the structure.
Link: https://lore.kernel.org/r/20231015074648.24185-1-nicolinc@nvidia.com
Signed-off-by: Nicolin Chen <nicolinc@nvidia.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
