[prev in list] [next in list] [prev in thread] [next in thread] 

List:       linux-net
Subject:    Re: Locked Network Buffers, Emerging Technologies, and Dummy modulesGirls who throw themselves at me
From:       Andrew Stanley-Jones <asj () cban ! com>
Date:       1996-06-30 13:28:52
[Download RAW message or body]

> try v2.3beta2. The crash problem is fixed, I believe ( a parameter change
> in the hard_header routine was the culprit). Hopefully you can get some
> better stats now.....

Ok, v2.3beta2 seems to have a couple problems, and the locked buffers 
problem still exists.

I had the module crash on me twice with this oops (crashed last 
night at 2:43am with this ooops): 

Jun 21 13:31:09 cban kernel: Unable to handle kernel NULL pointer dereference at \
                virtual address c0000000
Jun 21 13:31:09 cban kernel: current->tss.cr3 = 00101000,
Jun 21 13:31:09 cban kernel: *pde = 00102067
Jun 21 13:31:09 cban kernel: *pte = 00000027
Jun 21 13:31:09 cban kernel: Oops: 0002
Jun 21 13:31:09 cban kernel: CPU:    0
Jun 21 13:31:09 cban kernel: EIP:    0010:[<0204319c>]
Jun 21 13:31:09 cban kernel: EFLAGS: 00010206
Jun 21 13:31:09 cban kernel: eax: 00008f04   ebx: 00000000   ecx: 02040240   edx: \
                02040240
Jun 21 13:31:09 cban kernel: esi: 00008f04   edi: 00000001   ebp: 001a80c8   esp: \
                001a80b8
Jun 21 13:31:09 cban kernel: ds: 0018   es: 0018   fs: 002b   gs: 0018   ss: 0018
Jun 21 13:31:09 cban kernel: Process swapper (pid: 0, process nr: 0, \
                stackpage=001a6218)
Jun 21 13:31:09 cban kernel: Stack: 0204319c 01119e20 00000000 00000001 001a80e4 \
                0201dfd8 0204319c 00000000
Jun 21 13:31:09 cban kernel:        000000a6 0204319c 00000001 001a8108 0201db0d \
                0204319c 00000000 0204319c
Jun 21 13:31:09 cban kernel:        00000000 00000001 00144700 00000000 001a8134l
Jun 21 13:31:09 cban kernel: Call Trace: [<0204319c>] [<0201dfd8>] [<0204319c>] \
                [<0204319c>] [<0201db0d>] [<0204319c>] [<0204319c>]
Jun 21 13:31:09 cban kernel:        [<00144700>] [<0201d947>] [<02032970>] \
                [<0202b3ed>] [<0010c8dd>] [<0010bb46>] [<001096bc>] [<0010a5e2>]
Jun 21 13:31:09 cban kernel:        [<001093d0>] [<00109243>] [<00116760>] \
                [<0011049c>]
Jun 21 13:31:09 cban kernel: Code: 88 13 00 00 00 00 00 00 88 13 00 00 44 06 00 00 00 \
                00 00 07
Jun 21 13:31:09 cban kernel: Aiee, killing interrupt handler
Jun 21 13:31:09 cban kernel: kfree of non-kmalloced memory: 001a8260, next= 001a807c, \
                order=1137900
Jun 21 13:31:09 cban kernel: kfree of non-kmalloced memory: 001a8250, next= 00115d2c, \
                order=0
Jun 21 13:31:09 cban kernel: kfree of non-kmalloced memory: 001a8684, next= 001a807c, \
order=1138010

EIP: 0204319c

0204311c workscr        [et5025elf.2.3beta2]
0204319c lapb_wa        [et5025elf.2.3beta2] <==
02043e3c x25_lines      [et5025elf.2.3beta2]


I've had the module crash 2 or 3 times with:

Jun 29 19:25:47 cban kernel: general protection: 0000
Jun 29 19:25:47 cban kernel: CPU:    0
Jun 29 19:25:47 cban kernel: EIP:    0010:[<0201a934>]
Jun 29 19:25:47 cban kernel: EFLAGS: 00010246
Jun 29 19:25:47 cban kernel: eax: 00000000   ebx: 000000e2   ecx: 180b180b  edx: \
                020320e0
Jun 29 19:25:47 cban kernel: esi: 01730cd8   edi: 180b180b   ebp: 0125bee8   esp: \
                0125becc
Jun 29 19:25:47 cban kernel: ds: 0018   es: 0018   fs: 002b   gs: 002b   ss: 0018
Jun 29 19:25:47 cban kernel: Process sendmail (pid: 87, process nr: 9, \
                stackpage=0125b000)
Jun 29 19:25:47 cban kernel: Stack: 01730cd8 000000e2 00000000 0000f618 00000848 \
                00000200 180b180b 0125bf00 
Jun 29 19:25:47 cban kernel:        0201b49e 01730cd8 00000005 00000000 0204319c \
                0125bf18 0201e97e 0000f618 
Jun 29 19:25:47 cban kernel:        0204319c 0204319c 00000f08 0125bf30 0201e21e \
                0204319c 000000a6 0204319c 
Jun 29 19:25:47 cban kernel: Call Trace: [<0201b49e>] [<0204319c>] [<0201e97e>] \
                [<0204319c>] [<0204319c>] [<0201e21e>] [<0204319c>] 
Jun 29 19:25:47 cban kernel:        [<0204319c>] [<0201db0d>] [<0204319c>] \
                [<0204319c>] [<0011a206>] [<0201d947>] [<02032970>] [<0011cc00>] 
Jun 29 19:25:47 cban kernel:        [<0202b3ed>] [<0010c8dd>] [<0010bb46>] 
Jun 29 19:25:47 cban kernel: Code: 8b 04 8a ff d0 83 c4 0c eb 08 90 90 90 90 90  90 \
                31 c0 89 f9 
Jun 29 19:25:47 cban kernel: Aiee, killing interrupt handler

EIP: 0201a934

0201a344 l2_ppp_initcp  [et5025elf.2.3beta2]
0201a4a4 l2_ppp_init    [et5025elf.2.3beta2]
0201a594 l2_ppp_event   [et5025elf.2.3beta2] <==
0201a9a4 l2_ppp_proc    [et5025elf.2.3beta2]
0201abb4 l2_ppp_output  [et5025elf.2.3beta2]
0201afb4 l2_ppp_timeout [et5025elf.2.3beta2]


On the Lock Buffers issue.  I've switched to the IP Alias code, and still
locked buffers.  With 2.0.0 the rate at which they get
generated is much less, about 40 a day. 

After 4 days of uptime:

Jun 29 09:11:23 cban kernel: Networking buffers in use          : 138 
Jun 29 09:11:23 cban kernel: Network buffers locked by drivers  : 132
Jun 29 09:11:23 cban kernel: Total network buffer allocations   : 3399288 
Jun 29 09:11:23 cban kernel: Total failed network buffer allocs : 0 
Jun 29 09:11:23 cban kernel: Total free while locked events     : 132 
Jun 29 09:11:23 cban kernel: IP fragment buffer size            : 0 

If you need more information please let me know, Andrew

---
Andrew Stanley-Jones                Canadian Business Advertising Network
WWW: http://www.cban.com          Tel: (403) 486-7783 Fax: (403) 484-6004


[prev in list] [next in list] [prev in thread] [next in thread]