Re: [Xen-devel] Process irqbalance: BUG: unable to handle kernel paging request - with 2.6.32.18 pv-ops & xen 4.0.1

[Bruce Edge <bruce.edge@xxxxxxxxx>]

On Tue, Aug 31, 2010 at 11:17 AM, Bruce Edge <bruce.edge@xxxxxxxxx> wrote:

On Tue, Aug 31, 2010 at 11:00 AM, Jeremy Fitzhardinge <jeremy@xxxxxxxx> wrote:

 On 08/30/2010 04:12 PM, Bruce Edge wrote:
> Using Jeremy's latest 2.6.32 pv-ops kernel for both dom0 and domU, I
> get this BUG at domU boot time.
>
> [   14.780699] BUG: unable to handle kernel paging request at
> 00000002dd253558
> [   14.780787] RAX: 0000000000000000 RBX: 0000000000000020 RCX:
> 0000000000000020
> [   14.780793] RDX: 0000000000000020 RSI: 0000000000000ffe RDI:
> 0000000000000000
> [   14.780799] RBP: ffff8800717bde68 R08: 0000000000000023 R09:
> 0000000000000001
> [   14.780806] R10: ffffffff8182d004 R11: 0000000000000000 R12:
> 0000000000000000
> [   14.780812] R13: 00000002dd253558 R14: 0000000000000ffe R15:
> ffff88000b515000
> [   14.780822] FS:  00007feac5f86720(0000) GS:ffff880011423000(0000)
> knlGS:0000000000000000
> [   14.780829] CS:  e033 DS: 0000 ES: 0000 CR0: 000000008005003b
> [   14.780834] CR2: 00000002dd253558 CR3: 00000000789db000 CR4:
> 0000000000002660
> [   14.780841] DR0: 0000000000000000 DR1: 0000000000000000 DR2:
> 0000000000000000
> [   14.780847] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7:
> 0000000000000400
> [   14.780854] Process irqbalance (pid: 880, threadinfo
> ffff8800717bc000, task ffff88007306c7e0)
> [   14.780861] Stack:
> [   14.780864]  ffffffff819180a0 ffff88000b515000 ffffffff81944798
> ffff88000a861da8
> [   14.780874] <0> ffff8800717bdf48 ffff88000aa780a0 ffff8800717bde88
> ffffffff8135760b
> [   14.780885] <0> ffff88000a861da8 fffffffffffffffb ffff8800717bdea8
> ffffffff814061ef
> [   14.780898] Call Trace:
> [   14.780905]  [<ffffffff8135760b>] local_cpus_show+0x4b/0x70
> [   14.780913]  [<ffffffff814061ef>] dev_attr_show+0x2f/0x70
> [   14.780921]  [<ffffffff811df8ed>] sysfs_read_file+0x10d/0x1f0
> [   14.781108]  [<ffffffff8116d018>] vfs_read+0xc8/0x1a0
> [   14.781114]  [<ffffffff8116d1f5>] sys_read+0x55/0x90
> [   14.781123]  [<ffffffff810151f2>] system_call_fastpath+0x16/0x1b
> [   14.781129] Code: 45 31 e4 83 e1 e0 89 cb 83 eb 20 78 62 66 90 89
> d8 89 d9 41 b9 01 00 00 00 c1 f8 06 83 e1 3f 44 8d 42 03 48 98 49 63
> fc 44 89 f6 <49> 8b 44 c5 00 49 8d 3c 3f 44 29 e6 41 c1 f8 02 48 d3 e8
> 89 d1
> [   14.781206] RIP  [<ffffffff8134019c>] bitmap_scnprintf+0x5c/0xc0
> [   14.781213]  RSP <ffff8800717bde38>
> [   14.781217] CR2: 00000002dd253558
> [   14.781224] ---[ end trace ec14b09e94a4d7d5 ]---
>
> The system appears stable after this, although no additional testing
> has been done - this is a fresh conversion from hvm to pvops and has
> zero mileage on it.
>
> dom0/domU config is attached - it's a slightly modified version of
> Pasi's recommended configs.

How many cpus do you have configured and online?

   J

I re-ran the same test with 1 cpu/vcpu and the BUG was _not_ present.

Incrementing cpus and vcpus to 2 did cause the problem.

If there anything else I can do to provide more information that would assist in troubleshooting, just ask.

-Bruce

 

Lots :-)

I'm allocating 14 cores to this domU:

    (vcpus 14)

    (cpus ((2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15)))

    (on_reboot restart)

    (on_poweroff destroy)

    (maxmem 2048)

    (memory 2048)

    (online_vcpus 14)

-Bruce

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-devel