[ Upstream commit b306019848 ]
MLX cap pg_track_log_max_msg_size consists of 5 bits, value of which is
used as power of 2 for max_msg_size. This can lead to multiplication
overflow between max_msg_size (u32) and integer constant, and afterwards
incorrect value is being written to rq_size.
Fix this issue by extending integer constant to u64 type.
Found by Linux Verification Center (linuxtesting.org) with SVACE.
Suggested-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Artem Sadovnikov <a.sadovnikov@ispras.ru>
Reviewed-by: Yishai Hadas <yishaih@nvidia.com>
Link: https://lore.kernel.org/r/20250701144017.2410-2-a.sadovnikov@ispras.ru
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit b1779e4f20 ]
A large DMA mapping request can loop through dma address pinning for
many pages. In cases where THP can not be used, the repeated vmf_insert_pfn can
be costly, so let the task reschedule as need to prevent CPU stalls. Failure to
do so has potential harmful side effects, like increased memory pressure
as unrelated rcu tasks are unable to make their reclaim callbacks and
result in OOM conditions.
rcu: INFO: rcu_sched self-detected stall on CPU
rcu: 36-....: (20999 ticks this GP) idle=b01c/1/0x4000000000000000 softirq=35839/35839 fqs=3538
rcu: hardirqs softirqs csw/system
rcu: number: 0 107 0
rcu: cputime: 50 0 10446 ==> 10556(ms)
rcu: (t=21075 jiffies g=377761 q=204059 ncpus=384)
...
<TASK>
? asm_sysvec_apic_timer_interrupt+0x16/0x20
? walk_system_ram_range+0x63/0x120
? walk_system_ram_range+0x46/0x120
? pgprot_writethrough+0x20/0x20
lookup_memtype+0x67/0xf0
track_pfn_insert+0x20/0x40
vmf_insert_pfn_prot+0x88/0x140
vfio_pci_mmap_huge_fault+0xf9/0x1b0 [vfio_pci_core]
__do_fault+0x28/0x1b0
handle_mm_fault+0xef1/0x2560
fixup_user_fault+0xf5/0x270
vaddr_get_pfns+0x169/0x2f0 [vfio_iommu_type1]
vfio_pin_pages_remote+0x162/0x8e0 [vfio_iommu_type1]
vfio_iommu_type1_ioctl+0x1121/0x1810 [vfio_iommu_type1]
? futex_wake+0x1c1/0x260
x64_sys_call+0x234/0x17a0
do_syscall_64+0x63/0x130
? exc_page_fault+0x63/0x130
entry_SYSCALL_64_after_hwframe+0x4b/0x53
Signed-off-by: Keith Busch <kbusch@kernel.org>
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Link: https://lore.kernel.org/r/20250715184622.3561598-1-kbusch@meta.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit e908f58b6b ]
In the below noted Fixes commit we introduced a reflck mutex to allow
better scaling between devices for open and close. The reflck was
based on the hot reset granularity, device level for root bus devices
which cannot support hot reset or bus/slot reset otherwise. Overlooked
in this were SR-IOV VFs, where there's also no bus reset option, but
the default for a non-root-bus, non-slot-based device is bus level
reflck granularity.
The reflck mutex has since become the dev_set mutex (via commit
2cd8b14aaa ("vfio/pci: Move to the device set infrastructure")) and
is our defacto serialization for various operations and ioctls. It
still seems to be the case though that sets of vfio-pci devices really
only need serialization relative to hot resets affecting the entire
set, which is not relevant to SR-IOV VFs. As described in the Closes
link below, this serialization contributes to startup latency when
multiple VFs sharing the same "bus" are opened concurrently.
Mark the device itself as the basis of the dev_set for SR-IOV VFs.
Reported-by: Aaron Lewis <aaronlewis@google.com>
Closes: https://lore.kernel.org/all/20250626180424.632628-1-aaronlewis@google.com
Tested-by: Aaron Lewis <aaronlewis@google.com>
Fixes: e309df5b0c ("vfio/pci: Parallelize device open and release")
Reviewed-by: Yi Liu <yi.l.liu@intel.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Link: https://lore.kernel.org/r/20250626225623.1180952-1-alex.williamson@redhat.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 4518e5a60c ]
When setting up dirty page tracking at the vfio IOMMU backend for
device migration, if an error is encountered allocating a tracking
bitmap, the unwind loop fails to free previously allocated tracking
bitmaps. This occurs because the wrong loop index is used to
generate the tracking object. This results in unintended memory
usage for the life of the current DMA mappings where bitmaps were
successfully allocated.
Use the correct loop index to derive the tracking object for
freeing during unwind.
Fixes: d6a4c18566 ("vfio iommu: Implementation of ioctl for dirty pages tracking")
Signed-off-by: Li RongQing <lirongqing@baidu.com>
Link: https://lore.kernel.org/r/20250521034647.2877-1-lirongqing@baidu.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 3495cec078 ]
In order to ensure that the task packets of the accelerator
device are not lost during the migration process, it is necessary
to send an EQ and AEQ command to the device after the live migration
is completed and to update the completion position of the task queue.
Let the device recheck the completed tasks data and if there are
uncollected packets, device resend a task completion interrupt
to the software.
Fixes: b0eed08590 ("hisi_acc_vfio_pci: Add support for VFIO live migration")
Signed-off-by: Longfang Liu <liulongfang@huawei.com>
Reviewed-by: Shameer Kolothum <shameerali.kolothum.thodi@huawei.com>
Link: https://lore.kernel.org/r/20250510081155.55840-3-liulongfang@huawei.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 8bb7170c5a ]
The dma addresses of EQE and AEQE are wrong after migration and
results in guest kernel-mode encryption services failure.
Comparing the definition of hardware registers, we found that
there was an error when the data read from the register was
combined into an address. Therefore, the address combination
sequence needs to be corrected.
Even after fixing the above problem, we still have an issue
where the Guest from an old kernel can get migrated to
new kernel and may result in wrong data.
In order to ensure that the address is correct after migration,
if an old magic number is detected, the dma address needs to be
updated.
Fixes: b0eed08590 ("hisi_acc_vfio_pci: Add support for VFIO live migration")
Signed-off-by: Longfang Liu <liulongfang@huawei.com>
Reviewed-by: Shameer Kolothum <shameerali.kolothum.thodi@huawei.com>
Link: https://lore.kernel.org/r/20250510081155.55840-2-liulongfang@huawei.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 860be250fc ]
Some systems report INTx as not routed by setting pdev->irq to
IRQ_NOTCONNECTED, resulting in a -ENOTCONN error when trying to
setup eventfd signaling. Include this in the set of conditions
for which the PIN register is virtualized to zero.
