| 1 |
Brian Kroth wrote: |
| 2 |
>> Does the problem happen still if you do something like this? |
| 3 |
>> cp cactid cactid.nopax |
| 4 |
>> paxctl -mps cactid.nopax |
| 5 |
>> cactid.nopax --verbosity=5 -f 1 -l 10 |
| 6 |
> |
| 7 |
> Looping for 20 minutes without a problem. I'm going to start trying to |
| 8 |
> figure out which of the protections is making it angry. |
| 9 |
> |
| 10 |
> Brian |
| 11 |
|
| 12 |
paxctl -MPS /usr/sbin/cactid.nopax |
| 13 |
paxctl -mp /usr/sbin/cactid.nopax |
| 14 |
|
| 15 |
while [ "$?" == "0" ]; do /usr/sbin/cactid.nopax; done |
| 16 |
|
| 17 |
... |
| 18 |
|
| 19 |
Nov 4 01:11:13 tux-mc Bad page state in process 'cactid.nopax' |
| 20 |
Nov 4 01:11:13 tux-mc page:c13b8380 flags:0x40000001 mapping:00000000 |
| 21 |
mapcount:0 count:0 |
| 22 |
Nov 4 01:11:13 tux-mc Trying to fix it up, but a reboot is needed |
| 23 |
Nov 4 01:11:13 tux-mc Backtrace: |
| 24 |
Nov 4 01:11:13 tux-mc [<c0403da8>] show_trace_log_lvl+0x1a/0x30 |
| 25 |
Nov 4 01:11:13 tux-mc [<c0403dd0>] show_trace+0x12/0x14 |
| 26 |
Nov 4 01:11:13 tux-mc [<c0403eca>] dump_stack+0x16/0x18 |
| 27 |
Nov 4 01:11:13 tux-mc [<c044ba59>] bad_page+0x66/0x97 |
| 28 |
Nov 4 01:11:13 tux-mc [<c044c52e>] free_hot_cold_page+0x71/0x180 |
| 29 |
Nov 4 01:11:13 tux-mc [<c044c647>] free_hot_page+0xa/0xc |
| 30 |
Nov 4 01:11:13 tux-mc [<c044ef70>] __page_cache_release+0x84/0x90 |
| 31 |
Nov 4 01:11:13 tux-mc [<c044efdc>] put_page+0x27/0x36 |
| 32 |
Nov 4 01:11:13 tux-mc [<c04558aa>] do_wp_page+0x219/0x3fc |
| 33 |
Nov 4 01:11:13 tux-mc [<c0456b37>] __handle_mm_fault+0x277/0x2cd |
| 34 |
Nov 4 01:11:13 tux-mc [<c0414406>] do_page_fault+0x1cd/0x7bf |
| 35 |
Nov 4 01:11:13 tux-mc [<c06e6d55>] error_code+0x75/0x80 |
| 36 |
Nov 4 01:11:13 tux-mc ======================= |
| 37 |
|
| 38 |
|
| 39 |
paxctl -MPS ... |
| 40 |
paxctl -s ... |
| 41 |
|
| 42 |
No problems after 20+ minutes of looping the command. So SEGMEXEC seems |
| 43 |
to be the culprit, correct? Did something change in the way that was |
| 44 |
handled from 2.6.20 to 2.6.22? |
| 45 |
|
| 46 |
I'll have to see if it's the same for nagios, but that can wait until |
| 47 |
tomorrow. |
| 48 |
|
| 49 |
Brian |
Archives