Hiya all,
Somebody please help me! I was fighting for almost two weeks, but stumped
by now.
Simply can't keep my system up for more then a few hours or a day without
getting kernel Oops. This is the ha3pmf.pmmf.hu gateway. Dumped TNOS two
weeks ago and I have this problem since then. I've never got such
complaints before, though TNOS was not stable enough, spammable, etc.
Usually the first victim is mheardd, then comes netromd and the system
crashes. Worst, it's in the college, after a halt I always have to go up
there, fiddle this or that, come down and cross my fingers...
Last two Oops:
Warning: kfree_skb passed an skb still on a list (from c017b509).
Unable to handle kernel NULL pointer dereference at virtual address 00000004
current->tss.cr3 = 018c7000, %cr3 = 018c7000
*pde = 00000000
Oops: 0002
CPU: 0
EIP: 0010:[skb_recv_datagram+355/416]
EFLAGS: 00010047
eax: c18350d0 ebx: 00000000 ecx: 00000246 edx: 00000000
esi: c18c9ddc edi: c18c8000 ebp: c18ea514 esp: c18c9dcc
ds: 0018 es: 0018 ss: 0018
Process mheardd (pid: 1987, process nr: 19, stackpage=c18c9000)
Stack: c18c9e84 c18c9f48 c18352b0 c0187770 c18c8000 c19273d4 00000301 00050f3b
00000400 c0171f79 c18ea4d0 00000000 00000000 c18c9e20 c19273bc 000005dc
c18c9e84 c18c9f48 00000001 00000000 c18ea4d0 c013c863 00000301 00050f3b
Call Trace: [multwrite_intr+0/172] [packet_recvmsg+81/252] [inode_getblk+71/408]
[sock_recvmsg+65/188] [sys_recvfrom+172/264] [free_page_and_swap_cache+101/108]
[do_munmap+537/560]
[sys_socketcall+394/524] [sys_munmap+46/72] [common_interrupt+24/32]
[sys_time+19/32] [system_call+52/56] [startup_32+43/285]
Code: 89 6a 04 89 55 00 c7 00 00 00 00 00 c7 40 04 00 00 00 00 c7
Warning: kfree_skb passed an skb still on a list (from c017b509).
Unable to handle kernel NULL pointer dereference at virtual address 00000004
current->tss.cr3 = 01cea000, %cr3 = 01cea000
*pde = 00000000
Oops: 0002
CPU: 0
EIP: 0010:[skb_recv_datagram+355/416]
EFLAGS: 00010047
eax: c139daa0 ebx: 00000000 ecx: 00000246 edx: 00000000
esi: 00000200 edi: c1c99e84 ebp: c1fbb2a4 esp: c1c99dcc
ds: 0018 es: 0018 ss: 0018
Process netromd (pid: 1976, process nr: 11, stackpage=c1c99000)
Stack: c1c99e84 c1c99f48 c139d820 c1d6ec80 00000286 00000000 00000000 c1d6ec80
00000296 c0171f79 c1fbb260 00000000 00000000 c1c99e20 c1a15c3c 00000200
c1c99e84 c1c99f48 c1d6ec80 c1d6ec80 c1fbb260 00000010 c1d6ec80 c1fe3c68
Call Trace: [packet_recvmsg+81/252] [sock_recvmsg+65/188] [nr_release+73/240]
[sys_recvfrom+172/264] [free_pages+39/44] [free_wait+104/116] [do_select+481/504]
[do_select+437/504] [sys_select+1103/1296] [sys_socketcall+394/524]
[startup_32+43/285] [sys_time+19/32] [system_call+52/56] [startup_32+43/285]
Code: 89 6a 04 89 55 00 c7 00 00 00 00 00 c7 40 04 00 00 00 00 c7
Details:
- old 486DX2/66 CPU, 32M RAM, Baycom USCC (only two 1200Bd on-board modems
used now), SMC Ultra NIC and a small 540M hd
- kernel: 2.2.15pre14, distribution: Debian potato
- related softwares: libax25-0.0.7, ax25-tools-0.0.5 (with Tomi's netrom
patch), ax25-apps-0.0.4, node-0.3.0
- running ax25ipd, netromd, ax25d, mheardd
- no special services on the PC: smtp (postfix), ftp (proftpd), etc.
What I tried up to now:
- back to kernel-2.2.14 for awhile
- replaced CPU and RAM
- replaced motherboard with two different types, latter was a P1/133
- originally used Debian potato's libax25, ax25-apps, ax25-tools and node
packages, then compiled fresh ones from sources
- also tried awznode-v0.4-pre2 instead of node-0.3.0, thought it has
nothing to do with the issue I'm afraid
None of them helped.
I can try to get symbols by ksymoops if a kind fellow would like to
parse, unfortunatelly I'm not a coder.
One more observation: using listen usually results earlier Oops.
Any and all your ideas are welcome! (Except of drilling holes on the box
to let oops out of the box ;-) (c)Terry?)
Thanks, 73... Sanyi
--
Sandor Dibuz | email: [EMAIL PROTECTED]
Pecs, Hungary | AX.25: [EMAIL PROTECTED]
