linux-toradex.git/drivers/hv, branch v6.19

mshv: handle gpa intercepts for arm64

2026-01-15T07:29:14+00:00

The mshv driver now uses movable pages for guests. For arm64 guests
to be functional, handle gpa intercepts for arm64 too (the current
code implements handling only for x86).

Move some arch-agnostic functions out of #ifdefs so that they can be
re-used.

Fixes: b9a66cd5ccbb ("mshv: Add support for movable memory regions")
Signed-off-by: Anirudh Rayabharam (Microsoft) 
Signed-off-by: Wei Liu

mshv: Add __user attribute to argument passed to access_ok()

2026-01-15T07:12:04+00:00

access_ok() expects its first argument to have the __user attribute
since it is checking access to user space. Current code passes an
argument that lacks that attribute, resulting in 'sparse' flagging
the incorrect usage. However, the compiler doesn't generate code
based on the attribute, so there's no actual bug.

In the interest of general correctness and to avoid noise from sparse,
add the __user attribute. No functional change.

Reported-by: kernel test robot 
Closes: https://lore.kernel.org/oe-kbuild-all/202512141339.791TCKnB-lkp@intel.com/
Signed-off-by: Michael Kelley 
Reviewed-by: Nuno Das Neves 
Signed-off-by: Wei Liu

mshv: Store the result of vfs_poll in a variable of type __poll_t

2026-01-15T07:11:36+00:00

vfs_poll() returns a result of type __poll_t, but current code is using
an "unsigned int" local variable. The difference is that __poll_t carries
the "bitwise" attribute. This attribute is not interpreted by the C
compiler; it is only used by 'sparse' to flag incorrect usage of the
return value. The return value is used correctly here, so there's no
bug, but sparse complains about the type mismatch.

In the interest of general correctness and to avoid noise from sparse,
change the local variable to type __poll_t. No functional change.

Reported-by: kernel test robot 
Closes: https://lore.kernel.org/oe-kbuild-all/202512141339.791TCKnB-lkp@intel.com/
Signed-off-by: Michael Kelley 
Reviewed-by: Nuno Das Neves 
Signed-off-by: Wei Liu

mshv: Align huge page stride with guest mapping

2026-01-15T07:10:27+00:00

Ensure that a stride larger than 1 (huge page) is only used when page
points to a head of a huge page and both the guest frame number (gfn) and
the operation size (page_count) are aligned to the huge page size
(PTRS_PER_PMD). This matches the hypervisor requirement that map/unmap
operations for huge pages must be guest-aligned and cover a full huge page.

Add mshv_chunk_stride() to encapsulate this alignment and page-order
validation, and plumb a huge_page flag into the region chunk handlers.
This prevents issuing large-page map/unmap/share operations that the
hypervisor would reject due to misaligned guest mappings.

Fixes: abceb4297bf8 ("mshv: Fix huge page handling in memory region traversal")
Signed-off-by: Stanislav Kinsburskii 
Reviewed-by: Nuno Das Neves 
Reviewed-by: Michael Kelley 
Signed-off-by: Wei Liu

Drivers: hv: Always do Hyper-V panic notification in hv_kmsg_dump()

2026-01-15T07:01:33+00:00

hv_kmsg_dump() currently skips the panic notification entirely if it
doesn't get any message bytes to pass to Hyper-V due to an error from
kmsg_dump_get_buffer(). Skipping the notification is undesirable because
it leaves the Hyper-V host uncertain about the state of a panic'ed guest.

Fix this by always doing the panic notification, even if bytes_written
is zero. Also ensure that bytes_written is initialized, which fixes a
kernel test robot warning. The warning is actually bogus because
kmsg_dump_get_buffer() happens to set bytes_written even if it fails, and
in the kernel test robot's CONFIG_PRINTK not set case, hv_kmsg_dump() is
never called. But do the initialization for robustness and to quiet the
static checker.

Fixes: 9c318a1d9b50 ("Drivers: hv: move panic report code from vmbus to hv early init code")
Reported-by: kernel test robot 
Reported-by: Dan Carpenter 
Closes: https://lore.kernel.org/all/202512172103.OcUspn1Z-lkp@intel.com/
Signed-off-by: Michael Kelley 
Reviewed-by: Roman Kisel 
Signed-off-by: Wei Liu

Drivers: hv: vmbus: fix typo in function name reference

2026-01-15T07:00:13+00:00

Replace cmxchg by cmpxchg.

Signed-off-by: Julia Lawall 
Reviewed-by: Roman Kisel 
Signed-off-by: Wei Liu

mshv: release mutex on region invalidation failure

2025-12-18T20:00:10+00:00

In the region invalidation failure path in
mshv_region_interval_invalidate(), the region mutex is not released. Fix
it by releasing the mutex in the failure path.

Signed-off-by: Anirudh Rayabharam (Microsoft) 
Fixes: b9a66cd5ccbb ("mshv: Add support for movable memory regions")
Acked-by: Stanislav Kinsburskii 
Reviewed-by: Roman Kisel 
Signed-off-by: Wei Liu

mshv: hide x86-specific functions on arm64

2025-12-17T18:04:15+00:00

The hv_sleep_notifiers_register() and hv_machine_power_off() functions are
only called and declared on x86, but cause a warning on arm64:

drivers/hv/mshv_common.c:210:6: error: no previous prototype for 'hv_sleep_notifiers_register' [-Werror=missing-prototypes]
  210 | void hv_sleep_notifiers_register(void)
      |      ^~~~~~~~~~~~~~~~~~~~~~~~~~~
drivers/hv/mshv_common.c:224:6: error: no previous prototype for 'hv_machine_power_off' [-Werror=missing-prototypes]
  224 | void hv_machine_power_off(void)
      |      ^~~~~~~~~~~~~~~~~~~~

Hide both inside of an #ifdef block.

Fixes: f0be2600ac55 ("mshv: Use reboot notifier to configure sleep state")
Fixes: 615a6e7d83f9 ("mshv: Cleanly shutdown root partition with MSHV")
Signed-off-by: Arnd Bergmann 
Signed-off-by: Wei Liu

mshv: Initialize local variables early upon region invalidation

2025-12-15T18:31:45+00:00

Ensure local variables are initialized before use so that the warning can
print the right values if locking the region to invalidate fails due to
inability to lock the region.

Reported-by: Dan Carpenter 
Fixes: b9a66cd5ccbb ("mshv: Add support for movable memory regions")
Signed-off-by: Stanislav Kinsburskii 
Signed-off-by: Wei Liu

mshv: Use PMD_ORDER instead of HPAGE_PMD_ORDER when processing regions

2025-12-15T18:31:36+00:00

Fix page order determination logic when CONFIG_PGTABLE_HAS_HUGE_LEAVES
is undefined, as HPAGE_PMD_SHIFT is defined as BUILD_BUG in that case.

Fixes: abceb4297bf8 ("mshv: Fix huge page handling in memory region traversal")
Signed-off-by: Stanislav Kinsburskii 
Signed-off-by: Wei Liu