[Xen-devel] Re: OOPS and panic on 2.6.29-rc1 on xen-x86

["Bryan Donlan" <bdonlan@xxxxxxxxx>]

On Tue, Jan 13, 2009 at 9:59 PM, Nick Piggin <npiggin@xxxxxxx> wrote:
> On Mon, Jan 12, 2009 at 11:54:32PM -0500, Bryan Donlan wrote:
>> On Mon, Jan 12, 2009 at 12:26 PM, Bryan Donlan <bdonlan@xxxxxxxxx> wrote:
>> > [resending with log/config inline as my previous message seems to have
>> >  been eaten by vger's spam filters]
>> >
>> > Hi,
>> >
>> > After testing 2.6.29-rc1 on xen-x86 with a btrfs root filesystem, I
>> > got the OOPS quoted below and a hard freeze shortly after boot.
>> > Boot messages and config are attached.
>> >
>> > This is on a test system, so I'd be happy to test any patches.
>> >
>> > Thanks,
>> >
>> > Bryan Donlan
>>
>> I've bisected the bug in question, and the faulty commit appears to be:
>> commit e97a630eb0f5b8b380fd67504de6cedebb489003
>> Author: Nick Piggin <npiggin@xxxxxxx>
>> Date:   Tue Jan 6 14:39:19 2009 -0800
>>
>>     mm: vmalloc use mutex for purge
>>
>>     The vmalloc purge lock can be a mutex so we can sleep while a purge is
>>     going on (purge involves a global kernel TLB invalidate, so it can take
>>     quite a while).
>>
>>     Signed-off-by: Nick Piggin <npiggin@xxxxxxx>
>>     Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
>>     Signed-off-by: Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx>
>>
>> The bug is easily reproducable by a kernel build on -j4 - it will
>> generally OOPS and panic before the build completes.
>> Also, I've tested it with ext3, and it still occurs, so it seems
>> unrelated to btrfs at least :)
>>
>> >
>> > ------------[ cut here ]------------
>> > Kernel BUG at c05ef80d [verbose debug info unavailable]
>> > invalid opcode: 0000 [#1] SMP
>> > last sysfs file: /sys/block/xvdc/size
>> > Modules linked in:
>
> It is bugging in schedule somehow, but you don't have verbose debug
> info compiled in. Can you compile that in and reproduce if you have
> the time?

Sure - which config option would that be? CONFIG_DEBUG_BUGVERBOSE?

> Going bug here might indicate that there is some other problem with
> the Xen and/or vmalloc code, regardless of reverting this patch.
>
> Thanks,
> Nick
>

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