Joined February 2010
Replying to @lvwr
RAP protects returns with the type hash too and can do CGD without having to clone anything.
2
2
Replying to @lvwr
there's nothing new in your work that is patent eligible, but nice try for a copout. let's just see your code and i'll do the comparison ;).
1
2
Replying to @lvwr
RAP is open source as you already know. what prevented you from comparisons since you already claimed to have one for the program committee?
2
3
it was proactively 'fixed' by tedunangst.com/flak/post/my-… in 2013. a lesson in unforeseen consequences.
dear 'other side', brace yourselves for the consequences of gcc.gnu.org/git/?p=gcc.git;a… :)
2
6
it was introduced by gcc.gnu.org/git/?p=gcc.git;a… in 2001.
dear 'other side', brace yourselves for the consequences of gcc.gnu.org/git/?p=gcc.git;a… :)
4
2
dear 'other side', brace yourselves for the consequences of gcc.gnu.org/git/?p=gcc.git;a… :)
7
2
14
i did discuss those two (mis)features at the time to no avail. guess we just have different standards ;).
pointless when stuff like __ro_after_init and now refcount_t can go in despite being bad engineering...
1
1
IMHO bugs due to entirely avoidable design errors (NIH) are simply irresponsible and bad engineering.
1
the NIH'd version of vmap stack is full of bugs, still a whack-a-mole game a few releases later.
2
1
no, RAP doesn't rely on the secrecy of addresses (not even the probabilistic return address part).
dear infosec thought leaders: get your revenge by emailing spender@ with your testimonial about our technology for the revamped grsec site.
1
10
8
not the first time you guessed wrong, is it? ;) this is the wrong medium for legal questions.
wrong medium. if you have questions email our contact address please.
sorry but twitter is good for armchair 'experts' at most. go write some code instead, maybe you'll fare better ;).
1
your legal expertise is like your exploit writing skills, empty puffs without evidence ;)
3
Replying to @paxteam @comex
the asm instrumentation is really no different from what the plugin does though so you can study/work with that.
Replying to @comex
it's a general macro to help instrument assembly code. right now only KERNEXEC makes use of it but i'll add RAP too.
1
Replying to @comex
this is why the commercial version is claimed to be ROP-proof. lots of little things under the hood there ;).
1
1
Replying to @comex
pax_ret isn't implemented for RAP yet as it requires source changes (that of the kernel, not the plugin).
2