Additionally consolidate vfio_pci_get_irq_count() to use this
virtualized value in reporting INTx support via ioctl and sanity
checking ioctl paths since pdev->irq is re-used when the device
is in MSI mode.
The combination of these results in both the config space of the
device and the ioctl interface behaving as if the device does not
support INTx.
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Link: https://lore.kernel.org/r/20250311230623.1264283-1-alex.williamson@redhat.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 2b938e3db3 ]
Definitions of ioread64 and iowrite64 macros in asm/io.h called by vfio
pci implementations are enclosed inside check for CONFIG_GENERIC_IOMAP.
They don't get defined if CONFIG_GENERIC_IOMAP is defined. Include
linux/io-64-nonatomic-lo-hi.h to define iowrite64 and ioread64 macros
when they are not defined. io-64-nonatomic-lo-hi.h maps the macros to
generic implementation in lib/iomap.c. The generic implementation does
64 bit rw if readq/writeq is defined for the architecture, otherwise it
would do 32 bit back to back rw.
Note that there are two versions of the generic implementation that
differs in the order the 32 bit words are written if 64 bit support is
not present. This is not the little/big endian ordering, which is
handled separately. This patch uses the lo followed by hi word ordering
which is consistent with current back to back implementation in the
vfio/pci code.
Signed-off-by: Ramesh Thomas <ramesh.thomas@intel.com>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Link: https://lore.kernel.org/r/20241210131938.303500-2-ramesh.thomas@intel.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit ce9ff21ea8 upstream.
count and offset are passed from user space and not checked, only
offset is capped to 40 bits, which can be used to read/write out of
bounds of the device.
Fixes: 6e3f264560 (“vfio/platform: read and write support for the device fd”)
Cc: stable@vger.kernel.org
Reported-by: Mostafa Saleh <smostafa@google.com>
Reviewed-by: Eric Auger <eric.auger@redhat.com>
Reviewed-by: Mostafa Saleh <smostafa@google.com>
Tested-by: Mostafa Saleh <smostafa@google.com>
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit e24c155105 ]
Resolve a UBSAN shift-out-of-bounds issue in iova_bitmap_offset_to_index()
where shifting the constant "1" (of type int) by bitmap->mapped.pgshift
(an unsigned long value) could result in undefined behavior.
The constant "1" defaults to a 32-bit "int", and when "pgshift" exceeds
31 (e.g., pgshift = 63) the shift operation overflows, as the result
cannot be represented in a 32-bit type.
To resolve this, the constant is updated to "1UL", promoting it to an
unsigned long type to match the operand's type.
Fixes: 58ccf0190d ("vfio: Add an IOVA bitmap support")
Link: https://patch.msgid.link/r/20250113223820.10713-1-qasdev00@gmail.com
Reported-by: syzbot <syzbot+85992ace37d5b7b51635@syzkaller.appspotmail.com>
Closes: https://syzkaller.appspot.com/bug?extid=85992ace37d5b7b51635
Signed-off-by: Qasim Ijaz <qasdev00@gmail.com>
Reviewed-by: Joao Martins <joao.m.martins@oracle.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit ce9ff21ea8 upstream.
count and offset are passed from user space and not checked, only
offset is capped to 40 bits, which can be used to read/write out of
bounds of the device.
Fixes: 6e3f264560 (“vfio/platform: read and write support for the device fd”)
Cc: stable@vger.kernel.org
Reported-by: Mostafa Saleh <smostafa@google.com>
Reviewed-by: Eric Auger <eric.auger@redhat.com>
Reviewed-by: Mostafa Saleh <smostafa@google.com>
Tested-by: Mostafa Saleh <smostafa@google.com>
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 9c7c5430bc ]
Align the page tracking maximum message size with the device's
capability instead of relying on PAGE_SIZE.
This adjustment resolves a mismatch on systems where PAGE_SIZE is 64K,
but the firmware only supports a maximum message size of 4K.
Now that we rely on the device's capability for max_message_size, we
must account for potential future increases in its value.
Key considerations include:
- Supporting message sizes that exceed a single system page (e.g., an 8K
message on a 4K system).
- Ensuring the RQ size is adjusted to accommodate at least 4
WQEs/messages, in line with the device specification.
The above has been addressed as part of the patch.
Fixes: 79c3cf2799 ("vfio/mlx5: Init QP based resources for dirty tracking")
Reviewed-by: Cédric Le Goater <clg@redhat.com>
Tested-by: Yingshun Cui <yicui@redhat.com>
Signed-off-by: Yishai Hadas <yishaih@nvidia.com>
Link: https://lore.kernel.org/r/20241205122654.235619-1-yishaih@nvidia.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit fe4bf8d0b6 ]
There are cases where a PCIe extended capability should be hidden from
the user. For example, an unknown capability (i.e., capability with ID
greater than PCI_EXT_CAP_ID_MAX) or a capability that is intentionally
chosen to be hidden from the user.
Hiding a capability is done by virtualizing and modifying the 'Next
Capability Offset' field of the previous capability so it points to the
capability after the one that should be hidden.
The special case where the first capability in the list should be hidden
is handled differently because there is no previous capability that can
be modified. In this case, the capability ID and version are zeroed
while leaving the next pointer intact. This hides the capability and
leaves an anchor for the rest of the capability list.
However, today, hiding the first capability in the list is not done
properly if the capability is unknown, as struct
vfio_pci_core_device->pci_config_map is set to the capability ID during
initialization but the capability ID is not properly checked later when
used in vfio_config_do_rw(). This leads to the following warning [1] and
to an out-of-bounds access to ecap_perms array.
Fix it by checking cap_id in vfio_config_do_rw(), and if it is greater
than PCI_EXT_CAP_ID_MAX, use an alternative struct perm_bits for direct
read only access instead of the ecap_perms array.
Note that this is safe since the above is the only case where cap_id can
exceed PCI_EXT_CAP_ID_MAX (except for the special capabilities, which
are already checked before).
