NFSD3 : oops with sec=krb5 and RPCSEC_GSS_KRB5 not compiled in

DISCUSSION Quentin Godfroy 2 participants 2 messages Oct 29, 2007 – Nov 10, 2007

Subsystems (2)

NFS client NFS server

lore

Share & tools

lore

Quentin Godfroy

author

2007-10-29 23:35 UTC #0 lore raw

Hi,

on a 2.6.22.6, when trying to change an already mounted nfsv3 export from sec=sys to sec=krb5, with CONFIG_RPCSEC_GSS_KRB5 not compiled in (which meant that the rps.svcgssd service was refusing to run), i got the following oops on the server side :

Quoted 37 lines

BUG: unable to handle kernel NULL pointer dereference at virtual address

00000008
printing eip:
d0af084b
*pde = 00000000

Oops: 0000 [#1]
Modules linked in: nfsd exportfs lockd nfs_acl sunrpc ipt_LOG ipt_recent
ipt_REJECT xt_tcpudp iptable_filter ip_tables x_tables ipv6 ebtable_broute
ebtable_nat ebtable_filter ebtables tcp_diag inet_diag floppy uhci_hcd
iTCO_wdt usbcore
CPU:    0
EIP:    0060:[<d0af084b>]    Not tainted VLI
EFLAGS: 00010246   (2.6.22.6-pluton #1)

EIP is at encode_fsid+0x7b/0xa0 [nfsd]
eax: c58ed0e0 ebx: c9ebb004 ecx: 00000000 edx: 00000002
esi: c1f58048 edi: c9ebb0ec ebp: c9ebb004 esp: c599ff18
ds: 007b es: 007b fs: 0000 gs: 0000 ss: 0068
Process nfsd (pid: 13237, ti=c599e000 task=cbea9a50 task.ti=c599e000)

Stack: 00000000 c1f58038 d0af0922 c9d7cb40 d0aeeeef 00000000 c9ec4000
d0af1e30 
      d0b0f464 c1f58014 d0af1e50 c9ebb0ec c9ec4000 d0ae327f c9ec4000
cb01a504 
      c9ec4000 c1f58000 c9ec4000 00000000 c9ebb138 cab1f480 d0a89a34
c9ec4000 
Call Trace:
[<d0af0922>] encode_fattr3+0xb2/0x140 [nfsd]
[<d0aeeeef>] nfsd3_proc_getattr+0xff/0x110 [nfsd]
[<d0af1e30>] nfs3svc_encode_attrstat+0x0/0x40 [nfsd]
[<d0af1e50>] nfs3svc_encode_attrstat+0x20/0x40 [nfsd]
[<d0ae327f>] nfsd_dispatch+0xef/0x200 [nfsd]
[<d0a89a34>] svc_process+0x444/0x7e0 [sunrpc]
[<c010496b>] common_interrupt+0x23/0x28
[<d0ae3895>] nfsd+0x155/0x260 [nfsd]
[<d0ae3740>] nfsd+0x0/0x260 [nfsd]
[<c0104aeb>] kernel_thread_helper+0x7/0x1c
=======================
Code: c3 8b 83 88 00 00 00 8b 48 30 89 cb c1 fb 1f 89 d8 0f c8 89 06 89 c8

Quoted 3 lines

0f c8 89 46 04 8d 46 08 5b 5e c3 8b 83 88 00 00 00 8b 48 34 <8b> 59 08 8b 01 8b 51 04 31 d8 8b 59 0c 0f c8 89 46 04 8d 46 08
EIP: [<d0af084b>] encode_fsid+0x7b/0xa0 [nfsd] SS:ESP 0068:c599ff18

Moreover, all the other exported mounts freezed instantly, and the only solution was to reboot the server.

