1 |
hey guys and ladies, |
2 |
|
3 |
just wanted to share some good fun with you: |
4 |
|
5 |
-- >>> Regenerating /etc/ld.so.cache... |
6 |
* Caching service |
7 |
dependencies... [ ok ] |
8 |
>>> app-office/openoffice-1.1.0-r2 merged. |
9 |
|
10 |
11:30:49 [/usr/local/chroots/chroot001:18038.pty-s0.epoch] epoch ~ # cd |
11 |
/opt/OpenOffice.org1.1.0/program/ |
12 |
21:59:14 [/usr/local/chroots/chroot001:18038.pty-s0.epoch] epoch |
13 |
/opt/OpenOffice.org1.1.0/program # file soffice.bin |
14 |
soffice.bin: ELF 32-bit LSB shared object, Intel 80386, version 1 |
15 |
(SYSV), not stripped |
16 |
21:59:18 [/usr/local/chroots/chroot001:18038.pty-s0.epoch] epoch |
17 |
/opt/OpenOffice.org1.1.0/program # readelf -s soffice.bin | grep guard |
18 |
995: 00000000 32 OBJECT GLOBAL DEFAULT UND __guard@GLIBC_2.3.2 |
19 |
(12) |
20 |
1363: 00000000 32 OBJECT GLOBAL DEFAULT UND __guard@@GLIBC_2.3.2 |
21 |
|
22 |
yeeha, daddy made it into etdyn and propolice openoffice. |
23 |
|
24 |
thanks to the gcc-3.3.2 and the glibc guard, this stuff rocks now. |
25 |
|
26 |
when you use current hardened-gcc 3.3.2 versions with propolice and |
27 |
etdyn on a non pax kernel, you will see openoffice dying with a |
28 |
relocation error in Scrt1.o regarding a missing or irrelocatable main |
29 |
function. |
30 |
|
31 |
dont you open bugs for this, it will get fix0red in next hcc versions. |
32 |
|
33 |
on a pax kernel the openoffice build dies way earlier with regcomp dying |
34 |
because of java bytecode executed during the Makefile run, i am going to |
35 |
sleep about this also. |
36 |
|
37 |
also on my todo list is working through the mails from Peter Mazinger |
38 |
and tackling the kernel and nostdlib issues with propolice. |
39 |
|
40 |
thanks and greets to the channel, |
41 |
|
42 |
Alex |
43 |
|
44 |
Alexander Gabert <pappy@g.o> |
45 |
http://www.gentoo.org/proj/en/hardened |
46 |
|
47 |
|
48 |
-- |
49 |
gentoo-hardened@g.o mailing list |