[1]
WARNING: CPU: 118 PID: 5329 at drivers/vfio/pci/vfio_pci_config.c:1900 vfio_pci_config_rw+0x395/0x430 [vfio_pci_core]
CPU: 118 UID: 0 PID: 5329 Comm: simx-qemu-syste Not tainted 6.12.0+ #1
(snip)
Call Trace:
<TASK>
? show_regs+0x69/0x80
? __warn+0x8d/0x140
? vfio_pci_config_rw+0x395/0x430 [vfio_pci_core]
? report_bug+0x18f/0x1a0
? handle_bug+0x63/0xa0
? exc_invalid_op+0x19/0x70
? asm_exc_invalid_op+0x1b/0x20
? vfio_pci_config_rw+0x395/0x430 [vfio_pci_core]
? vfio_pci_config_rw+0x244/0x430 [vfio_pci_core]
vfio_pci_rw+0x101/0x1b0 [vfio_pci_core]
vfio_pci_core_read+0x1d/0x30 [vfio_pci_core]
vfio_device_fops_read+0x27/0x40 [vfio]
vfs_read+0xbd/0x340
? vfio_device_fops_unl_ioctl+0xbb/0x740 [vfio]
? __rseq_handle_notify_resume+0xa4/0x4b0
__x64_sys_pread64+0x96/0xc0
x64_sys_call+0x1c3d/0x20d0
do_syscall_64+0x4d/0x120
entry_SYSCALL_64_after_hwframe+0x76/0x7e
Fixes: 89e1f7d4c6 ("vfio: Add PCI device driver")
Signed-off-by: Avihai Horon <avihaih@nvidia.com>
Reviewed-by: Yi Liu <yi.l.liu@intel.com>
Tested-by: Yi Liu <yi.l.liu@intel.com>
Link: https://lore.kernel.org/r/20241124142739.21698-1-avihaih@nvidia.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit 95feb3160e upstream.
Due to an erratum with the SPR_DSA and SPR_IAX devices, it is not secure to assign
these devices to virtual machines. Add the PCI IDs of these devices to the VFIO
denylist to ensure that this is handled appropriately by the VFIO subsystem.
The SPR_DSA and SPR_IAX devices are on-SOC devices for the Sapphire Rapids
(and related) family of products that perform data movement and compression.
Signed-off-by: Arjan van de Ven <arjan@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 7447d911af ]
The eventfd_ctx trigger pointer of the vfio_fsl_mc_irq object is
initially NULL and may become NULL if the user sets the trigger
eventfd to -1. The interrupt handler itself is guaranteed that
trigger is always valid between request_irq() and free_irq(), but
the loopback testing mechanisms to invoke the handler function
need to test the trigger. The triggering and setting ioctl paths
both make use of igate and are therefore mutually exclusive.
The vfio-fsl-mc driver does not make use of irqfds, nor does it
support any sort of masking operations, therefore unlike vfio-pci
and vfio-platform, the flow can remain essentially unchanged.
Cc: Diana Craciun <diana.craciun@oss.nxp.com>
Cc: <stable@vger.kernel.org>
Fixes: cc0ee20bd9 ("vfio/fsl-mc: trigger an interrupt via eventfd")
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Eric Auger <eric.auger@redhat.com>
Link: https://lore.kernel.org/r/20240308230557.805580-8-alex.williamson@redhat.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 675daf435e ]
The vfio-platform SET_IRQS ioctl currently allows loopback triggering of
an interrupt before a signaling eventfd has been configured by the user,
which thereby allows a NULL pointer dereference.
Rather than register the IRQ relative to a valid trigger, register all
IRQs in a disabled state in the device open path. This allows mask
operations on the IRQ to nest within the overall enable state governed
by a valid eventfd signal. This decouples @masked, protected by the
@locked spinlock from @trigger, protected via the @igate mutex.
In doing so, it's guaranteed that changes to @trigger cannot race the
IRQ handlers because the IRQ handler is synchronously disabled before
modifying the trigger, and loopback triggering of the IRQ via ioctl is
safe due to serialization with trigger changes via igate.
For compatibility, request_irq() failures are maintained to be local to
the SET_IRQS ioctl rather than a fatal error in the open device path.
This allows, for example, a userspace driver with polling mode support
to continue to work regardless of moving the request_irq() call site.
This necessarily blocks all SET_IRQS access to the failed index.
Cc: Eric Auger <eric.auger@redhat.com>
Cc: <stable@vger.kernel.org>
Fixes: 57f972e2b3 ("vfio/platform: trigger an interrupt via eventfd")
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Eric Auger <eric.auger@redhat.com>
Link: https://lore.kernel.org/r/20240308230557.805580-7-alex.williamson@redhat.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 18c198c96a ]
A vulnerability exists where the eventfd for INTx signaling can be
deconfigured, which unregisters the IRQ handler but still allows
eventfds to be signaled with a NULL context through the SET_IRQS ioctl
or through unmask irqfd if the device interrupt is pending.
Ideally this could be solved with some additional locking; the igate
mutex serializes the ioctl and config space accesses, and the interrupt
handler is unregistered relative to the trigger, but the irqfd path
runs asynchronous to those. The igate mutex cannot be acquired from the
atomic context of the eventfd wake function. Disabling the irqfd
relative to the eventfd registration is potentially incompatible with
existing userspace.
As a result, the solution implemented here moves configuration of the
INTx interrupt handler to track the lifetime of the INTx context object
and irq_type configuration, rather than registration of a particular
trigger eventfd. Synchronization is added between the ioctl path and
eventfd_signal() wrapper such that the eventfd trigger can be
dynamically updated relative to in-flight interrupts or irqfd callbacks.
Cc: <stable@vger.kernel.org>
Fixes: 89e1f7d4c6 ("vfio: Add PCI device driver")
Reported-by: Reinette Chatre <reinette.chatre@intel.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Reinette Chatre <reinette.chatre@intel.com>
Reviewed-by: Eric Auger <eric.auger@redhat.com>
Link: https://lore.kernel.org/r/20240308230557.805580-5-alex.williamson@redhat.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit b620ecbd17 ]
In order to synchronize changes that can affect the thread callback,
introduce an interface to force a flush of the inject workqueue. The
irqfd pointer is only valid under spinlock, but the workqueue cannot
be flushed under spinlock. Therefore the flush work for the irqfd is
queued under spinlock. The vfio_irqfd_cleanup_wq workqueue is re-used
for queuing this work such that flushing the workqueue is also ordered
relative to shutdown.
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Reinette Chatre <reinette.chatre@intel.com>
Reviewed-by: Eric Auger <eric.auger@redhat.com>
Link: https://lore.kernel.org/r/20240308230557.805580-4-alex.williamson@redhat.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit fe9a708268 ]
Currently for devices requiring masking at the irqchip for INTx, ie.
devices without DisINTx support, the IRQ is enabled in request_irq()
and subsequently disabled as necessary to align with the masked status
flag. This presents a window where the interrupt could fire between
these events, resulting in the IRQ incrementing the disable depth twice.
This would be unrecoverable for a user since the masked flag prevents
nested enables through vfio.
Instead, invert the logic using IRQF_NO_AUTOEN such that exclusive INTx
is never auto-enabled, then unmask as required.
