linux-nfs.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Trond Myklebust <trondmy@hammerspace.com>
To: "chuck.lever@oracle.com" <chuck.lever@oracle.com>
Cc: "linux-rdma@vger.kernel.org" <linux-rdma@vger.kernel.org>,
	"linux-nfs@vger.kernel.org" <linux-nfs@vger.kernel.org>
Subject: Re: [PATCH v4 06/30] xprtrdma: Don't wake pending tasks until disconnect is done
Date: Mon, 17 Dec 2018 18:55:27 +0000	[thread overview]
Message-ID: <c639c203e1c9a5c0016a46e95ceb690744480583.camel@hammerspace.com> (raw)
In-Reply-To: <FD0CDDF9-EDEF-41C6-83F5-C956F3FF098E@oracle.com>

On Mon, 2018-12-17 at 13:37 -0500, Chuck Lever wrote:
> > On Dec 17, 2018, at 12:28 PM, Trond Myklebust <
> > trondmy@hammerspace.com> wrote:
> > 
> > On Mon, 2018-12-17 at 11:39 -0500, Chuck Lever wrote:
> > > Transport disconnect processing does a "wake pending tasks" at
> > > various points.
> > > 
> > > Suppose an RPC Reply is being processed. The RPC task that Reply
> > > goes with is waiting on the pending queue. If a disconnect wake-
> > > up
> > > happens before reply processing is done, that reply, even if it
> > > is
> > > good, is thrown away, and the RPC has to be sent again.
> > > 
> > > This window apparently does not exist for socket transports
> > > because
> > > there is a lock held while a reply is being received which
> > > prevents
> > > the wake-up call until after reply processing is done.
> > > 
> > > To resolve this, all RPC replies being processed on an RPC-over-
> > > RDMA
> > > transport have to complete before pending tasks are awoken due to
> > > a
> > > transport disconnect.
> > > 
> > > Callers that already hold the transport write lock may invoke
> > > ->ops->close directly. Others use a generic helper that schedules
> > > a close when the write lock can be taken safely.
> > > 
> > > Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
> > > ---
> > > include/linux/sunrpc/xprt.h                |    1 +
> > > net/sunrpc/xprt.c                          |   19
> > > +++++++++++++++++++
> > > net/sunrpc/xprtrdma/backchannel.c          |   13 +++++++------
> > > net/sunrpc/xprtrdma/svc_rdma_backchannel.c |    8 +++++---
> > > net/sunrpc/xprtrdma/transport.c            |   16 ++++++++++-----
> > > -
> > > net/sunrpc/xprtrdma/verbs.c                |    5 ++---
> > > 6 files changed, 44 insertions(+), 18 deletions(-)
> > > 
> > > diff --git a/include/linux/sunrpc/xprt.h
> > > b/include/linux/sunrpc/xprt.h
> > > index a4ab4f8..ee94ed0 100644
> > > --- a/include/linux/sunrpc/xprt.h
> > > +++ b/include/linux/sunrpc/xprt.h
> > > @@ -401,6 +401,7 @@ static inline __be32
> > > *xprt_skip_transport_header(struct rpc_xprt *xprt, __be32 *
> > > bool			xprt_request_get_cong(struct rpc_xprt
> > > *xprt,
> > > struct rpc_rqst *req);
> > > void			xprt_disconnect_done(struct rpc_xprt
> > > *xprt);
> > > void			xprt_force_disconnect(struct rpc_xprt
> > > *xprt);
> > > +void			xprt_disconnect_nowake(struct rpc_xprt
> > > *xprt);
> > > void			xprt_conditional_disconnect(struct
> > > rpc_xprt
> > > *xprt, unsigned int cookie);
> > > 
> > > bool			xprt_lock_connect(struct rpc_xprt *,
> > > struct
> > > rpc_task *, void *);
> > > diff --git a/net/sunrpc/xprt.c b/net/sunrpc/xprt.c
> > > index ce92700..afe412e 100644
> > > --- a/net/sunrpc/xprt.c
> > > +++ b/net/sunrpc/xprt.c
> > > @@ -685,6 +685,25 @@ void xprt_force_disconnect(struct rpc_xprt
> > > *xprt)
> > > }
> > > EXPORT_SYMBOL_GPL(xprt_force_disconnect);
> > > 
> > > +/**
> > > + * xprt_disconnect_nowake - force a call to xprt->ops->close
> > > + * @xprt: transport to disconnect
> > > + *
> > > + * The caller must ensure that xprt_wake_pending_tasks() is
> > > + * called later.
> > > + */
> > > +void xprt_disconnect_nowake(struct rpc_xprt *xprt)
> > > +{
> > > +       /* Don't race with the test_bit() in xprt_clear_locked()
> > > */
> > > +       spin_lock_bh(&xprt->transport_lock);
> > > +       set_bit(XPRT_CLOSE_WAIT, &xprt->state);
> > > +       /* Try to schedule an autoclose RPC call */
> > > +       if (test_and_set_bit(XPRT_LOCKED, &xprt->state) == 0)
> > > +               queue_work(xprtiod_workqueue, &xprt-
> > > >task_cleanup);
> > > +       spin_unlock_bh(&xprt->transport_lock);
> > > +}
> > > +EXPORT_SYMBOL_GPL(xprt_disconnect_nowake);
> > > +
> > 
> > We shouldn't need both xprt_disconnect_nowake() and
> > xprt_force_disconnect() to be exported given that you can build the
> > latter from the former + xprt_wake_pending_tasks() (which is also
> > already exported).
> 
> Thanks for your review!
> 
> I can get rid of xprt_disconnect_nowake. There are some variations,
> depending on why wake_pending_tasks is protected by xprt-
> >transport_lock.

