Re: [kernel-hardening] Re: [PATCHv4 0/4] WX checking for arm64

[Ard Biesheuvel]

On 7 November 2016 at 19:49, Mark Rutland <mark.rutland@xxxxxxx> wrote:
> On Mon, Nov 07, 2016 at 03:38:02PM +0000, Mark Rutland wrote:
>> On Sun, Oct 30, 2016 at 03:03:07PM +0000, Catalin Marinas wrote:
>> > On Thu, Oct 27, 2016 at 09:27:30AM -0700, Laura Abbott wrote:
>> > > Laura Abbott (4):
>> > >   arm64: dump: Make ptdump debugfs a separate option
>> > >   arm64: dump: Make the page table dumping seq_file optional
>> > >   arm64: dump: Remove max_addr
>> > >   arm64: dump: Add checking for writable and exectuable pages
>> >
>> > Queued for 4.10. Thanks.
>>
>> Catalin mentioned to me that he saw some KASAN splats when testing; it
>> looks like need a fixup something like the below.
>
> As an aside, it looks like any ptdump usage when KASAN is enabled takes
> several minutes, which at boot time looks like a hang.
>
> AFAICT, this is because KASAN allocates *huge* VA ranges (4TB+) worth of
> zeroed shadow memory at pte granularity (reusing the same pmd, pud,
> tables), and the ptdump code dutifully walks this with, with the added
> KASAN instrumentation overhead.
>
> I'll try to dig into that tomorrow; I suspect/hope it's not necessary to
> keep all of that mapped.
>

I have noticed that in the past, but I see how this delay at boot time
is an issue. However, I don't think there is a huge cost involved in
terms of memory footprint: AFAIK, the same PMD/PTE/kasan zero page are
mapped over and over across the range.