Cc: <stable@vger.kernel.org>
Fixes: 89e1f7d4c6 ("vfio: Add PCI device driver")
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Eric Auger <eric.auger@redhat.com>
Link: https://lore.kernel.org/r/20240308230557.805580-2-alex.williamson@redhat.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit fcdc0d3d40 ]
irqfds for mask and unmask that are not specifically disabled by the
user are leaked. Remove any irqfds during cleanup
Cc: Eric Auger <eric.auger@redhat.com>
Cc: <stable@vger.kernel.org>
Fixes: a7fa7c77cf ("vfio/platform: implement IRQ masking/unmasking via an eventfd")
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Eric Auger <eric.auger@redhat.com>
Link: https://lore.kernel.org/r/20240308230557.805580-6-alex.williamson@redhat.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 810cd4bb53 ]
Mask operations through config space changes to DisINTx may race INTx
configuration changes via ioctl. Create wrappers that add locking for
paths outside of the core interrupt code.
In particular, irq_type is updated holding igate, therefore testing
is_intx() requires holding igate. For example clearing DisINTx from
config space can otherwise race changes of the interrupt configuration.
This aligns interfaces which may trigger the INTx eventfd into two
camps, one side serialized by igate and the other only enabled while
INTx is configured. A subsequent patch introduces synchronization for
the latter flows.
Cc: <stable@vger.kernel.org>
Fixes: 89e1f7d4c6 ("vfio: Add PCI device driver")
Reported-by: Reinette Chatre <reinette.chatre@intel.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Reinette Chatre <reinette.chatre@intel.com>
Reviewed-by: Eric Auger <eric.auger@redhat.com>
Link: https://lore.kernel.org/r/20240308230557.805580-3-alex.williamson@redhat.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 6578ed85c7 ]
User space provides the vector as an unsigned int that is checked
early for validity (vfio_set_irqs_validate_and_prepare()).
A later negative check of the provided vector is not necessary.
Remove the negative check and ensure the type used
for the vector is consistent as an unsigned int.
Signed-off-by: Reinette Chatre <reinette.chatre@intel.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Acked-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Link: https://lore.kernel.org/r/28521e1b0b091849952b0ecb8c118729fc8cdc4f.1683740667.git.reinette.chatre@intel.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Stable-dep-of: fe9a708268 ("vfio/pci: Disable auto-enable of exclusive INTx IRQ")
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit a65f35cfd5 ]
vfio_msi_disable() releases all previously allocated state
associated with each interrupt before disabling MSI/MSI-X.
vfio_msi_disable() iterates twice over the interrupt state:
first directly with a for loop to do virqfd cleanup, followed
by another for loop within vfio_msi_set_block() that removes
the interrupt handler and its associated state using
vfio_msi_set_vector_signal().
Simplify interrupt cleanup by iterating over allocated interrupts
once.
Signed-off-by: Reinette Chatre <reinette.chatre@intel.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Acked-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Link: https://lore.kernel.org/r/837acb8cbe86a258a50da05e56a1f17c1a19abbe.1683740667.git.reinette.chatre@intel.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Stable-dep-of: fe9a708268 ("vfio/pci: Disable auto-enable of exclusive INTx IRQ")
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 0886196ca8 ]
Use GFP_KERNEL_ACCOUNT for userspace persistent allocations.
The GFP_KERNEL_ACCOUNT option lets the memory allocator know that this
is untrusted allocation triggered from userspace and should be a subject
of kmem accounting, and as such it is controlled by the cgroup
mechanism.
The way to find the relevant allocations was for example to look at the
close_device function and trace back all the kfrees to their
allocations.
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Yishai Hadas <yishaih@nvidia.com>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Link: https://lore.kernel.org/r/20230108154427.32609-4-yishaih@nvidia.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Stable-dep-of: fe9a708268 ("vfio/pci: Disable auto-enable of exclusive INTx IRQ")
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 4bbcbc6ea2 ]
For small bitmaps that aren't PAGE_SIZE aligned *and* that are less than
512 pages in bitmap length, use an extra page to be able to cover the
entire range e.g. [1M..3G] which would be iterated more efficiently in a
single iteration, rather than two.
Fixes: b058ea3ab5 ("vfio/iova_bitmap: refactor iova_bitmap_set() to better handle page boundaries")
Link: https://lore.kernel.org/r/20240202133415.23819-10-joao.m.martins@oracle.com
Signed-off-by: Joao Martins <joao.m.martins@oracle.com>
Tested-by: Avihai Horon <avihaih@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit d18411ec30 ]
iova_bitmap_mapped_length() don't deal correctly with the small bitmaps
(< 2M bitmaps) when the starting address isn't u64 aligned, leading to
skipping a tiny part of the IOVA range. This is materialized as not
marking data dirty that should otherwise have been.
Fix that by using a u8 * in the internal state of IOVA bitmap. Most of the
data structures use the type of the bitmap to adjust its indexes, thus
changing the type of the bitmap decreases the granularity of the bitmap
indexes.
Fixes: b058ea3ab5 ("vfio/iova_bitmap: refactor iova_bitmap_set() to better handle page boundaries")
Link: https://lore.kernel.org/r/20240202133415.23819-3-joao.m.martins@oracle.com
Signed-off-by: Joao Martins <joao.m.martins@oracle.com>
Tested-by: Avihai Horon <avihaih@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit a4ab7dedae ]
Dirty IOMMU hugepages reported on a base page page-size granularity can
lead to an attempt to set dirty pages in the bitmap beyond the limits that
are pinned.
Bounds check the page index of the array we are trying to access is within
the limits before we kmap() and return otherwise.
While it is also a defensive check, this is also in preparation to defer
setting bits (outside the mapped range) to the next iteration(s) when the
pages become available.
Fixes: b058ea3ab5 ("vfio/iova_bitmap: refactor iova_bitmap_set() to better handle page boundaries")
Link: https://lore.kernel.org/r/20240202133415.23819-2-joao.m.martins@oracle.com
Signed-off-by: Joao Martins <joao.m.martins@oracle.com>
Tested-by: Avihai Horon <avihaih@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit cd24e2a60a ]
Fix an information leak where an uninitialized hole in struct
vfio_iommu_type1_info_cap_migration on the stack is exposed to userspace.
