On 02/09/2024 14:25, Matias Ezequiel Vara Larsen wrote:
> Hello Steven,
>
> On Wed, Aug 21, 2024 at 04:38:22PM +0100, Steven Price wrote:
...
>> +static int private_memslot_fault(struct kvm_vcpu *vcpu,
>> + phys_addr_t fault_ipa,
>> + struct kvm_memory_slot *memslot)
>> +{
>> + struct kvm *kvm = vcpu->kvm;
>> + gpa_t gpa_stolen_mask = kvm_gpa_stolen_bits(kvm);
>> + gfn_t gfn = (fault_ipa & ~gpa_stolen_mask) >> PAGE_SHIFT;
>> + bool is_priv_gfn = !((fault_ipa & gpa_stolen_mask) == gpa_stolen_mask);
>> + bool priv_exists = kvm_mem_is_private(kvm, gfn);
>> + struct kvm_mmu_memory_cache *memcache = &vcpu->arch.mmu_page_cache;
>> + kvm_pfn_t pfn;
>> + int ret;
>> +
>> + if (priv_exists != is_priv_gfn) {
>> + kvm_prepare_memory_fault_exit(vcpu,
>> + fault_ipa & ~gpa_stolen_mask,
>> + PAGE_SIZE,
>> + kvm_is_write_fault(vcpu),
>> + false, is_priv_gfn);
>> +
>> + return 0;
>> + }
>
> If I understand correctly, `kvm_prepare_memory_fault_exit()` ends up
> returning to the VMM with the KVM_EXIT_MEMORY_FAULT exit reason. The
> documentation says (https://docs.kernel.org/virt/kvm/api.html#kvm-run):
>
> "Note! KVM_EXIT_MEMORY_FAULT is unique among all KVM exit reasons in that
> it accompanies a return code of ‘-1’, not ‘0’! errno will always be set
> to EFAULT or EHWPOISON when KVM exits with KVM_EXIT_MEMORY_FAULT,
> userspace should assume kvm_run.exit_reason is stale/undefined for all
> other error numbers."
>
> Shall the return code be different for KVM_EXIT_MEMORY_FAULT?
>
> Thanks, Matias.
Ah, good spot - I've no idea why KVM_EXIT_MEMORY_FAULT is special in
this regard, but yes I guess we should be returning -EFAULT here.
Thanks,
Steve
