2.6.31.6: XFS DEBUG: Assertions cause kernel OOPS.

[Dave Chinner]

On Tue, Dec 01, 2009 at 01:39:06PM -0500, Justin Piszcz wrote:
>
>
> On Tue, 1 Dec 2009, Dave Chinner wrote:
>
>> On Mon, Nov 30, 2009 at 08:22:03AM -0500, Justin Piszcz wrote:
>>> Hi,
>>>
>>> I wanted to compile my kernel with DEBUG for XFS and also kernel frame pointers
>>> to catch any issues.
>>>
>>> However, DEBUG for XFS does more harm than good?
>>
>> DEBUG is there for developers, not end users. Often the debug code
>> will assert fail or panic where the situation is not fatal but
>> indicative of some problem that should be looked into further.
> Dave,
>
> Recall that issue that I was having with asterisk installed?

Yes, it appeared to be caused by log IO not completing, right?

> How should  
> one get more information on a filesystem lockup if we cannot have debug  
> enabled for the filesystem?

The debug code in XFS won't tell you anything extra about a lock-up
unless it trips an ASSERT before the lockup that is related to the
lock-up. My experience with hangs like the above are that they
generally are not related to the filesystem at all; most of the time
it is IO not being completed by the lower layers for some reason,
and no amount of filesystem level debug will help solve that
problem.

> Should there be a user-debug option which it 
> will include more verbosity if/when the xfs kernel processes lockup?

See /proc/sys/fs/xfs/error_level. That can turn up the verbosity of
error reporting without needing DEBUG compiled in.

Cheers,

Dave.
-- 
Dave Chinner
david at fromorbit.com