The definition of struct vfio_iommu_type1_info_cap_migration contains a hole as
shown in this pahole(1) output:
struct vfio_iommu_type1_info_cap_migration {
struct vfio_info_cap_header header; /* 0 8 */
__u32 flags; /* 8 4 */
/* XXX 4 bytes hole, try to pack */
__u64 pgsize_bitmap; /* 16 8 */
__u64 max_dirty_bitmap_size; /* 24 8 */
/* size: 32, cachelines: 1, members: 4 */
/* sum members: 28, holes: 1, sum holes: 4 */
/* last cacheline: 32 bytes */
};
The cap_mig variable is filled in without initializing the hole:
static int vfio_iommu_migration_build_caps(struct vfio_iommu *iommu,
struct vfio_info_cap *caps)
{
struct vfio_iommu_type1_info_cap_migration cap_mig;
cap_mig.header.id = VFIO_IOMMU_TYPE1_INFO_CAP_MIGRATION;
cap_mig.header.version = 1;
cap_mig.flags = 0;
/* support minimum pgsize */
cap_mig.pgsize_bitmap = (size_t)1 << __ffs(iommu->pgsize_bitmap);
cap_mig.max_dirty_bitmap_size = DIRTY_BITMAP_SIZE_MAX;
return vfio_info_add_capability(caps, &cap_mig.header, sizeof(cap_mig));
}
The structure is then copied to a temporary location on the heap. At this point
it's already too late and ioctl(VFIO_IOMMU_GET_INFO) copies it to userspace
later:
int vfio_info_add_capability(struct vfio_info_cap *caps,
struct vfio_info_cap_header *cap, size_t size)
{
struct vfio_info_cap_header *header;
header = vfio_info_cap_add(caps, size, cap->id, cap->version);
if (IS_ERR(header))
return PTR_ERR(header);
memcpy(header + 1, cap + 1, size - sizeof(*header));
return 0;
}
This issue was found by code inspection.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Fixes: ad721705d0 ("vfio iommu: Add migration capability to report supported features")
Link: https://lore.kernel.org/r/20230801155352.1391945-1-stefanha@redhat.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit ff598081e5 ]
The pointer to mdev_bus_compat_class is statically defined at the top
of mdev_core, and was originally (commit 7b96953bc6 ("vfio: Mediated
device Core driver") serialized by the parent_list_lock. The blamed
commit removed this mutex, leaving the pointer initialization
unserialized. As a result, the creation of multiple MDEVs in parallel
(such as during boot) can encounter errors during the creation of the
sysfs entries, such as:
[ 8.337509] sysfs: cannot create duplicate filename '/class/mdev_bus'
[ 8.337514] vfio_ccw 0.0.01d8: MDEV: Registered
[ 8.337516] CPU: 13 PID: 946 Comm: driverctl Not tainted 6.4.0-rc7 #20
[ 8.337522] Hardware name: IBM 3906 M05 780 (LPAR)
[ 8.337525] Call Trace:
[ 8.337528] [<0000000162b0145a>] dump_stack_lvl+0x62/0x80
[ 8.337540] [<00000001622aeb30>] sysfs_warn_dup+0x78/0x88
[ 8.337549] [<00000001622aeca6>] sysfs_create_dir_ns+0xe6/0xf8
[ 8.337552] [<0000000162b04504>] kobject_add_internal+0xf4/0x340
[ 8.337557] [<0000000162b04d48>] kobject_add+0x78/0xd0
[ 8.337561] [<0000000162b04e0a>] kobject_create_and_add+0x6a/0xb8
[ 8.337565] [<00000001627a110e>] class_compat_register+0x5e/0x90
[ 8.337572] [<000003ff7fd815da>] mdev_register_parent+0x102/0x130 [mdev]
[ 8.337581] [<000003ff7fdc7f2c>] vfio_ccw_sch_probe+0xe4/0x178 [vfio_ccw]
[ 8.337588] [<0000000162a7833c>] css_probe+0x44/0x80
[ 8.337599] [<000000016279f4da>] really_probe+0xd2/0x460
[ 8.337603] [<000000016279fa08>] driver_probe_device+0x40/0xf0
[ 8.337606] [<000000016279fb78>] __device_attach_driver+0xc0/0x140
[ 8.337610] [<000000016279cbe0>] bus_for_each_drv+0x90/0xd8
[ 8.337618] [<00000001627a00b0>] __device_attach+0x110/0x190
[ 8.337621] [<000000016279c7c8>] bus_rescan_devices_helper+0x60/0xb0
[ 8.337626] [<000000016279cd48>] drivers_probe_store+0x48/0x80
[ 8.337632] [<00000001622ac9b0>] kernfs_fop_write_iter+0x138/0x1f0
[ 8.337635] [<00000001621e5e14>] vfs_write+0x1ac/0x2f8
[ 8.337645] [<00000001621e61d8>] ksys_write+0x70/0x100
[ 8.337650] [<0000000162b2bdc4>] __do_syscall+0x1d4/0x200
[ 8.337656] [<0000000162b3c828>] system_call+0x70/0x98
[ 8.337664] kobject: kobject_add_internal failed for mdev_bus with -EEXIST, don't try to register things with the same name in the same directory.
[ 8.337668] kobject: kobject_create_and_add: kobject_add error: -17
[ 8.337674] vfio_ccw: probe of 0.0.01d9 failed with error -12
[ 8.342941] vfio_ccw_mdev aeb9ca91-10c6-42bc-a168-320023570aea: Adding to iommu group 2
Move the initialization of the mdev_bus_compat_class pointer to the
init path, to match the cleanup in module exit. This way the code
in mdev_register_parent() can simply link the new parent to it,
rather than determining whether initialization is required first.
Fixes: 89345d5177 ("vfio/mdev: embedd struct mdev_parent in the parent data structure")
Reported-by: Alexander Egorenkov <egorenar@linux.ibm.com>
Signed-off-by: Eric Farman <farman@linux.ibm.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Tony Krowiak <akrowiak@linux.ibm.com>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Link: https://lore.kernel.org/r/20230626133642.2939168-1-farman@linux.ibm.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 4752354af7 ]
Check physical PFN is valid before converting the PFN to a struct page
pointer to be returned to caller of vfio_pin_pages().
vfio_pin_pages() pins user pages with contiguous IOVA.
If the IOVA of a user page to be pinned belongs to vma of vm_flags
VM_PFNMAP, pin_user_pages_remote() will return -EFAULT without returning
struct page address for this PFN. This is because usually this kind of PFN
(e.g. MMIO PFN) has no valid struct page address associated.
Upon this error, vaddr_get_pfns() will obtain the physical PFN directly.
While previously vfio_pin_pages() returns to caller PFN arrays directly,
after commit
34a255e676 ("vfio: Replace phys_pfn with pages for vfio_pin_pages()"),
PFNs will be converted to "struct page *" unconditionally and therefore
the returned "struct page *" array may contain invalid struct page
addresses.
Given current in-tree users of vfio_pin_pages() only expect "struct page *
returned, check PFN validity and return -EINVAL to let the caller be
aware of IOVAs to be pinned containing PFN not able to be returned in
"struct page *" array. So that, the caller will not consume the returned
pointer (e.g. test PageReserved()) and avoid error like "supervisor read
access in kernel mode".
