Re: reiserfs deadlock

[Frederic Weisbecker]

On Fri, Feb 05, 2010 at 04:46:38AM +0300, Alexander Beregalov wrote:
> Hi Frederic
> 
> Does it help?


Nope :-S

I need the full stracktraces. Plus the fact a thread holds
the reiserfs lock doesn't help much in this regard. This
is a drawback of lockdep: "holding" a lock in such lock dump
means either: the task holds the lock _or_ the task requests the
lock and block on it.

The only way to determine if a task really holds the reiserfs
lock is to look at the stack trace and see if it blocks on
reiserfs_write_lock_*() or something else. If this is something
else, then it means it really holds it, and there are fair
chances this is exactly the stacktrace we want: the one that
raised the softlockup.

Thanks!

--
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/