| Message ID | 20240809205158.1340255-3-amoorthy@google.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | Set up KVM_EXIT_MEMORY_FAULTs when arm64/x86 stage-2 fault handlers fail | expand |
On Fri, Aug 09, 2024, Anish Moorthy wrote: > Right now userspace just gets a bare EFAULT when the stage-2 fault > handler fails to fault in the relevant page. Set up a memory fault exit > when this happens, which at the very least eases debugging and might > also let userspace decide on/take some specific action other than > crashing the VM. Heh, most of the way there (from my hack-a-patch response), just need to add the KVM_BUG_ON() + -EIO conversions. Can you send x86 and arm64 as separate series for v3? E.g. for x86, just this patch and the -EIO changes. I'm pretty sure the docs updates can go in the arm64 series (I need to send another response to that patch).
diff --git a/arch/x86/kvm/mmu/mmu.c b/arch/x86/kvm/mmu/mmu.c index 901be9e420a4..c22c807696ae 100644 --- a/arch/x86/kvm/mmu/mmu.c +++ b/arch/x86/kvm/mmu/mmu.c @@ -3264,6 +3264,7 @@ static int kvm_handle_error_pfn(struct kvm_vcpu *vcpu, struct kvm_page_fault *fa return RET_PF_RETRY; } + kvm_mmu_prepare_memory_fault_exit(vcpu, fault); return -EFAULT; }
Right now userspace just gets a bare EFAULT when the stage-2 fault handler fails to fault in the relevant page. Set up a memory fault exit when this happens, which at the very least eases debugging and might also let userspace decide on/take some specific action other than crashing the VM. Signed-off-by: Anish Moorthy <amoorthy@google.com> --- arch/x86/kvm/mmu/mmu.c | 1 + 1 file changed, 1 insertion(+)