linux-nfs.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: "J. Bruce Fields" <bfields@fieldses.org>
To: Benny Halevy <bhalevy@tonian.com>
Cc: Benny Halevy <benny@tonian.com>,
	"J. Bruce Fields" <bfields@redhat.com>,
	linux-nfs@vger.kernel.org
Subject: Re: [PATCH 5/7] nfsd4: implement NFS4_SHARE_WANT_NO_DELEG, NFS4_OPEN_DELEGATE_NONE_EXT
Date: Thu, 20 Oct 2011 08:47:13 -0400	[thread overview]
Message-ID: <20111020124713.GP5444@fieldses.org> (raw)
In-Reply-To: <1456421465-1319114189-cardhu_decombobulator_blackberry.rim.net-173541571-@b18.c3.bise3.blackberry>

On Thu, Oct 20, 2011 at 12:36:25PM +0000, Benny Halevy wrote:
> OK, I read you.  Just please take note that WANT_NO_DELEG flag is important to simplified clients that do not support delegations, like the one I was testing against in the BAT, and it is the only one in use that I could tell.

If you can suggest an implementation that meets the minimum required by
the spec, I'd happily take it....

It might not be much harder: it's probably just a matter of returrning
the right OPEN_DELEGATE_NONE_EXT?

Though unfortunately there's not a "why" for "I just didn't feel like
giving you one", so we probably would have to make an effort to detect
the different case.

Also, I wonder if it's wise for a client to depend on this--it really is
an optional feature as far as I can tell.  It would be safer to teach
the client just to handle delegations in the simplest way possible
(possibly just return them immediately?).

--b.

> 
> Benny
> -----Original Message-----
> From: "J. Bruce Fields" <bfields@fieldses.org>Date: Thu, 20 Oct 2011 07:52:24 
> To: Benny Halevy<benny@tonian.com>
> Cc: J. Bruce Fields<bfields@redhat.com>; <linux-nfs@vger.kernel.org>; Benny Halevy<bhalevy@tonian.com>
> Subject: Re: [PATCH 5/7] nfsd4: implement NFS4_SHARE_WANT_NO_DELEG,
>  NFS4_OPEN_DELEGATE_NONE_EXT
> 
> On Wed, Oct 19, 2011 at 07:13:21PM -0700, Benny Halevy wrote:
> > From: Benny Halevy <bhalevy@tonian.com>
> > 
> > Respect client request for not getting a delegation in NFSv4.1
> > Appropriately return delegation "type" NFS4_OPEN_DELEGATE_NONE_EXT
> > and WND4_NOT_WANTED reason.
> 
> As I read it, if we're going to implement part of this, then we have to
> implement all of it.  ("If the server supports the new _WANT_ flags and
> the client sends one or more of the new flags, then in the event the
> server does not return a delegation, it MUST return a delegation type of
> OPEN_DELEGATE_NONE_EXT.  The field ond_why in the reply indicates why no
> delegation was returned and will be one of....")
> 
> As long as we don't implement any want flags, we escape that MUST.
> 
> So I'd rather put this aside till it's more complete.
> 
> --b.
> 
> > 
> > Signed-off-by: Benny Halevy <bhalevy@tonian.com>
> > ---
> >  fs/nfsd/nfs4state.c  |   10 ++++++++--
> >  fs/nfsd/nfs4xdr.c    |    3 +++
> >  include/linux/nfs4.h |   15 ++++++++++++++-
> >  3 files changed, 25 insertions(+), 3 deletions(-)
> > 
> > diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c
> > index ec361bb..5c3377c 100644
> > --- a/fs/nfsd/nfs4state.c
> > +++ b/fs/nfsd/nfs4state.c
> > @@ -2931,15 +2931,21 @@ static int nfs4_set_delegation(struct nfs4_delegation *dp, int flag)
> >  	update_stateid(&stp->st_stid.sc_stateid);
> >  	memcpy(&open->op_stateid, &stp->st_stid.sc_stateid, sizeof(stateid_t));
> >  
> > -	if (nfsd4_has_session(&resp->cstate))
> > +	if (nfsd4_has_session(&resp->cstate)) {
> >  		open->op_openowner->oo_flags |= NFS4_OO_CONFIRMED;
> >  
> > +		if (open->op_share_access & NFS4_SHARE_WANT_NO_DELEG) {
> > +			open->op_delegate_type = NFS4_OPEN_DELEGATE_NONE_EXT;
> > +			goto nodeleg;
> > +		}
> > +	}
> > +
> >  	/*
> >  	* Attempt to hand out a delegation. No error return, because the
> >  	* OPEN succeeds even if we fail.
> >  	*/
> >  	nfs4_open_delegation(current_fh, open, stp);
> > -
> > +nodeleg:
> >  	status = nfs_ok;
> >  
> >  	dprintk("%s: stateid=" STATEID_FMT "\n", __func__,
> > diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c
> > index 706ada1..1a419c0 100644
> > --- a/fs/nfsd/nfs4xdr.c
> > +++ b/fs/nfsd/nfs4xdr.c
> > @@ -2842,6 +2842,9 @@ static __be32 nfsd4_encode_bind_conn_to_session(struct nfsd4_compoundres *resp,
> >  		WRITE32(0);   /* XXX: is NULL principal ok? */
> >  		ADJUST_ARGS();
> >  		break;
> > +	case NFS4_OPEN_DELEGATE_NONE_EXT: /* 4.1 */
> > +		WRITE32(WND4_NOT_WANTED);	/* only reason for now */
> > +		break;
> >  	default:
> >  		BUG();
> >  	}
> > diff --git a/include/linux/nfs4.h b/include/linux/nfs4.h
> > index 32345c2..8cdde4d 100644
> > --- a/include/linux/nfs4.h
> > +++ b/include/linux/nfs4.h
> > @@ -441,7 +441,20 @@ enum limit_by4 {
> >  enum open_delegation_type4 {
> >  	NFS4_OPEN_DELEGATE_NONE = 0,
> >  	NFS4_OPEN_DELEGATE_READ = 1,
> > -	NFS4_OPEN_DELEGATE_WRITE = 2
> > +	NFS4_OPEN_DELEGATE_WRITE = 2,
> > +	NFS4_OPEN_DELEGATE_NONE_EXT = 3, /* 4.1 */
> > +};
> > +
> > +enum why_no_delegation4 { /* new to v4.1 */
> > +	WND4_NOT_WANTED = 0,
> > +	WND4_CONTENTION = 1,
> > +	WND4_RESOURCE = 2,
> > +	WND4_NOT_SUPP_FTYPE = 3,
> > +	WND4_WRITE_DELEG_NOT_SUPP_FTYPE = 4,
> > +	WND4_NOT_SUPP_UPGRADE = 5,
> > +	WND4_NOT_SUPP_DOWNGRADE = 6,
> > +	WND4_CANCELLED = 7,
> > +	WND4_IS_DIR = 8,
> >  };
> >  
> >  enum lock_type4 {
> > -- 
> > 1.7.6
> > 