Fixes: 34a255e676 ("vfio: Replace phys_pfn with pages for vfio_pin_pages()")
Cc: Sean Christopherson <seanjc@google.com>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Yan Zhao <yan.y.zhao@intel.com>
Reviewed-by: Sean Christopherson <seanjc@google.com>
Link: https://lore.kernel.org/r/20230519065843.10653-1-yan.y.zhao@intel.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit 90fdd158a6 upstream.
When a vfio container is preserved across exec or fork-exec, the new
task's mm has a locked_vm count of 0. After a dma vaddr is updated using
VFIO_DMA_MAP_FLAG_VADDR, locked_vm remains 0, and the pinned memory does
not count against the task's RLIMIT_MEMLOCK.
To restore the correct locked_vm count, when VFIO_DMA_MAP_FLAG_VADDR is
used and the dma's mm has changed, add the dma's locked_vm count to
the new mm->locked_vm, subject to the rlimit, and subtract it from the
old mm->locked_vm.
Fixes: c3cbab24db ("vfio/type1: implement interfaces to update vaddr")
Cc: stable@vger.kernel.org
Signed-off-by: Steve Sistare <steven.sistare@oracle.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Link: https://lore.kernel.org/r/1675184289-267876-5-git-send-email-steven.sistare@oracle.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 18e292705b upstream.
Track locked_vm per dma struct, and create a new subroutine, both for use
in a subsequent patch. No functional change.
Fixes: c3cbab24db ("vfio/type1: implement interfaces to update vaddr")
Cc: stable@vger.kernel.org
Signed-off-by: Steve Sistare <steven.sistare@oracle.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Link: https://lore.kernel.org/r/1675184289-267876-4-git-send-email-steven.sistare@oracle.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 046eca5018 upstream.
When a vfio container is preserved across exec, the task does not change,
but it gets a new mm with locked_vm=0, and loses the count from existing
dma mappings. If the user later unmaps a dma mapping, locked_vm underflows
to a large unsigned value, and a subsequent dma map request fails with
ENOMEM in __account_locked_vm.
To avoid underflow, grab and save the mm at the time a dma is mapped.
Use that mm when adjusting locked_vm, rather than re-acquiring the saved
task's mm, which may have changed. If the saved mm is dead, do nothing.
locked_vm is incremented for existing mappings in a subsequent patch.
Fixes: 73fa0d10d0 ("vfio: Type1 IOMMU implementation")
Cc: stable@vger.kernel.org
Signed-off-by: Steve Sistare <steven.sistare@oracle.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Link: https://lore.kernel.org/r/1675184289-267876-3-git-send-email-steven.sistare@oracle.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit ef3a3f6a29 upstream.
Disable the VFIO_UPDATE_VADDR capability if mediated devices are present.
Their kernel threads could be blocked indefinitely by a misbehaving
userland while trying to pin/unpin pages while vaddrs are being updated.
Do not allow groups to be added to the container while vaddr's are invalid,
so we never need to block user threads from pinning, and can delete the
vaddr-waiting code in a subsequent patch.
Fixes: c3cbab24db ("vfio/type1: implement interfaces to update vaddr")
Cc: stable@vger.kernel.org
Signed-off-by: Steve Sistare <steven.sistare@oracle.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Link: https://lore.kernel.org/r/1675184289-267876-2-git-send-email-steven.sistare@oracle.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 895c0747f7 ]
Since commit cbf7827bc5 ("iommu/s390: Fix potential s390_domain
aperture shrinking") the s390 IOMMU driver uses reserved regions for the
system provided DMA ranges of PCI devices. Previously it reduced the
size of the IOMMU aperture and checked it on each mapping operation.
On current machines the system denies use of DMA addresses below 2^32 for
all PCI devices.
Usually mapping IOVAs in a reserved regions is harmless until a DMA
actually tries to utilize the mapping. However on s390 there is
a virtual PCI device called ISM which is implemented in firmware and
used for cross LPAR communication. Unlike real PCI devices this device
does not use the hardware IOMMU but inspects IOMMU translation tables
directly on IOTLB flush (s390 RPCIT instruction). If it detects IOVA
mappings outside the allowed ranges it goes into an error state. This
error state then causes the device to be unavailable to the KVM guest.
Analysing this we found that vfio_test_domain_fgsp() maps 2 pages at DMA
address 0 irrespective of the IOMMUs reserved regions. Even if usually
harmless this seems wrong in the general case so instead go through the
freshly updated IOVA list and try to find a range that isn't reserved,
and fits 2 pages, is PAGE_SIZE * 2 aligned. If found use that for
testing for fine grained super pages.
Fixes: af029169b8 ("vfio/type1: Check reserved region conflict and update iova list")
Signed-off-by: Niklas Schnelle <schnelle@linux.ibm.com>
Reviewed-by: Matthew Rosato <mjrosato@linux.ibm.com>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Link: https://lore.kernel.org/r/20230110164427.4051938-2-schnelle@linux.ibm.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit b058ea3ab5 ]
Commit f38044e5ef ("vfio/iova_bitmap: Fix PAGE_SIZE unaligned bitmaps")
had fixed the unaligned bitmaps by capping the remaining iterable set at
the start of the bitmap. Although, that mistakenly worked around
iova_bitmap_set() incorrectly setting bits across page boundary.
Fix this by reworking the loop inside iova_bitmap_set() to iterate over a
range of bits to set (cur_bit .. last_bit) which may span different pinned
pages, thus updating @page_idx and @offset as it sets the bits. The
previous cap to the first page is now adjusted to be always accounted
rather than when there's only a non-zero pgoff.
While at it, make @page_idx , @offset and @nbits to be unsigned int given
that it won't be more than 512 and 4096 respectively (even a bigger
PAGE_SIZE or a smaller struct page size won't make this bigger than the
above 32-bit max). Also, delete the stale kdoc on Return type.
Cc: Avihai Horon <avihaih@nvidia.com>
Fixes: f38044e5ef ("vfio/iova_bitmap: Fix PAGE_SIZE unaligned bitmaps")
Co-developed-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Joao Martins <joao.m.martins@oracle.com>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Tested-by: Avihai Horon <avihaih@nvidia.com>
Link: https://lore.kernel.org/r/20221129131235.38880-1-joao.m.martins@oracle.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit f38044e5ef ]
iova_bitmap_set() doesn't consider the end of the page boundary when the
first bitmap page offset isn't zero, and wrongly changes the consecutive
page right after. Consequently this leads to missing dirty pages from
reported by the device as seen from the VMM.
The current logic iterates over a given number of base pages and clamps it
to the remaining indexes to iterate in the last page. Instead of having to
consider extra pages to pin (e.g. first and extra pages), just handle the
first page as its own range and let the rest of the bitmap be handled as if
it was base page aligned.
