On Fri, Oct 18, 2024 at 08:22:38AM +1100, NeilBrown wrote:
> On Fri, 18 Oct 2024, Jeff Layton wrote:
> > On Thu, 2024-10-17 at 19:16 +0000, Chuck Lever III wrote:
> > >
> > > > On Oct 17, 2024, at 3:13 PM, Jeff Layton <jlayton@xxxxxxxxxx> wrote:
> > > >
> > > > On Thu, 2024-10-17 at 09:36 -0400, cel@xxxxxxxxxx wrote:
> > > > > From: Chuck Lever <chuck.lever@xxxxxxxxxx>
> > > > >
> > > > > Since commit 75c7940d2a86 ("lockd: set missing fl_flags field when
> > > > > retrieving args"), nlmsvc_retrieve_args() initializes the flc_flags
> > > > > field. svcxdr_decode_lock() no longer needs to do this.
> > > > >
> > > > > This clean up removes one dependency on the nlm_lock:fl field. No
> > > > > behavior change is expected.
> > > > >
> > > > > Analysis:
> > > > >
> > > > > svcxdr_decode_lock() is called by:
> > > > >
> > > > > nlm4svc_decode_testargs()
> > > > > nlm4svc_decode_lockargs()
> > > > > nlm4svc_decode_cancargs()
> > > > > nlm4svc_decode_unlockargs()
> > > > >
> > > > > nlm4svc_decode_testargs() is used by:
> > > > > - NLMPROC4_TEST and NLMPROC4_TEST_MSG, which call nlmsvc_retrieve_args()
> > > > > - NLMPROC4_GRANTED and NLMPROC4_GRANTED_MSG, which don't pass the
> > > > > lock's file_lock to the generic lock API
> > > > >
> > > > > nlm4svc_decode_lockargs() is used by:
> > > > > - NLMPROC4_LOCK and NLM4PROC4_LOCK_MSG, which call nlmsvc_retrieve_args()
> > > > > - NLMPROC4_UNLOCK and NLM4PROC4_UNLOCK_MSG, which call nlmsvc_retrieve_args()
> > > > > - NLMPROC4_NM_LOCK, which calls nlmsvc_retrieve_args()
> > > > >
> > > > > nlm4svc_decode_cancargs() is used by:
> > > > > - NLMPROC4_CANCEL and NLMPROC4_CANCEL_MSG, which call nlmsvc_retrieve_args()
> > > > >
> > > > > nlm4svc_decode_unlockargs() is used by:
> > > > > - NLMPROC4_UNLOCK and NLMPROC4_UNLOCK_MSG, which call nlmsvc_retrieve_args()
> > > > >
> > > > > All callers except GRANTED/GRANTED_MSG eventually call
> > > > > nlmsvc_retrieve_args() before using nlm_lock::fl.c.flc_flags. Thus
> > > > > this change is safe.
> > > > >
> > > > > Signed-off-by: Chuck Lever <chuck.lever@xxxxxxxxxx>
> > > > > ---
> > > > > fs/lockd/svc4proc.c | 5 +++--
> > > > > fs/lockd/xdr4.c | 1 -
> > > > > 2 files changed, 3 insertions(+), 3 deletions(-)
> > > > >
> > > > > diff --git a/fs/lockd/svc4proc.c b/fs/lockd/svc4proc.c
> > > > > index 2cb603013111..109e5caae8c7 100644
> > > > > --- a/fs/lockd/svc4proc.c
> > > > > +++ b/fs/lockd/svc4proc.c
> > > > > @@ -46,14 +46,15 @@ nlm4svc_retrieve_args(struct svc_rqst *rqstp, struct nlm_args *argp,
> > > > > if (filp != NULL) {
> > > > > int mode = lock_to_openmode(&lock->fl);
> > > > >
> > > > > + lock->fl.c.flc_flags = FL_POSIX;
> > > > > +
> > > > > error = nlm_lookup_file(rqstp, &file, lock);
> > > > > if (error)
> > > > > goto no_locks;
> > > > > *filp = file;
> > > > >
> > > > > /* Set up the missing parts of the file_lock structure */
> > > > > - lock->fl.c.flc_flags = FL_POSIX;
> > > > > - lock->fl.c.flc_file = file->f_file[mode];
> > > > > + lock->fl.c.flc_file = file->f_file[mode];
> > > > > lock->fl.c.flc_pid = current->tgid;
> > > > > lock->fl.fl_start = (loff_t)lock->lock_start;
> > > > > lock->fl.fl_end = lock->lock_len ?
> > > > > diff --git a/fs/lockd/xdr4.c b/fs/lockd/xdr4.c
> > > > > index 60466b8bac58..e343c820301f 100644
> > > > > --- a/fs/lockd/xdr4.c
> > > > > +++ b/fs/lockd/xdr4.c
> > > > > @@ -89,7 +89,6 @@ svcxdr_decode_lock(struct xdr_stream *xdr, struct nlm_lock *lock)
> > > > > return false;
> > > > >
> > > > > locks_init_lock(fl);
> > > > > - fl->c.flc_flags = FL_POSIX;
> > > > > fl->c.flc_type = F_RDLCK;
> > > > > nlm4svc_set_file_lock_range(fl, lock->lock_start, lock->lock_len);
> > > > > return true;
> > > >
> > > > 1-4 look fine. You can add my R-b to those.
> > >
> > > Thanks!
> > >
> > >
> > > > For this one, I think I'd rather see this go the other way, and just
> > > > eliminate the setting of flc_flags in nlm4svc_retrieve_args. We only
> > > > deal with FL_POSIX locks in svc lockd, and that does it right after
> > > > locks_init_lock, so I think that means it'll be done earlier, no?
> > >
> > > Have a look at the nlm4 branch in my kernel.org <http://kernel.org/> repo to see where
> > > this is headed.
> > >
> >
> > (For everyone following along: It's actually in Chuck's xdrgen branch)
> >
> > Oh ok, I see. This is an interim step toward moving all of the lock
> > initialization into nlm4svc_retrieve_args(). That probably is better. I
> > withdraw my objection.
>
> Adding that observation to the commit message would help with review.
I was hoping to break these up to make them easier to review, but
for 5/5 I guess it didn't help.
I can send more of these clean-up/refactors along for v6.13. But I'm
not sure the actual conversion later in the xdrgen branch is ready
for prime time.
> I agree that moving the initialisation to nlm4svc_retrieve_args() seems
> sensible. The half-way version in this patch looks really odd without
> that explanation.
>
> But the whole series
> Reviewed-by: NeilBrown <neilb@xxxxxxx>
>
> Thanks,
> NeilBrown
>
>
>
> >
> > >
> > > > Also, I think the same duplication is in nlmsvc_retrieve_args and the
> > > > nlmv3 version of svcxdr_decode_lock.
> > >
> > > Which is going away when NFSv2 is removed. I'm not too concerned
> > > about that duplication.
> > >
> >
> > Fair enough. I'm fine with leaving that to wither for now:
> >
> > Reviewed-by: Jeff Layton <jlayton@xxxxxxxxxx>
> >
>
--
Chuck Lever
