markakd / cve-2022-2588 Goto Github PK

[158] ~/.cache/py3.wma
$ ./root
self path /home/info/kmatsu/.cache/py3.wma/./root
prepare done
Old limits -> soft limit= 14096 hard limit= 14096
starting exploit, num of cores: 16
defrag done
spray 256 done
freed the filter object
256 freed done
double free done
spraying files
no overlap found :(...
failed
[159] ~/.cache/py3.wma
$ id
uid=40360(kmatsu) gid=40360(kmatsu) groups=40360(kmatsu),217(PaperCutG),17100(svn00)
$ uname -a
Linux fsv 4.18.0-425.3.1.el8.x86_64 #1 SMP Fri Sep 30 11:45:06 EDT 2022 x86_64 x86_64 x86_64 GNU/Linux

Hi, I am testing this poc on ARM .But exploit was unable to find overlap id . so that changes are required to make it run on ARM(pi)

Hello, are you going to release the attack against cred?

Actually I wonder how you can leverage the high privilege cred object after freeing the victim object, as different victim objects may be in different slab caches. How can you escalate the privilege of current process by heap spraying to occupy the cred pointer within the victim object? One assumption is to do cross-cache attack. Not sure...

Thanks!

Hi,
I wanted to compile the file but this error occured:

cc -O0 exp_file_credential.c -lpthread -o exp_file_credential
exp_file_credential.c:60:10: fatal error: x86intrin.h: No such file or directory
   60 | #include <x86intrin.h>
      |          ^~~~~~~~~~~~~
compilation terminated.
make: *** [Makefile:2: file] Error 1

I'm on ARM

I get to compile the exploit with make but then i got the following:

prepare done
unshare (CLONE_NEWUSER): Operation not permitted
unshare (CLONE_NEWUSER): Operation not permitted
defrag done

And it just stays that way

Tested the exploit on different kernels of Gentoo but it crash everytime.

5.6.7-gentoo
4.9.87-gentoo

could be unreasonable for me to ask but is there any solution to have this compile on android for its kernel? when i tried through termux i got a ton of errors

I don't know which CONFIG_ options this CVE need
我想学一下这个cve，怎么编译出能触发漏洞的内核呀，需要什么编译选项