  reply	other threads:[~2011-10-20 12:47 UTC|newest]

Thread overview: 18+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-10-20  2:10 [PATCH 0/7] Bakeathon server fixes Benny Halevy
2011-10-20  2:12 ` [PATCH 1/7] nfsd4: implement new 4.1 open reclaim types Benny Halevy
2011-10-20 12:23   ` J. Bruce Fields
2011-10-20  2:12 ` [PATCH 2/7] nfsd41: use SEQ4_STATUS_BACKCHANNEL_FAULT when cb_sequence is invalid Benny Halevy
2011-10-20  2:13 ` [PATCH 3/7] nfsd4: seq->status_flags may be used unitialized Benny Halevy
2011-10-20  2:13 ` [PATCH 4/7] nfsd4: allow NFS4_SHARE_SIGNAL_DELEG_WHEN_RESRC_AVAIL | NFS4_SHARE_PUSH_DELEG_WHEN_UNCONTENDED Benny Halevy
2011-10-20  2:13 ` [PATCH 5/7] nfsd4: implement NFS4_SHARE_WANT_NO_DELEG, NFS4_OPEN_DELEGATE_NONE_EXT Benny Halevy
2011-10-20 11:52   ` J. Bruce Fields
2011-10-20 12:36     ` Benny Halevy
2011-10-20 12:47       ` J. Bruce Fields [this message]
2011-10-20 19:16         ` Benny Halevy
2011-10-20 19:21           ` J. Bruce Fields
2011-10-20  2:13 ` [PATCH 6/7] nfsd4: typo logical vs bitwise negate for want_mask Benny Halevy
2011-10-20 11:53   ` J. Bruce Fields
2011-10-20  2:13 ` [PATCH 7/7] nfsd4: share_access_to_flags should consider only nfs4.x share_access flags Benny Halevy
2011-10-20 11:56   ` J. Bruce Fields
2011-10-20 18:59     ` Benny Halevy
2011-10-20 11:58 ` [PATCH 0/7] Bakeathon server fixes J. Bruce Fields

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20111020124713.GP5444@fieldses.org \
    --to=bfields@fieldses.org \
    --cc=benny@tonian.com \
    --cc=bfields@redhat.com \
    --cc=bhalevy@tonian.com \
    --cc=linux-nfs@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).