@Sh1n0d4 @subgraph @grsecurity @marcan42 it wasn't about the bugreport, but his subsequent *lies* and the resulting noise.
1
2
@chrisrohlf length of code doesn't really matter, RAP works fine for c++ thunks too ;).
@chrisrohlf how would the exact hash generation method influence whether you can redirect to a gaJIT bypassing the checks?
@chrisrohlf RAP doesn't care about aligned/unaligned insns. so i still don't see how you'd enter a gaJIT sequence.
@chrisrohlf how would you execute a gaJIT without violating a RAP-like defense?
new grsecurity blog post on today's Linux kernel JIT spray attack news: forums.grsecurity.net/viewto…
49
51
@GCsVentures so it works on the configs i tested it, let me know if anything breaks (will be in the next patch).
1
1
@GCsVentures it's WIP, but after fixing the C side, i'm not sure what will be left and how fixable that is.
1
1
spender's prepared a small FAQ on RAP at grsecurity.net/rap_faq.php .
1
56
55
@copumpkin that said, i've done enough of that work to be able to have a RAP protected chromium since 2014AD :).
1
2
@copumpkin the bigger issue will be fixing all the bad fptr casts that the plugin will flag for you but can't fix itself.
1
@copumpkin i developed it under gentoo but any distro will work that allows passing cflags/etc to packages.
1
@rootkovska so for the kernel it's GPLv2, the commercial version GPLv3 (to meet the eligible compilation condition needed for userland).
1
@rootkovska as the copyright holder on my code i get to choose the license ;). there're circumstances that steer that decision of course.
2
2