Skip to contentSkip to conversation

It looks like there will be no more public versions of PaX and Grsec.

kernel-hardening [email protected] 3 participants 3 messages May 2, 2017

lore
Share & tools
lore
#1 lore raw
Quoted 11 lines · On 2017-05-02 16:22, [email protected] wrote:

Hi David, I read your paper [1] about how RAP wouldn't defend against data-only attacks. Do you plan on contributing your PT-Rand mitigation to KSPP?

That was never mentioned in the grsec RAP announcements for obvious marketing reasons. With KERNSEAL not being public no one can verify its effectiveness either.

[1]
https://www.internetsociety.org/sites/default/files/ndss2017_05B-4_Davi_paper.pdf

That is certainly doable (and it would be cool), but mainly depends on how interesting such a feature would be for upstream. My impression was that the interest is rather limited, since messing with the linear mapping generally is a big nono (because of TLB pressure).

But we're really getting off-topic.

Best,
David

#2 lore raw
In reply to [email protected]
Quoted 7 lines · On 2 May 2017 at 16:22, <[email protected]> wrote:

Hi David, I read your paper [1] about how RAP wouldn't defend against data-only attacks. Do you plan on contributing your PT-Rand mitigation to KSPP?

That was never mentioned in the grsec RAP announcements for obvious marketing reasons. With KERNSEAL not being public no one can verify its effectiveness either.

That's nonsense. RAP's scope was made clear from the beginning, focusing on control flow integrity, i.e. prevent unintended code reuse. Just look at [1], page 6 and see for yourself that data-only attacks are out-of-scope for RAP.

[1] https://pax.grsecurity.net/docs/PaXTeam-H2HC15-RAP-RIP-ROP.pdf

Regards,
Mathias