[Cluster-devel] Re: [PATCH 1/2] dlm: initialize file_lock struct in GETLK before copying conflicting lock

[J. Bruce Fields <bfields@fieldses.org>]

On Wed, Jan 21, 2009 at 09:26:08PM -0500, Jeff Layton wrote:
> On Wed, 21 Jan 2009 18:42:39 -0500
> "J. Bruce Fields" <bfields@fieldses.org> wrote:
> 
> > On Wed, Jan 21, 2009 at 11:34:50AM -0500, Jeff Layton wrote:
> > > dlm_posix_get fills out the relevant fields in the file_lock before
> > > returning when there is a lock conflict, but doesn't clean out any of
> > > the other fields in the file_lock.
> > > 
> > > When nfsd does a NFSv4 lockt call, it sets the fl_lmops to
> > > nfsd_posix_mng_ops before calling the lower fs. When the lock comes back
> > > after testing a lock on GFS2, it still has that field set. This confuses
> > > nfsd into thinking that the file_lock is a nfsd4 lock.
> > 
> > I think of the lock system as supporting two types of objects, both
> > stored in "struct lock"'s:
> > 
> > 	- Heavyweight locks: these have callbacks set and the filesystem
> > 	  or lock manager could in theory have some private data
> > 	  associated with them, so it's important that the appropriate
> > 	  callbacks be called when they're released or copied.  These
> > 	  are what are actually passed to posix_lock_file() and kept on
> > 	  the inode lock lists.
> > 	- Lightweight locks: just start, end, pid, flags, and type, with
> > 	  everything zeroed out and/or ignored.
> > 
> > I don't see any reason why the lock passed into dlm_posix_get() needs to
> > be a heavyweight lock.  In any case, if it were, then dlm_posix_get()
> > would need to release the passed-in-lock before initializing the new one
> > that it's returning.
> > 
> 
> 
> From what I can tell, dlm_posix_lock is always passed a "lightweight"
> lock.

Right, so in your second patch, I think the fl_lmops assignment in
nfsd4_lockt should also be removed.

> > The returned lock should probably also be a lightweight lock that's a
> > copy of whatever conflicting lock was found; otherwise we need to
> > require the caller to for example release the thing correctly.
> > 
> > That's unfortunate for nfsv4 since that doesn't allow returning the
> > lockowner information to the client.  But it's not terribly important
> > to get that right.
> > 
> > Since gfs2 doesn't report the conflicting lock, I guess we just punt and
> > return a copy of the passed-in lock, OK.
> > 
> 
> I'm not sure I follow you here...
> 
> GFS2/DLM does report the conflicting lock. It's just that when there is
> one, it's only overwriting some of the fields in the lock.

Whoops, sorry, OK.

> The idea with this patch is to basically try and make dlm_posix_get()
> fill out the same fields as __locks_copy_lock() and make sure the rest
> are initialized.

Yes, this patch seems fine.  I'm less sure of the second.

--b.