Re: [PATCH 7/7] driver-core : convert semaphore to mutex in struct class

[Jarek Poplawski <jarkao2@gmail.com>]

On Thu, Jan 17, 2008 at 09:31:55PM +0100, Jarek Poplawski wrote:
> On Thu, Jan 17, 2008 at 02:57:36PM -0500, Alan Stern wrote:
...
> > If I recall correctly the nature of the warning was that a method
> > routine for one class (called with the class's mutex held) was creating
> > a second class and locking that class's mutex.  In principle this is
> > perfectly legal and should be allowed for arbitrary depths of nesting,
> > even though it is the sort of thing lockdep is currently unable to
> > handle.

BTW, Dave, if it's only about one such "second class" here, then it
shouldn't be so hard to try this one more level of nesting. I think,
the real problem for lockdep starts when there are more such "second
classes", but it's probably in another place.

You could also have a look at e.g. enum_inode_i_mutex_lock_class in
include/linux/fs.h and fh_lock_nested() in include/linux/nfsd/nfsfh.h,
and maybe define similar enum for this.

Jarek P.