Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[dizel]

I don't know if this is excatly aMule bug, but since in the time of sistem crash kernel log displays also the line Process amule (pid: 10456, stackpage=c6e21000), I decided to post here. If I am wrong, please excuse me (just want to help) ?(.

I am using Debian Stable + baskports. Kernel 2.4.26-1-686 (by backports.org), aMule 2.0.0rc5 from deb http://debian.thermoman.de/ woody amule wxwindows2.4.
Proc: Celeron
Ram: 256 SDRam

The sistem crash usually occurs after more than 2 days of Mule uptime.
Its got something to do with kswap and aMule.
Here is the kernel log, before sistem hangs:

Code: [Select]

Sep  9 12:21:00 damijan kernel: Unable to handle kernel paging request at virtual address 54492b42
Sep  9 12:21:00 damijan kernel:  printing eip:
Sep  9 12:21:00 damijan kernel: c012e8dd
Sep  9 12:21:00 damijan kernel: *pde = 00000000
Sep  9 12:21:00 damijan kernel: Oops: 0002
Sep  9 12:21:00 damijan kernel: CPU:    0
Sep  9 12:21:00 damijan kernel: EIP:    0010:[refill_inactive+145/344]    Tainted: P
Sep  9 12:21:00 damijan kernel: EFLAGS: 00010202
Sep  9 12:21:00 damijan kernel: eax: c024a034   ebx: c1000980   ecx: c024a120   edx: 54492b42
Sep  9 12:21:00 damijan kernel: esi: c100099c   edi: 54492b42   ebp: 00000007   esp: c1313f6c
Sep  9 12:21:00 damijan kernel: ds: 0018   es: 0018   ss: 0018
Sep  9 12:21:00 damijan kernel: Process kswapd (pid: 4, stackpage=c1313000)
Sep  9 12:21:00 damijan kernel: Stack: 00000020 000001d0 c024a1fc c024a1fc c012e9c3 0000003c 000001d0 00000020
Sep  9 12:21:00 damijan kernel:        c012ea40 c1313fa8 c024a1fc 00000001 c1312000 00000000 c1313fa8 00000000
Sep  9 12:21:00 damijan kernel:        c012ebce c024a120 00000000 c1312249 0008e000 c012ec36 c024a120 c1312000
Sep  9 12:21:00 damijan kernel: Call Trace:    [shrink_caches+31/60] [try_to_free_pages_zone+96/236] [kswapd_balance_pgdat+74/152] [kswapd_balance+26/48] [kswapd+153/180]
Sep  9 12:21:00 damijan kernel:   [arch_kernel_thread+40/56]
Sep  9 12:21:00 damijan kernel:
Sep  9 12:21:00 damijan kernel: Code: 89 02 c7 43 1c 00 00 00 00 c7 46 04 00 00 00 00 b8 07 00 00
Sep  9 12:21:19 damijan kernel:  <1>Unable to handle kernel paging request at virtual address 54492b46
Sep  9 12:21:19 damijan kernel:  printing eip:
Sep  9 12:21:19 damijan kernel: c012e88a
Sep  9 12:21:19 damijan kernel: *pde = 00000000
Sep  9 12:21:19 damijan kernel: Oops: 0000
Sep  9 12:21:19 damijan kernel: CPU:    0
Sep  9 12:21:19 damijan kernel: EIP:    0010:[refill_inactive+62/344]    Tainted: P
Sep  9 12:21:19 damijan kernel: EFLAGS: 00010a93
Sep  9 12:21:19 damijan kernel: eax: 0000000d   ebx: 54492b26   ecx: 000cf1c0   edx: 000015f5
Sep  9 12:21:19 damijan kernel: esi: 000001d2   edi: 54492b42   ebp: 0000000d   esp: c6e21e04
Sep  9 12:21:19 damijan kernel: ds: 0018   es: 0018   ss: 0018
Sep  9 12:21:19 damijan kernel: Process amule (pid: 10456, stackpage=c6e21000)
Sep  9 12:21:19 damijan kernel: Stack: 00000020 000001d2 c024a1fc c024a1fc c012e9c3 0000003c 000001d2 00000020
Sep  9 12:21:19 damijan kernel:        c012ea40 c6e21e40 c6e20000 00000000 00000010 00000000 c6e21e40 00000000
Sep  9 12:21:19 damijan kernel:        c012f4d5 c024a3e0 00000000 00000010 0000000c 00000db6 00000282 c024a15c
Sep  9 12:21:19 damijan kernel: Call Trace:    [shrink_caches+31/60] [try_to_free_pages_zone+96/236] [balance_classzone+61/448] [__alloc_pages+394/652] [lru_cache_add+90/96]
Sep  9 12:21:19 damijan kernel:   [_alloc_pages+22/24] [do_anonymous_page+52/228] [do_no_page+51/432] [handle_mm_fault+82/176] [do_page_fault+352/1184] [do_page_fault+0/1184]

[Jacobo221]

Please go to http://wiki.amule.org and look for the backtraces tutorial. then submit a baktrace of the crash rather than the kernel log.
thanks for your report, it will be very helpful :-)

[dizel]

Fine, thanks, will check out backtraces, and submit that in the future.

[Kry]

Nono, this is really interesting. Seems from my knowledge that kswap was trying to get another virtual ram page and kernel crashed, and probably it's a coincidence that amule is involved.


But notice us if it happens again.

Thanks,
   Kry

[dizel]

@Kry: It did happen again, but I think now I figured out what the problem is, aMule does really nothing have to do with it (a least not directly).
My computer has got two NICs, eth0 connected to ADSL modem, eth1 connected to my brothers computer, I have IP forwarding (NAT) set up for his access to internet.
Everything runs perfectly for days and days of uptime (of my aMule), but when he starts aMule (for his aMule I configured TCP port 5662 and UDP 5672, so they don't interfere with my standard ports 4664,4672), on my computer a configured port forwarding for his 5665 and 5672 ports. Under this conditions (both computers running aMule) my computer after cca. 5-6 hours becomes a "zombie", eventually after a day or so it crashes.
It is (I think) a kswap-kernel-iptables-IPforwarding combination problem, aMule alone works 100% OK  :baby:  

Just to let you know.
thanks for your time and great work,
d