This is done by changing iova_bitmap_mapped_remaining() to return PAGE_SIZE
- pgoff (on the first bitmap page), and leads to pgoff being set to 0 on
following iterations.
Fixes: 58ccf0190d ("vfio: Add an IOVA bitmap support")
Reported-by: Avihai Horon <avihaih@nvidia.com>
Tested-by: Avihai Horon <avihaih@nvidia.com>
Signed-off-by: Joao Martins <joao.m.martins@oracle.com>
Link: https://lore.kernel.org/r/20221025193114.58695-3-joao.m.martins@oracle.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit e67e070632 ]
The ACPI _RST method has no return value, there's no need to pass a return
buffer to acpi_evaluate_object().
Fixes: d30daa33ec ("vfio: platform: call _RST method when using ACPI")
Signed-off-by: Rafael Mendonca <rafaelmendsr@gmail.com>
Reviewed-by: Eric Auger <eric.auger@redhat.com>
Link: https://lore.kernel.org/r/20221018152825.891032-1-rafaelmendsr@gmail.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
vfio_pci_dev_set_needs_reset() inspects the open_count of every device
in the set to determine whether a reset is allowed. The current device
always has open_count == 1 within vfio_pci_core_disable(), effectively
disabling the reset logic. This field is also documented as private in
vfio_device, so it should not be used to determine whether other devices
in the set are open.
Checking for vfio_device_set_open_count() > 1 on the device set fixes
both issues.
After commit 2cd8b14aaa ("vfio/pci: Move to the device set
infrastructure"), failure to create a new file for a device would cause
the reset to be skipped due to open_count being decremented after
calling close_device() in the error path.
After commit eadd86f835 ("vfio: Remove calls to
vfio_group_add_container_user()"), releasing a device would always skip
the reset due to an ordering change in vfio_device_fops_release().
Failing to reset the device leaves it in an unknown state, potentially
causing errors when it is accessed later or bound to a different driver.
This issue was observed with a Radeon RX Vega 56 [1002:687f] (rev c3)
assigned to a Windows guest. After shutting down the guest, unbinding
the device from vfio-pci, and binding the device to amdgpu:
[ 548.007102] [drm:psp_hw_start [amdgpu]] *ERROR* PSP create ring failed!
[ 548.027174] [drm:psp_hw_init [amdgpu]] *ERROR* PSP firmware loading failed
[ 548.027242] [drm:amdgpu_device_fw_loading [amdgpu]] *ERROR* hw_init of IP block <psp> failed -22
[ 548.027306] amdgpu 0000:0a:00.0: amdgpu: amdgpu_device_ip_init failed
[ 548.027308] amdgpu 0000:0a:00.0: amdgpu: Fatal error during GPU init
Fixes: 2cd8b14aaa ("vfio/pci: Move to the device set infrastructure")
Fixes: eadd86f835 ("vfio: Remove calls to vfio_group_add_container_user()")
Signed-off-by: Anthony DeRossi <ajderossi@gmail.com>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Link: https://lore.kernel.org/r/20221110014027.28780-4-ajderossi@gmail.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
The open count of a device set is the sum of the open counts of all
devices in the set. Drivers can use this value to determine whether
shared resources are in use without tracking them manually or accessing
the private open_count in vfio_device.
Signed-off-by: Anthony DeRossi <ajderossi@gmail.com>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Yi Liu <yi.l.liu@intel.com>
Link: https://lore.kernel.org/r/20221110014027.28780-3-ajderossi@gmail.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
In vfio_device_open(), vfio_device_container_register() is always called
when open_count == 1. On error, vfio_device_container_unregister() is
only called when open_count == 1 and close_device is set. This leaks a
registration for devices without a close_device implementation.
In vfio_device_fops_release(), vfio_device_container_unregister() is
called unconditionally. This can cause a device to be unregistered
multiple times.
Treating container device registration/unregistration uniformly (always
when open_count == 1) fixes both issues.
Fixes: ce4b4657ff ("vfio: Replace the DMA unmapping notifier with a callback")
Signed-off-by: Anthony DeRossi <ajderossi@gmail.com>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Yi Liu <yi.l.liu@intel.com>
Link: https://lore.kernel.org/r/20221110014027.28780-2-ajderossi@gmail.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
- Prune private items from vfio_pci_core.h to a new internal header,
fix missed function rename, and refactor vfio-pci interrupt defines.
(Jason Gunthorpe)
- Create consistent naming and handling of ioctls with a function per
ioctl for vfio-pci and vfio group handling, use proper type args
where available. (Jason Gunthorpe)
- Implement a set of low power device feature ioctls allowing userspace
to make use of power states such as D3cold where supported.
(Abhishek Sahu)
- Remove device counter on vfio groups, which had restricted the page
pinning interface to singleton groups to account for limitations in
the type1 IOMMU backend. Document usage as limited to emulated IOMMU
devices, ie. traditional mdev devices where this restriction is
consistent. (Jason Gunthorpe)
- Correct function prefix in hisi_acc driver incurred during previous
refactoring. (Shameer Kolothum)
- Correct typo and remove redundant warning triggers in vfio-fsl driver.
(Christophe JAILLET)
- Introduce device level DMA dirty tracking uAPI and implementation in
the mlx5 variant driver (Yishai Hadas & Joao Martins)
- Move much of the vfio_device life cycle management into vfio core,
simplifying and avoiding duplication across drivers. This also
facilitates adding a struct device to vfio_device which begins the
introduction of device rather than group level user support and fills
a gap allowing userspace identify devices as vfio capable without
implicit knowledge of the driver. (Kevin Tian & Yi Liu)
- Split vfio container handling to a separate file, creating a more
well defined API between the core and container code, masking IOMMU
backend implementation from the core, allowing for an easier future
transition to an iommufd based implementation of the same.
(Jason Gunthorpe)
- Attempt to resolve race accessing the iommu_group for a device
between vfio releasing DMA ownership and removal of the device from
the IOMMU driver. Follow-up with support to allow vfio_group to
exist with NULL iommu_group pointer to support existing userspace
use cases of holding the group file open. (Jason Gunthorpe)
- Fix error code and hi/lo register manipulation issues in the hisi_acc
variant driver, along with various code cleanups. (Longfang Liu)
- Fix a prior regression in GVT-g group teardown, resulting in
unreleased resources. (Jason Gunthorpe)
- A significant cleanup and simplification of the mdev interface,
consolidating much of the open coded per driver sysfs interface
support into the mdev core. (Christoph Hellwig)
- Simplification of tracking and locking around vfio_groups that
fall out from previous refactoring. (Jason Gunthorpe)
- Replace trivial open coded f_ops tests with new helper.
