| 1 |
On Thursday 06 October 2005 21:31, Marco Matthies wrote: |
| 2 |
> Hemmann, Volker Armin wrote: |
| 3 |
> > to do this fancy jump-around stuff, you need to know exactly where which |
| 4 |
> > snippet resides in ram - one error and all your work is lost. |
| 5 |
> > |
| 6 |
> > So, and because on gentoo everything is compiled by yourself with a |
| 7 |
> > little bit different setting like the next gentoo user, it is a lot |
| 8 |
> > harder to guess the correct adresses. Not like the binary distributions, |
| 9 |
> > where one is laid out like the other. |
| 10 |
> |
| 11 |
> Well, lets find out. On my home system, i did this: |
| 12 |
> |
| 13 |
> strace gzip > /dev/null |
| 14 |
> hit ctrl-c |
| 15 |
> find the mmap that maps /lib/libc.so.6 by: |
| 16 |
> - find the open call that opens libc, note that the open call returns |
| 17 |
> the file descriptor |
| 18 |
> - find the next mmap that acts on that fd (second to last argument |
| 19 |
> to mmap) |
| 20 |
> - mmap returns the address it mmapped to |
| 21 |
> |
| 22 |
> Short excerpt: |
| 23 |
> $ strace gzip > /dev/null |
| 24 |
> [...snip...] |
| 25 |
> open("/lib/libc.so.6", O_RDONLY) = 3 |
| 26 |
> [... reads, lseeks and fstats snipped ...] |
| 27 |
> mmap(NULL, 2248680, PROT_READ|PROT_EXEC, MAP_PRIVATE|MAP_DENYWRITE, 3, |
| 28 |
> 0) = 0x2aaaaabc1000 |
| 29 |
> [...snip...] |
| 30 |
> |
| 31 |
> I got 0x2aaaaabc1000 on my machine, i would be interested to know what |
| 32 |
> your machine did. But judging from the fact that the people from |
| 33 |
> hardened added address space layout randomization, i'd guess we're going |
| 34 |
> to get the same address or something that varies but not very randomly |
| 35 |
> (low entropy). Make sure that you test with gzip though, other apps will |
| 36 |
> have slighly different addresses as they seem to load libs in a |
| 37 |
> different order -- this won't help though, as an exploit will be |
| 38 |
> targeted at a specific app. |
| 39 |
> |
| 40 |
> But I'm still interested if we all get the same addresses or slightly |
| 41 |
> different. |
| 42 |
> |
| 43 |
|
| 44 |
and I get: |
| 45 |
|
| 46 |
strace gzip > /dev/null |
| 47 |
execve("/bin/gzip", ["gzip"], [/* 63 vars */]) = 0 |
| 48 |
<snip> |
| 49 |
open("/lib/tls/libc.so.6", O_RDONLY) = 3 |
| 50 |
<snip> |
| 51 |
mmap(NULL, 2261000, PROT_READ|PROT_EXEC, MAP_PRIVATE|MAP_DENYWRITE, 3, 0) = |
| 52 |
0x2aaaaabc3000 |
| 53 |
<snip> |
| 54 |
|
| 55 |
-- |
| 56 |
Conclusions |
| 57 |
In a straight-up fight, the Empire squashes the Federation like a bug. Even |
| 58 |
with its numerical advantage removed, the Empire would still squash the |
| 59 |
Federation like a bug. Accept it. -Michael Wong |
| 60 |
-- |
| 61 |
gentoo-amd64@g.o mailing list |
Archives