| 1 |
On 20 Apr 2009 at 21:03, Peter Hjalmarsson wrote: |
| 2 |
|
| 3 |
> I realised earlier today that foldingathome (installed with the help of |
| 4 |
> portage) had not started a new WU since 5 of april, and when I started |
| 5 |
> to investigate I found out that the "cores" had problem running. |
| 6 |
> |
| 7 |
> cd /opt/foldingathome && |
| 8 |
> paxctl -c FahCore_*.exe && |
| 9 |
> paxctl -PEMrXS FahCore_*.exe |
| 10 |
> |
| 11 |
> makes it work again. |
| 12 |
> foldingathome has worked in the past without problems, and I do not |
| 13 |
> really know what have changed more then some kernel-upgrades (but |
| 14 |
> booting the last kept 2.6.27-hardened did not help either) and keeping |
| 15 |
> the userland up to stable x86. |
| 16 |
> What can I do to not have to do this dance? |
| 17 |
|
| 18 |
can you re-enable pax on the binaries and see if they produce any logs |
| 19 |
(pax kills)? if they do, try to remove mprotect only and see if that |
| 20 |
helps. the other issue could be a bad glibc and lack of GNU_STACK headers, |
| 21 |
execstack -c would fix that without having to compromise on pax. |
Archives