| 1 |
On 30.04.2007, at 22:06, pageexec@××××××××.hu wrote: |
| 2 |
|
| 3 |
>> In the latest mm-sources there is the possibility to choose VIA C7 as |
| 4 |
>> processor. This is enabled by the x86_64-mm-add-an-option-for-the- |
| 5 |
>> via- |
| 6 |
>> c7-which-sets-appropriate-l1-cache.patch patch. I did not see the NX |
| 7 |
>> bit enabled message after booting (maybe there are too many messages? |
| 8 |
> |
| 9 |
> did you enable PAE (HIGHMEM64) in your .config? |
| 10 |
|
| 11 |
stoile@chris ~ $ zgrep HIGHMEM64 /proc/config.gz |
| 12 |
CONFIG_HIGHMEM64G=y |
| 13 |
|
| 14 |
I should haev that in every kernel since i always took my |
| 15 |
hardened .config and did make oldconfig. |
| 16 |
|
| 17 |
>> Is there a way to enlarge the kernel ring buffer?) but the second |
| 18 |
>> part of the test program mentioned in <http://www.ussg.iu.edu/ |
| 19 |
>> hypermail/linux/kernel/0612.1/0509.html> segfaults, so it seems to be |
| 20 |
>> working. |
| 21 |
> |
| 22 |
> what does paxtest report on the kernels you tried (if there's an |
| 23 |
> NX bit in use, at least the non-mprotect test should succeed else |
| 24 |
> there's a problem somewhere indeed)? |
| 25 |
|
| 26 |
I'm not sure wich tests you mean. I'll post the complete log of the |
| 27 |
test with hardened sources and the differences between the tests as |
| 28 |
soon as i finish them. |
| 29 |
|
| 30 |
PaXtest - Copyright(c) 2003,2004 by Peter Busser <peter@×××××××××.org> |
| 31 |
Released under the GNU Public Licence version 2 or later |
| 32 |
|
| 33 |
Mode: blackhat |
| 34 |
Linux chris 2.6.20-hardened-r2 #4 Tue May 1 15:52:59 UTC 2007 i686 |
| 35 |
VIA Esther pr |
| 36 |
ocessor 1300MHz GNU/Linux |
| 37 |
|
| 38 |
Executable anonymous mapping : Killed |
| 39 |
Executable bss : Killed |
| 40 |
Executable data : Killed |
| 41 |
Executable heap : Killed |
| 42 |
Executable stack : Killed |
| 43 |
Executable anonymous mapping (mprotect) : Killed |
| 44 |
Executable bss (mprotect) : Killed |
| 45 |
Executable data (mprotect) : Killed |
| 46 |
Executable heap (mprotect) : Killed |
| 47 |
Executable stack (mprotect) : Killed |
| 48 |
Executable shared library bss (mprotect) : Killed |
| 49 |
Executable shared library data (mprotect): Killed |
| 50 |
Writable text segments : Killed |
| 51 |
Anonymous mapping randomisation test : 17 bits (guessed) |
| 52 |
Heap randomisation test (ET_EXEC) : 13 bits (guessed) |
| 53 |
Heap randomisation test (ET_DYN) : 23 bits (guessed) |
| 54 |
Main executable randomisation (ET_EXEC) : No randomisation |
| 55 |
Main executable randomisation (ET_DYN) : 15 bits (guessed) |
| 56 |
Shared library randomisation test : 17 bits (guessed) |
| 57 |
Stack randomisation test (SEGMEXEC) : 23 bits (guessed) |
| 58 |
Stack randomisation test (PAGEEXEC) : No randomisation |
| 59 |
Return to function (strcpy) : Vulnerable |
| 60 |
Return to function (memcpy) : Vulnerable |
| 61 |
Return to function (strcpy, RANDEXEC) : Vulnerable |
| 62 |
Return to function (memcpy, RANDEXEC) : Vulnerable |
| 63 |
Executable shared library bss : Killed |
| 64 |
Executable shared library data : Killed |
| 65 |
|
| 66 |
|
| 67 |
Philipp |
| 68 |
|
| 69 |
|
| 70 |
-- |
| 71 |
gentoo-hardened@g.o mailing list |
Archives