Re: strange stuff in dmesg

[AmÃrico Wang]

On Thu, Dec 24, 2009 at 11:22 PM, BuraphaLinux Server
<buraphalinuxserver@xxxxxxxxx> wrote:
> On 12/24/09, Justin P. Mattock <justinmattock@xxxxxxxxx> wrote:
>> On 12/24/09 02:03, BuraphaLinux Server wrote:
>>> On my Dell OptiPlex 330 machines with kernel 2.6.32.2 I get a strange
>>> WARNING. ÂDo I need to worry? ÂHere is the warning:
>>>
>>> [ Â Â0.000000] ------------[ cut here ]------------
>>> [ Â Â0.000000] WARNING: at mm/page_alloc.c:1805
>>> __alloc_pages_nodemask+0x1b6/0x730()
>>> [ Â Â0.000000] Hardware name: OptiPlex 330
>>> [ Â Â0.000000] Modules linked in:
>>> [ Â Â0.000000] Pid: 0, comm: swapper Not tainted 2.6.32.2 #1
>>> [ Â Â0.000000] Call Trace:
>>> [ Â Â0.000000] Â[<ffffffff8108e806>] ? __alloc_pages_nodemask+0x1b6/0x730
>>> [ Â Â0.000000] Â[<ffffffff81043f68>] warn_slowpath_common+0x78/0xd0
>>> [ Â Â0.000000] Â[<ffffffff81043fcf>] warn_slowpath_null+0xf/0x20
>>> [ Â Â0.000000] Â[<ffffffff8108e806>] __alloc_pages_nodemask+0x1b6/0x730
>>> [ Â Â0.000000] Â[<ffffffff810b4e28>] alloc_pages_current+0x78/0xf0
>>> [ Â Â0.000000] Â[<ffffffff8108da69>] __get_free_pages+0x9/0x50
>>> [ Â Â0.000000] Â[<ffffffff810bb912>] __kmalloc+0x112/0x120
>>> [ Â Â0.000000] Â[<ffffffff8146e773>] vgacon_scrollback_startup+0x13/0x70
>>> [ Â Â0.000000] Â[<ffffffff811d29b3>] vgacon_startup+0x2a3/0x420
>>> [ Â Â0.000000] Â[<ffffffff816fc556>] con_init+0x1b/0x230
>>> [ Â Â0.000000] Â[<ffffffff816fba00>] console_init+0x22/0x42
>>> [ Â Â0.000000] Â[<ffffffff816d4b8f>] start_kernel+0x240/0x3be
>>> [ Â Â0.000000] Â[<ffffffff816d4289>] x86_64_start_reservations+0x99/0xb9
>>> [ Â Â0.000000] Â[<ffffffff816d4389>] x86_64_start_kernel+0xe0/0xf2
>>> [ Â Â0.000000] ---[ end trace 4eaa2a86a8e2da22 ]---
>>>
>>> I also got it on 2.6.31.9, but had waited to ask hoping 2.6.32.2 would fix
>>> it.
>>>
>>> Attached is my config
>>>
>> was there a kernel that did not do this?
>> if so can you try a bisect on this?
>>
>> Justin P. Mattock
>>
>
> It took a while, but I have verified that 2.6.30.10 works without any
> message, and 2.6.31 has the error message (but otherwise seems to run
> ok). ÂThe hex codes are different, but the function names match and
> are in the same places.
>
> Does it have to be git bisect, or will trying the 2.6.31rc[1-9] be
> enough instead? ÂThe Documentation/BUG-HUNTING does not give detailed
> enough instructions for me to do the bisect thing.

Hey, the problem is that you configured VGACON_SOFT_SCROLLBACK_SIZE
to be CONFIG_VGACON_SOFT_SCROLLBACK_SIZE=65536, which will be
65536Kb, i.e. 16384 pages, how huge! :)

We shoud limit the number of VGACON_SOFT_SCROLLBACK_SIZE. I will cook
a patch now...
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/