oops during nfs4 mount

[Trond Myklebust]

On Fri, 2008-03-28 at 07:14 -0400, Michael Spang wrote:
> The following oops occured during a mount attempt:
> 
> [64711.272242] BUG: unable to handle kernel NULL pointer dereference at 00000010
> [64711.272279] IP: [<c011b8bf>] task_rq_lock+0x38/0x73
> [64711.272306] *pde = 00000000 
> [64711.272323] Oops: 0000 [#1] SMP 
> [64711.272341] Modules linked in: des_generic cbc crypto_blkcipher
> rpcsec_gss_krb5 auth_rpcgss ppdev lp autofs4 nfs lockd nfs_acl sunrpc
> ipv6 ext2 fuse eeprom it87 hwmon_vid loop snd_via82xx gameport
> snd_ac97_codec ac97_bus snd_pcm parport_pc parport snd_timer
> snd_page_alloc snd_mpu401_uart floppy snd_rawmidi snd_seq_device rtc
> i2c_viapro snd i2c_core soundcore shpchp pci_hotplug button via_agp
> agpgart evdev ext3 jbd mbcache dm_mirror dm_snapshot dm_mod raid1 md_mod
> ide_generic ide_cd_mod cdrom ide_disk ata_generic uhci_hcd ehci_hcd
> sata_via via_rhine mii usbcore via82cxxx ide_core libata scsi_mod dock
> skge thermal processor fan
> [64711.272588] 
> [64711.272601] Pid: 5991, comm: mount Not tainted (2.6.25-0-686 #1)
> [64711.272619] EIP: 0060:[<c011b8bf>] EFLAGS: 00210012 CPU: 0
> [64711.272637] EIP is at task_rq_lock+0x38/0x73
> [64711.272652] EAX: 00000000 EBX: c0406840 ECX: 00000000 EDX: e5be0020
> [64711.272670] ESI: c0406840 EDI: e537db94 EBP: e537db84 ESP: e537db74
> [64711.272687]  DS: 007b ES: 007b FS: 00d8 GS: 0033 SS: 0068
> [64711.272704] Process mount (pid: 5991, ti=e537c000 task=e5bda020 task.ti=e537c000)
> [64711.272723] Stack: e5be0020 e5be0020 e88fc0c0 e537dcb0 e537dba4 c011baaf 00000000 0000000f 
> [64711.272764]        00200286 e5be0020 e88fc0c0 e537dcb0 e537dbac c011bb8f e537dbb8 c0131b4e 
> [64711.272803]        e98585d8 e537dbc0 f8c7a3c4 e537dbcc f8c59a43 e98585d8 e537dbd8 f8c59bbf 
> [64711.272842] Call Trace:
> [64711.272870]  [<c011baaf>] ? try_to_wake_up+0x19/0xd1
> [64711.272903]  [<c011bb8f>] ? wake_up_process+0xf/0x11
> [64711.272927]  [<c0131b4e>] ? kthread_stop+0x37/0x72
> [64711.272954]  [<f8c7a3c4>] ? nfs_callback_down+0x31/0x42 [nfs]
> [64711.273006]  [<f8c59a43>] ? nfs_free_client+0x81/0xac [nfs]
> [64711.273046]  [<f8c59bbf>] ? nfs_put_client+0x5f/0x64 [nfs]
> [64711.273085]  [<f8c59c35>] ? nfs_free_server+0x71/0xaa [nfs]
> [64711.273123]  [<f8c5aaea>] ? nfs4_create_server+0x327/0x349 [nfs]
> [64711.273161]  [<c0132070>] ? remove_wait_queue+0x22/0x27
> [64711.273189]  [<c0185c70>] ? free_poll_entry+0x18/0x1b
> [64711.273215]  [<c0185c8f>] ? poll_freewait+0x1c/0x4d
> [64711.273253]  [<c0177127>] ? cache_alloc_debugcheck_after+0xe8/0x15b
> [64711.273278]  [<c017823d>] ? __kmalloc+0xd4/0x129
> [64711.273300]  [<c0178258>] ? __kmalloc+0xef/0x129
> [64711.273300]  [<c01e8cb2>] ? copy_from_user+0x2c/0x4f
> [64711.273300]  [<f8c61cce>] ? nfs4_get_sb+0x363/0x48c [nfs]
> [64711.273300]  [<c0177dac>] ? cache_alloc_refill+0x223/0x5e0
> [64711.273300]  [<c0177127>] ? cache_alloc_debugcheck_after+0xe8/0x15b
> [64711.273300]  [<c017823d>] ? __kmalloc+0xd4/0x129
> [64711.273300]  [<c0178258>] ? __kmalloc+0xef/0x129
> [64711.273300]  [<c018e17b>] ? alloc_vfsmnt+0x8c/0xb4
> [64711.273300]  [<c018e10a>] ? alloc_vfsmnt+0x1b/0xb4
> [64711.273300]  [<c018e17b>] ? alloc_vfsmnt+0x8c/0xb4
> [64711.273300]  [<c017cc38>] ? vfs_kern_mount+0x81/0xf7
> [64711.273300]  [<c017ccf2>] ? do_kern_mount+0x32/0xba
> [64711.273300]  [<c018f060>] ? do_new_mount+0x46/0x74
> [64711.273300]  [<c018f217>] ? do_mount+0x189/0x1a7
> [64711.273300]  [<c0160c9f>] ? __get_free_pages+0x45/0x4c
> [64711.273300]  [<c018d58c>] ? copy_mount_options+0x27/0x10e
> [64711.273300]  [<c018f299>] ? sys_mount+0x64/0x9b
> [64711.273300]  [<c0104a28>] ? sysenter_past_esp+0x6d/0xc5
> [64711.273300]  =======================
> [64711.273300] Code: 83 ec 04 89 45 f0 9c 58 0f 1f 84 00 00 00 00 00 89
> c2 fa 0f 1f 84 00 00 00 00 00 90 89 17 89 f3 e8 ef f4 01 00 8b 55 f0 8b
> 42 04 <8b> 40 10 03 1c 85 20 12 3c c0 89 d8 e8 37 ed 1a 00 8b 55 f0 8b 
> [64711.273300] EIP: [<c011b8bf>] task_rq_lock+0x38/0x73 SS:ESP
> 0068:e537db74
> [64711.273300] ---[ end trace 0b6e7a1b96c8ec26 ]---
> 
> Kernel is 2.6.25-rc6-CITI_NFS4_ALL-1-299-gae05ece. I don't quite know
> how to reproduce it, at this point this is just an FYI.

Hmm... Bruce & Jeff,

Doesn't the line

	nfs_callback_info.task = NULL;

in nfs_callback_svc() need protection by the BKL?

Cheers
  Trond