6.12-stable review patch. If anyone has any objections, please let me know.
------------------
From: Will Deacon <will@xxxxxxxxxx>
[ Upstream commit d44679fb954ffea961036ed1aeb7d65035f78489 ]
Calling the MMIO_GUARD hypercall from guests which have not been
enrolled (e.g. because they are running without pvmfw) results in
-EINVAL being returned. In this case, MMIO_GUARD is not active
and so we can simply proceed with the normal ioremap() routine.
Don't fail ioremap() if MMIO_GUARD fails; instead WARN_ON_ONCE()
to highlight that the pvm environment is slightly wonky.
Fixes: 0f1269495800 ("drivers/virt: pkvm: Intercept ioremap using pKVM MMIO_GUARD hypercall")
Signed-off-by: Will Deacon <will@xxxxxxxxxx>
Link: https://lore.kernel.org/r/20241202145731.6422-2-will@xxxxxxxxxx
Signed-off-by: Catalin Marinas <catalin.marinas@xxxxxxx>
Signed-off-by: Sasha Levin <sashal@xxxxxxxxxx>
---
drivers/virt/coco/pkvm-guest/arm-pkvm-guest.c | 6 +-----
1 file changed, 1 insertion(+), 5 deletions(-)
diff --git a/drivers/virt/coco/pkvm-guest/arm-pkvm-guest.c b/drivers/virt/coco/pkvm-guest/arm-pkvm-guest.c
index 56a3859dda8a1..4230b817a80bd 100644
--- a/drivers/virt/coco/pkvm-guest/arm-pkvm-guest.c
+++ b/drivers/virt/coco/pkvm-guest/arm-pkvm-guest.c
@@ -87,12 +87,8 @@ static int mmio_guard_ioremap_hook(phys_addr_t phys, size_t size,
while (phys < end) {
const int func_id = ARM_SMCCC_VENDOR_HYP_KVM_MMIO_GUARD_FUNC_ID;
- int err;
-
- err = arm_smccc_do_one_page(func_id, phys);
- if (err)
- return err;
+ WARN_ON_ONCE(arm_smccc_do_one_page(func_id, phys));
phys += PAGE_SIZE;
}
--
2.43.0
