[PATCH][RFC] arm64: kaslr: add pseudo-RNG in kernel

[Ard Biesheuvel]

On 29 February 2016 at 13:47, Laurentiu Tudor <laurentiu.tudor at nxp.com> wrote:
> On 02/26/2016 02:37 PM, Ard Biesheuvel wrote:
>> On 26 February 2016 at 13:18, Mark Rutland <mark.rutland at arm.com> wrote:
>>> On Fri, Feb 26, 2016 at 12:51:25PM +0100, Ard Biesheuvel wrote:
>>>> On 26 February 2016 at 12:01, Laurentiu Tudor <laurentiu.tudor at nxp.com> wrote:
>>>>> In case the bootloader doesn't provide randomness
>>>>> (through device tree or by uefi protocol) generate
>>>>> a pseudo-random seed based on the timer counter.
>>>>> People might find this "week rng" approach convenient
>>>>> as it gets rid of the bootloader dependency.
>>>>>
>>>>> The patch tries to mimic the x86's rdtsc
>>>>> based implementation authored by Kees Cook.
>>>>>
>>>>> Signed-off-by: Laurentiu Tudor <laurentiu.tudor at nxp.com>
>>>>> Cc: Ard Biesheuvel <ard.biesheuvel at linaro.org>
>>>>> Cc: Kees Cook <keescook at chromium.org>
>>>>
>>>> Hi Laurentiu,
>>>>
>>>> I appreciate the interest in this work, but to be honest, I don't like
>>>> this at all. I went out of my way to ensure that
>>>> a) the kernel itself does not take part in generating the random bits, and
>>>> b) the random bits are used in such a way that there is no correlation
>>>> between the randomization of the core kernel, the module region and
>>>> the linear region if there is no correlation between the random bits.
>>>>
>>>> The limited entropy of the cycle counter at early boot defeats that,
>>>> and even worse, it will not encourage platform vendors to implement
>>>> this properly in their boot code, given that it will appear to work,
>>>> and the only thing more dangerous than no security is a false sense of
>>>> security imo.
>>>
>>> I agree that a false sense of security is a worry here.
>>>
>>> Either way, I think we need numbers. It's non-obvious how much entropy
>>> we can acquire through counters or other means at early boot time.
>>>
>>> Has anyone done an analysis of environmental entropy available (through
>>> any means) at early boot, VM vs native?
>>>
>>
>> Define 'environmental'. The only architected thing you can rely on is
>> the timer, which really does not hold a lot of entropy on modern solid
>> state platforms. I don't have the numbers, but I do have the
>> experience to back this up, unfortunately (at my former employer).
>>
>> You could argue that KASLR does not require strong entropy like key
>> generation, but I would prefer to simply steer clear of that entire
>> debate. The bootloader simply needs to do the best job it can, either
>> based on some peripheral that the kernel has no awareness of this
>> early in the boot, or perhaps by other means if it doesn't (which
>> could include storing a random seed in the file system, or even in a
>> EFI variable). Punting it to the kernel is really the last resort.
>>
>>> It's also not obvious that vendors will correctly implement the EFI RNG
>>> protocol; depending on the above we may want to mix in additional entropy
>>> regardless.
>>>
>>
>> True, but that is not the point. The main risk I see is that vendors
>> will not bother at all once the digits start to look random to the
>> human eye.
>>
>>>> What I would ack, for development purposes, is something similar to
>>>> what Mark Rutland implemented for randomizing TEXT_OFFSET, so that
>>>> developers get to exercise this code even if their boot environment
>>>> does not provide any entropy. Anything beyond that is a nack as far as
>>>> I am concerned.
>>>
>>> FWIW, I would not like to see that approach. I can easily see a build-time
>>> constant KASLR seed being abused to give a false sense of security. Having a
>>> bootloader or hypervisor provide different random seeds to the same Image gives
>>> you a much better turnaround time for testing regardless (vs rebuilding,
>>> copying, etc).
>>>
>
> Hi guys,
>
> And thanks for all the great comments. In conclusion, seems that this
> patch is not the right direction, at least not in its current form.
>
>> If it would be *instead* of the ordinary handling, with a big fat
>> blurb saying that KASLR is disabled, I would not mind.
>
> Ard,
>
> Let me see if i understood your thoughts correctly.
> Have something like a KConfig option (suggestions for name, please?
> CONFIG_DEBUG_RANDOMIZE_BASE?) that disables the normal RNG seed
> handling and replaces it with this counter based weak rng,
> plus the big fat warning.
>

No, that is not what I meant.

What i meant was, that, in order for the KASLR plumbing to be
exercised by developers whose bootloader does not supply a random
seed, I could imagine a Kconfig option that rolls the dice at build
time, and hardcodes that value into kaslr_early_init rather than
retrieve it from the DT. This means your kernel will execute from the
same 'random' address each time, even if you rebuild it, and only a
Kconfig change may trigger a new value to be generated.

Since this is not KASLR but a build time hack to get testing coverage
for the code, it should announce that in some way.

However, while I said I was not opposed to the idea, I am not crazy
about it either, and it is not up to me to ultimately decide what gets
merged anyway. So before you sit down and spend the time, you may want
to get the maintainer's opinion first.

-- 
Ard.