The configure file is on
http://www.eleves.ens.fr/home/godfroy/config-2.6.22.6

Quentin Godfroy

-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now >> http://get.splunk.com/
_______________________________________________
NFS maillist - [email protected]
https://lists.sourceforge.net/lists/listinfo/nfs

J. Bruce Fields

2007-11-10 19:17 UTC #1 lore raw

Quoted 4 lines · Quentin Godfroy

on a 2.6.22.6, when trying to change an already mounted nfsv3 export from sec=sys to sec=krb5, with CONFIG_RPCSEC_GSS_KRB5 not compiled in (which meant that the rps.svcgssd service was refusing to run), i got the following oops on the server side :

Hm. This looks a little like b8da0d1c27f144: "knfsd: Validate filehandle type in fsid_source", though I'm not sure why changing security flavors would trigger this. Is the problem reproduceable? If so, could you see if it's fixed in 2.6.23.1 (which includes that patch)?

--b.

Quoted 51 lines

Quoted 38 lines
BUG: unable to handle kernel NULL pointer dereference at virtual address
00000008
printing eip:
d0af084b
*pde = 00000000
Oops: 0000 [#1]
Modules linked in: nfsd exportfs lockd nfs_acl sunrpc ipt_LOG ipt_recent
ipt_REJECT xt_tcpudp iptable_filter ip_tables x_tables ipv6 ebtable_broute
ebtable_nat ebtable_filter ebtables tcp_diag inet_diag floppy uhci_hcd
iTCO_wdt usbcore
CPU:    0
EIP:    0060:[<d0af084b>]    Not tainted VLI
EFLAGS: 00010246   (2.6.22.6-pluton #1)
EIP is at encode_fsid+0x7b/0xa0 [nfsd]
eax: c58ed0e0 ebx: c9ebb004 ecx: 00000000 edx: 00000002
esi: c1f58048 edi: c9ebb0ec ebp: c9ebb004 esp: c599ff18
ds: 007b es: 007b fs: 0000 gs: 0000 ss: 0068
Process nfsd (pid: 13237, ti=c599e000 task=cbea9a50 task.ti=c599e000)
Stack: 00000000 c1f58038 d0af0922 c9d7cb40 d0aeeeef 00000000 c9ec4000
d0af1e30 
      d0b0f464 c1f58014 d0af1e50 c9ebb0ec c9ec4000 d0ae327f c9ec4000
cb01a504 
      c9ec4000 c1f58000 c9ec4000 00000000 c9ebb138 cab1f480 d0a89a34
c9ec4000 
Call Trace:
[<d0af0922>] encode_fattr3+0xb2/0x140 [nfsd]
[<d0aeeeef>] nfsd3_proc_getattr+0xff/0x110 [nfsd]
[<d0af1e30>] nfs3svc_encode_attrstat+0x0/0x40 [nfsd]
[<d0af1e50>] nfs3svc_encode_attrstat+0x20/0x40 [nfsd]
[<d0ae327f>] nfsd_dispatch+0xef/0x200 [nfsd]
[<d0a89a34>] svc_process+0x444/0x7e0 [sunrpc]
[<c010496b>] common_interrupt+0x23/0x28
[<d0ae3895>] nfsd+0x155/0x260 [nfsd]
[<d0ae3740>] nfsd+0x0/0x260 [nfsd]
[<c0104aeb>] kernel_thread_helper+0x7/0x1c
=======================
Code: c3 8b 83 88 00 00 00 8b 48 30 89 cb c1 fb 1f 89 d8 0f c8 89 06 89 c8
Quoted 3 lines
0f c8 89 46 04 8d 46 08 5b 5e c3 8b 83 88 00 00 00 8b 48 34 <8b> 59 08 8b 01 8b 51 04 31 d8 8b 59 0c 0f c8 89 46 04 8d 46 08
EIP: [<d0af084b>] encode_fsid+0x7b/0xa0 [nfsd] SS:ESP 0068:c599ff18
Moreover, all the other exported mounts freezed instantly, and the only solution was to reboot the server.
The configure file is on
http://www.eleves.ens.fr/home/godfroy/config-2.6.22.6
Quentin Godfroy

Quoted 8 lines

-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now >> http://get.splunk.com/
_______________________________________________
NFS maillist - [email protected]
https://lists.sourceforge.net/lists/listinfo/nfs