(Alex Williamson)
-----BEGIN PGP SIGNATURE-----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=Z8+P
-----END PGP SIGNATURE-----
Merge tag 'vfio-v6.1-rc1' of https://github.com/awilliam/linux-vfio
Pull VFIO updates from Alex Williamson:
- Prune private items from vfio_pci_core.h to a new internal header,
fix missed function rename, and refactor vfio-pci interrupt defines
(Jason Gunthorpe)
- Create consistent naming and handling of ioctls with a function per
ioctl for vfio-pci and vfio group handling, use proper type args
where available (Jason Gunthorpe)
- Implement a set of low power device feature ioctls allowing userspace
to make use of power states such as D3cold where supported (Abhishek
Sahu)
- Remove device counter on vfio groups, which had restricted the page
pinning interface to singleton groups to account for limitations in
the type1 IOMMU backend. Document usage as limited to emulated IOMMU
devices, ie. traditional mdev devices where this restriction is
consistent (Jason Gunthorpe)
- Correct function prefix in hisi_acc driver incurred during previous
refactoring (Shameer Kolothum)
- Correct typo and remove redundant warning triggers in vfio-fsl driver
(Christophe JAILLET)
- Introduce device level DMA dirty tracking uAPI and implementation in
the mlx5 variant driver (Yishai Hadas & Joao Martins)
- Move much of the vfio_device life cycle management into vfio core,
simplifying and avoiding duplication across drivers. This also
facilitates adding a struct device to vfio_device which begins the
introduction of device rather than group level user support and fills
a gap allowing userspace identify devices as vfio capable without
implicit knowledge of the driver (Kevin Tian & Yi Liu)
- Split vfio container handling to a separate file, creating a more
well defined API between the core and container code, masking IOMMU
backend implementation from the core, allowing for an easier future
transition to an iommufd based implementation of the same (Jason
Gunthorpe)
- Attempt to resolve race accessing the iommu_group for a device
between vfio releasing DMA ownership and removal of the device from
the IOMMU driver. Follow-up with support to allow vfio_group to exist
with NULL iommu_group pointer to support existing userspace use cases
of holding the group file open (Jason Gunthorpe)
- Fix error code and hi/lo register manipulation issues in the hisi_acc
variant driver, along with various code cleanups (Longfang Liu)
- Fix a prior regression in GVT-g group teardown, resulting in
unreleased resources (Jason Gunthorpe)
- A significant cleanup and simplification of the mdev interface,
consolidating much of the open coded per driver sysfs interface
support into the mdev core (Christoph Hellwig)
- Simplification of tracking and locking around vfio_groups that fall
out from previous refactoring (Jason Gunthorpe)
- Replace trivial open coded f_ops tests with new helper (Alex
Williamson)
* tag 'vfio-v6.1-rc1' of https://github.com/awilliam/linux-vfio: (77 commits)
vfio: More vfio_file_is_group() use cases
vfio: Make the group FD disassociate from the iommu_group
vfio: Hold a reference to the iommu_group in kvm for SPAPR
vfio: Add vfio_file_is_group()
vfio: Change vfio_group->group_rwsem to a mutex
vfio: Remove the vfio_group->users and users_comp
vfio/mdev: add mdev available instance checking to the core
vfio/mdev: consolidate all the description sysfs into the core code
vfio/mdev: consolidate all the available_instance sysfs into the core code
vfio/mdev: consolidate all the name sysfs into the core code
vfio/mdev: consolidate all the device_api sysfs into the core code
vfio/mdev: remove mtype_get_parent_dev
vfio/mdev: remove mdev_parent_dev
vfio/mdev: unexport mdev_bus_type
vfio/mdev: remove mdev_from_dev
vfio/mdev: simplify mdev_type handling
vfio/mdev: embedd struct mdev_parent in the parent data structure
vfio/mdev: make mdev.h standalone includable
drm/i915/gvt: simplify vgpu configuration management
drm/i915/gvt: fix a memory leak in intel_gvt_init_vgpu_types
...
Allow the vfio_group struct to exist with a NULL iommu_group pointer. When
the pointer is NULL the vfio_group users promise not to touch the
iommu_group. This allows a driver to be hot unplugged while userspace is
keeping the group FD open.
Remove all the code waiting for the group FD to close.
This fixes a userspace regression where we learned that virtnodedevd
leaves a group FD open even though the /dev/ node for it has been deleted
and all the drivers for it unplugged.
Fixes: ca5f21b257 ("vfio: Follow a strict lifetime for struct iommu_group")
Reported-by: Christian Borntraeger <borntraeger@linux.ibm.com>
Tested-by: Matthew Rosato <mjrosato@linux.ibm.com>
Tested-by: Christian Borntraeger <borntraeger@de.ibm.com>
Tested-by: Eric Farman <farman@linux.ibm.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Link: https://lore.kernel.org/r/3-v2-15417f29324e+1c-vfio_group_disassociate_jgg@nvidia.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
SPAPR exists completely outside the normal iommu driver framework, the
groups it creates are fake and are only created to enable VFIO's uAPI.
Thus, it does not need to follow the iommu core rule that the iommu_group
will only be touched while a driver is attached.
Carry a group reference into KVM and have KVM directly manage the lifetime
of this object independently of VFIO. This means KVM no longer relies on
the vfio group file being valid to maintain the group reference.
Tested-by: Matthew Rosato <mjrosato@linux.ibm.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Link: https://lore.kernel.org/r/2-v2-15417f29324e+1c-vfio_group_disassociate_jgg@nvidia.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
This replaces uses of vfio_file_iommu_group() which were only detecting if
the file is a VFIO file with no interest in the actual group.
The only remaning user of vfio_file_iommu_group() is in KVM for the SPAPR
stuff. It passes the iommu_group into the arch code through kvm for some
reason.
Tested-by: Matthew Rosato <mjrosato@linux.ibm.com>
Tested-by: Christian Borntraeger <borntraeger@de.ibm.com>
Tested-by: Eric Farman <farman@linux.ibm.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Link: https://lore.kernel.org/r/1-v2-15417f29324e+1c-vfio_group_disassociate_jgg@nvidia.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
These days not much is using the read side:
- device first open
- ioctl_get_status
- device FD release
- check enforced_coherent
None of this is performance, so just make it into a normal mutex.
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Link: https://lore.kernel.org/r/2-v1-917e3647f123+b1a-vfio_group_users_jgg@nvidia.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Kevin points out that the users is really just tracking if
group->opened_file is set, so we can simplify this code to a wait_queue
that looks for !opened_file under the group_rwsem.
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Link: https://lore.kernel.org/r/1-v1-917e3647f123+b1a-vfio_group_users_jgg@nvidia.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
