milo-d / house-of-gods

House of Gods Heap Exploitation Technique

House of Gods - Arena Hijacking Technique

House-of-Gods is a new heap exploitation technique for older versions of the GNU C Library (< glibc 2.27). There are currently two different revisions of the same exploit. The first revision is way more inefficient and less applicable in real world scenarios compared to the second revision.

The goal of both revisions is to hijack the thread_arena by corrupting the malloc_state.next pointer.

Revision 1

This revision requires a lot of allocations i.e. 200 allocations on my test-system. The exact number is system dependent. The rootcause is a repeatable WAF on an unsorted-chunk.

Paper: https://github.com/Milo-D/house-of-gods/blob/master/rev1/HOUSE_OF_GODS.TXT

Revision 2

This revision hijacks the arena within only 8 allocations and drops a shell after a total of 10 allocations. Furthermore, only a single WAF bug on an unsorted-chunk is needed in order to launch the attack.

Paper: https://github.com/Milo-D/house-of-gods/blob/master/rev2/HOUSE_OF_GODS.TXT

how2heap

A how2heap tutorial of the second revision can be now found here https://github.com/shellphish/how2heap/blob/master/glibc_2.24/house_of_gods.c

glibc bug report

The second revision takes advantage of inappropriate malloc behaviour, which is still present in the most recent version of glibc. Although a rather minor issue with low priority, the bug was filed here https://sourceware.org/bugzilla/show_bug.cgi?id=29709

Proof of Concept

For both revision, type

./patch && ./exploit.py

house-of-gods's Issues

Segfault if compile the .c source code

Hey Milo do you have any clue why the binary crashes despite the fact that gcc doesn't warn about anything?