On Mon, 2009-11-30 at 00:06 +0000, Daniel J Blueman wrote: > Hi Trond, > > It looks less likely this is fixed in the NFS-related changes in > 2.6.31.6, so I'm reporting this 2.6.31.4 rpcauth_checkverf oops [1]. > Info from /proc/mounts included [2]. Client is 2.6.32-rc8 and seems > quite seldom to reproduce. > > What info would help here, such as line numbers and disassembly? > > Thanks, > Daniel > > --- [1] > > BUG: unable to handle kernel paging request at 0000000400001038 > > IP: [<ffffffffa03c3c22>] rpcauth_checkverf+0x32/0x70 [sunrpc] > > PGD 483fa067 PUD 0 > > Oops: 0000 [#1] SMP > > last sysfs file: > /sys/devices/pci0000:00/0000:00:1f.2/host0/target0:0:0/0:0:0:0/block/sda/uevent > > CPU 5 > > Modules linked in: pl2303 usbserial nls_iso8859_1 nls_cp437 vfat fat > ppdev kvm_intel kvm binfmt_misc bridge stp bnep btusb joydev > snd_hda_codec_atihdmi microcode snd_hda_codec_idt dm_crypt arc4 ecb > nfsd exportfs nfs lockd nfs_acl snd_hda_intel snd_seq_dummy > auth_rpcgss snd_seq_oss snd_hda_codec snd_seq_midi snd_hwdep > snd_rawmidi snd_pcm_oss iwlagn snd_seq_midi_event snd_mixer_oss > mmc_block dell_wmi sunrpc iwlcore mac80211 snd_seq snd_pcm > iptable_filter snd_seq_device snd_timer ip_tables uvcvideo x_tables lp > dell_laptop psmouse videodev v4l1_compat v4l2_compat_ioctl32 usbhid > dcdbas sdhci_pci snd sdhci cfg80211 soundcore parport serio_raw > snd_page_alloc fglrx(P) led_class r8169 mii video output > > Pid: 826, comm: rpciod/5 Tainted: P 2.6.31-15-generic > #50-Ubuntu Studio 1557 > > RIP: 0010:[<ffffffffa03c3c22>] [<ffffffffa03c3c22>] > rpcauth_checkverf+0x32/0x70 [sunrpc] > > RSP: 0000:ffff88012e02bd50 EFLAGS: 00010246 > > RAX: 0000000400001000 RBX: ffff88012dc86ec8 RCX: 000000000000080f > > RDX: 0000000000000000 RSI: ffff880129c141c8 RDI: ffff88012dc86ec8 > > RBP: ffff88012e02bd70 R08: ffff88012e02a000 R09: 0000026edec938f8 > > R10: 0000000000000000 R11: 00000000ffffffff R12: ffff88012692e540 > > R13: ffff880129c141c8 R14: ffff88012a508000 R15: ffffc90011124308 > > FS: 0000000000000000(0000) GS:ffff8800280b9000(0000) knlGS:0000000000000000 > > CS: 0010 DS: 0018 ES: 0018 CR0: 000000008005003b > > CR2: 0000000400001038 CR3: 000000009eaae000 CR4: 00000000000026e0 > > DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 > > DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 > > Process rpciod/5 (pid: 826, threadinfo ffff88012e02a000, task ffff88012e030000) > > Stack: > > ffff8801319b4a00 ffff88012dc86ec8 ffff88013092e000 ffff88013092e000 > > <0> ffff88012e02bdb0 ffffffffa03bb415 0000000000000010 0000000000000202 > > <0> ffff88012e02bda0 ffff88012dc86ec8 ffff88013092e000 ffffffffa04d4ff0 > > Call Trace: > > [<ffffffffa03bb415>] rpc_verify_header+0x1b5/0x5a0 [sunrpc] > > [<ffffffffa04d4ff0>] ? nfs4_xdr_dec_read+0x0/0x80 [nfs] > > [<ffffffffa03bb930>] call_decode+0x130/0x220 [sunrpc] > > [<ffffffffa03c2e4a>] __rpc_execute+0xba/0x220 [sunrpc] > > [<ffffffffa03c2fe0>] ? rpc_async_schedule+0x0/0x20 [sunrpc] > > [<ffffffffa03c2ff0>] rpc_async_schedule+0x10/0x20 [sunrpc] > > [<ffffffff810737a5>] run_workqueue+0x95/0x170 > > [<ffffffff81073924>] worker_thread+0xa4/0x120 > > [<ffffffff81078b30>] ? autoremove_wake_function+0x0/0x40 > > [<ffffffff81073880>] ? worker_thread+0x0/0x120 > > [<ffffffff81078746>] kthread+0xa6/0xb0 > > [<ffffffff810130ea>] child_rip+0xa/0x20 > > [<ffffffff810786a0>] ? kthread+0x0/0xb0 > > [<ffffffff810130e0>] ? child_rip+0x0/0x20 > > Code: 20 f6 05 51 d6 02 00 10 48 89 5d e8 4c 89 6d f8 48 89 fb 4c 89 > 65 f0 49 89 f5 4c 8b 67 50 75 1c 49 8b 44 24 38 4c 89 ee 48 89 df <ff> > 50 38 48 8b 5d e8 4c 8b 65 f0 4c 8b 6d f8 c9 c3 49 8b 44 24 > > RIP [<ffffffffa03c3c22>] rpcauth_checkverf+0x32/0x70 [sunrpc] > > RSP <ffff88012e02bd50> > > CR2: 0000000400001038 > > ---[ end trace c9ba33d8dceb9bfc ]--- > > > --- [2] > > x1:/ /net nfs4 rw,relatime,vers=4,rsize=262144,wsize=262144,namlen=255,acregmin=30,hard,proto=tcp,timeo=600,retrans=2,sec=sys,clientaddr=192.168.10.2,addr=192.168.10.250 > 0 0 ...and it is reproducible without the fglrx module? Sorry, but I do have to check that... The other question is, whether this always happens in the read and write code? If so, does it go away when you reduce the rsize/wsize to 128k? Trond -- To unsubscribe from this list: send the line "unsubscribe linux-nfs" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
