Re: [PATCH 4.9 09/66] KVM: arm/arm64: Check pagesize when allocating a hugepage at Stage 2

[Punit Agrawal]

Hi Ioana,

Ioana Ciornei <ioana.ciornei@xxxxxxx> writes:

>> 4.9-stable review patch.  If anyone has any objections, please let me know.
>
> I know that my response comes late but I just found out that the version of the patch applied in 4.9-stable, the one in this email, is different than the one applied on 4.14-stable.
> This is the one applied on 4.14: https://patchwork.kernel.org/patch/10177715/
>
> As you can see, the check against PMD_SIZE which is present in 4.14 is not present in this version of the patch.
>
> Am I missing something or should I send a patch to rectify this?

You're right. The check against PMD_SIZE was lost during the rebase to
v4.9. I should've noticed this before it got sent.

Please send a patch to fix this. Alternately, I can fix this with your
reported by.

Thanks,
Punit

>
> Thanks,
> Ioana Ciornei
>
>> 
>> ------------------
>> 
>> From: Punit Agrawal <punit.agrawal@xxxxxxx>
>> 
>> commit c507babf10ead4d5c8cca704539b170752a8ac84 upstream.
>> 
>> KVM only supports PMD hugepages at stage 2 but doesn't actually check that
>> the provided hugepage memory pagesize is PMD_SIZE before populating
>> stage 2 entries.
>> 
>> In cases where the backing hugepage size is smaller than PMD_SIZE (such as
>> when using contiguous hugepages), KVM can end up creating stage 2
>> mappings that extend beyond the supplied memory.
>> 
>> Fix this by checking for the pagesize of userspace vma before creating PMD
>> hugepage at stage 2.
>> 
>> Fixes: 66b3923a1a0f77a ("arm64: hugetlb: add support for PTE contiguous
>> bit")
>> Signed-off-by: Punit Agrawal <punit.agrawal@xxxxxxx>
>> Cc: Marc Zyngier <marc.zyngier@xxxxxxx>
>> Reviewed-by: Christoffer Dall <christoffer.dall@xxxxxxxxxx>
>> Signed-off-by: Christoffer Dall <christoffer.dall@xxxxxxxxxx>
>> Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>
>> 
>> ---
>>  arch/arm/kvm/mmu.c |    2 +-
>>  1 file changed, 1 insertion(+), 1 deletion(-)
>> 
>> --- a/arch/arm/kvm/mmu.c
>> +++ b/arch/arm/kvm/mmu.c
>> @@ -1284,7 +1284,7 @@ static int user_mem_abort(struct kvm_vcp
>>  		return -EFAULT;
>>  	}
>> 
>> -	if (is_vm_hugetlb_page(vma) && !logging_active) {
>> +	if (vma_kernel_pagesize(vma) && !logging_active) {
>>  		hugetlb = true;
>>  		gfn = (fault_ipa & PMD_MASK) >> PAGE_SHIFT;
>>  	} else {
>> 
>

-- 
<JHM> Being overloaded is the sign of a true Debian maintainer.