[asterisk-dev] strange oops with dahdi-linux 2.2.1

[Felix Salfelder]

Hi all.

I am running asterisk on a debian (lenny) machine with an A400M card. asterisk
and the kernel modules are backported from squeeze (2.2.1). i am using the
debian kernel (2.6.26) but get close to equivalent results when using a vanilla
2.6.27 or 2.6.32-bpo (backports.org)

when i happen to use a telephone (FXS calling some SIP address), and try to
log in via ssh at the same time, i get the following oops, which, sometimes
recovers and sometimes completely freezes the machine:

[  649.036159] BUG: unable to handle kernel NULL pointer dereference at 00000000                                                                                         
[  649.040108] IP: [<c0102893>] __switch_to+0x23/0xf7                                                                                                                    
[  649.040108] *pde = 00000000                                                                                                                                           
[  649.040108] Oops: 0002 [#1]                                                                                                                                           
[  649.040108] Modules linked in: dahdi_echocan_oslec echo xt_TCPMSS xt_tcpmss iptable_mangle xt_tcpudp ipt_MASQUERADE iptable_nat nf_nat nf_conntrack_ipv4 iptable_filter ip_tables bridge ipv6 authenc esp4 aead xfrm4_mode_transport deflate zlib_deflate zlib_inflate ctr twofish twofish_common camellia serpent blowfish des_generic cbc xcbc sha256_generic sha1_generic crypto_null af_key usblp leds_alix led_class evdev geode_rng rng_core ieee80211_crypt wcopenpci pcspkr pppoe pppox cs5535_gpio ppp_generic slhc geode_aes cdc_ether crypto_blkcipher usbnet wctdm xt_state dahdi crc_ccitt nf_conntrack x_tables ext2 mbcache sd_mod usb_storage scsi_mod ohci_hcd ehci_hcd ssb usbcore pcmcia pcmcia_core firmware_class via_rhine mii thermal_sys ide_generic amd74xx ide_disk ide_core                                                                    
[  649.040108]                                                                                                                                                           
[  649.040108] Pid: 2608, comm: bash Not tainted (2.6.26-2-486 #1)                                                                                                       
[  649.040108] EIP: 0060:[<c0102893>] EFLAGS: 00010002 CPU: 0                                                                                                            
[  649.040108] EIP is at __switch_to+0x23/0xf7                                                                                                                           
[  649.040108] EAX: 00000000 EBX: cf6e6c90 ECX: cf6e6fe0 EDX: cf6e6a40                                                                                                   
[  649.040108] ESI: cf6e6a40 EDI: cf6e6e50 EBP: cf6e70a0 ESP: cf6fbd84                                                                                                   
[  649.040108]  DS: 007b ES: 007b FS: 0000 GS: 0033 SS: 0068                                                                                                             
[  649.040108] Process bash (pid: 2608, ti=cf6fa000 task=cf6e6e50 task.ti=cecc6000)                                                                                      
[  649.040108] Stack: cf6e6a40 cf8484e0 cf974820 cf6e6e50 c02a666e cf6e6a40 00000082 00000000                                                                            
[  649.040108]        00000000 1ce000d0 00000097 cf6e6bcc 00011ba8 00000000 7fffffff 00000000                                                                            
[  649.040108]        cf6fbe5c ceaf2880 c02a681b cf6b55e0 00000080 cf6fbe0c cf409560 c024c3a8                                                                            
[  649.040108] Call Trace:                                                                                                                                               
[  649.040108]  [<c02a666e>] schedule+0x310/0x351                                                                                                                        
[  649.040108]  [<c02a681b>] schedule_timeout+0x13/0x86                                                                                                                  
[  649.040108]  [<c024c3a8>] __alloc_skb+0x53/0xff                                                                                                                       
[  649.040108]  [<c01738ec>] __d_lookup+0x8d/0xc1
 .... ( tell me if it makes sense to post the rest of these 500 lines )

i dont know why exactly incoming ssh triggers this bug, but a netcat does not.
now since ive switch to dahdi modules version 2.3.0.1 (only tried with the
2.6.32 kernel), it seems to run stable.

could you give me a hint on how to track this down? or is 2.2.1 just too old to
care (in that case i'd file a debian bugreport).

regards
felix