I'm having some second thoughts about the patch that Scott sent out
last week to fix the issue that Dave and he were seeing. I think that
what we really need to do to fix his issue is to call
xprt_disconnect_done() after we've released the TCP socket.

Given that realisation, I think that we can fix up
xprt_force_disconnect() to only wake up the task that holds the
XPRT_LOCKED instead of doing a thundering herd wakeup like we do today.
That should (I think) obviate the need for a separate
xprt_disconnect_nowake().

A patch is forthcoming later today. I'll make sure you are Cced so you
can comment.

-- 
Trond Myklebust
Linux NFS client maintainer, Hammerspace
trond.myklebust@hammerspace.com



  reply	other threads:[~2018-12-17 18:55 UTC|newest]

Thread overview: 40+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-12-17 16:39 [PATCH v4 00/30] NFS/RDMA client for next Chuck Lever
2018-12-17 16:39 ` [PATCH v4 01/30] xprtrdma: Yet another double DMA-unmap Chuck Lever
2018-12-17 16:39 ` [PATCH v4 02/30] xprtrdma: Ensure MRs are DMA-unmapped when posting LOCAL_INV fails Chuck Lever
2018-12-17 16:39 ` [PATCH v4 03/30] xprtrdma: Refactor Receive accounting Chuck Lever
2018-12-17 16:39 ` [PATCH v4 04/30] xprtrdma: Replace rpcrdma_receive_wq with a per-xprt workqueue Chuck Lever
2018-12-17 16:39 ` [PATCH v4 05/30] xprtrdma: No qp_event disconnect Chuck Lever
2018-12-17 16:39 ` [PATCH v4 06/30] xprtrdma: Don't wake pending tasks until disconnect is done Chuck Lever
2018-12-17 17:28   ` Trond Myklebust
2018-12-17 18:37     ` Chuck Lever
2018-12-17 18:55       ` Trond Myklebust [this message]
2018-12-17 19:00         ` Chuck Lever
2018-12-17 19:09           ` Trond Myklebust
2018-12-17 19:19             ` Chuck Lever
2018-12-17 19:26               ` Trond Myklebust
2018-12-17 16:39 ` [PATCH v4 07/30] xprtrdma: Fix ri_max_segs and the result of ro_maxpages Chuck Lever
2018-12-18 19:35   ` Anna Schumaker
2018-12-18 19:39     ` Chuck Lever
2018-12-17 16:40 ` [PATCH v4 08/30] xprtrdma: Reduce max_frwr_depth Chuck Lever
2018-12-17 16:40 ` [PATCH v4 09/30] xprtrdma: Remove support for FMR memory registration Chuck Lever
2018-12-17 16:40 ` [PATCH v4 10/30] xprtrdma: Remove rpcrdma_memreg_ops Chuck Lever
2018-12-17 16:40 ` [PATCH v4 11/30] xprtrdma: Plant XID in on-the-wire RDMA offset (FRWR) Chuck Lever
2018-12-17 16:40 ` [PATCH v4 12/30] NFS: Make "port=" mount option optional for RDMA mounts Chuck Lever
2018-12-17 16:40 ` [PATCH v4 13/30] xprtrdma: Recognize XDRBUF_SPARSE_PAGES Chuck Lever
2018-12-17 16:40 ` [PATCH v4 14/30] xprtrdma: Remove request_module from backchannel Chuck Lever
2018-12-17 16:40 ` [PATCH v4 15/30] xprtrdma: Expose transport header errors Chuck Lever
2018-12-17 16:40 ` [PATCH v4 16/30] xprtrdma: Simplify locking that protects the rl_allreqs list Chuck Lever
2018-12-17 16:40 ` [PATCH v4 17/30] xprtrdma: Cull dprintk() call sites Chuck Lever
2018-12-17 16:40 ` [PATCH v4 18/30] xprtrdma: Remove unused fields from rpcrdma_ia Chuck Lever
2018-12-17 16:41 ` [PATCH v4 19/30] xprtrdma: Clean up of xprtrdma chunk trace points Chuck Lever
2018-12-17 16:41 ` [PATCH v4 20/30] xprtrdma: Relocate the xprtrdma_mr_map " Chuck Lever
2018-12-17 16:41 ` [PATCH v4 21/30] xprtrdma: Add trace points for calls to transport switch methods Chuck Lever
2018-12-17 16:41 ` [PATCH v4 22/30] xprtrdma: Trace mapping, alloc, and dereg failures Chuck Lever
2018-12-17 16:41 ` [PATCH v4 23/30] NFS: Fix NFSv4 symbolic trace point output Chuck Lever
2018-12-17 16:41 ` [PATCH v4 24/30] SUNRPC: Simplify defining common RPC trace events Chuck Lever
2018-12-17 16:41 ` [PATCH v4 25/30] SUNRPC: Fix some kernel doc complaints Chuck Lever
2018-12-17 16:41 ` [PATCH v4 26/30] xprtrdma: Update comments in frwr_op_send Chuck Lever
2018-12-17 16:41 ` [PATCH v4 27/30] xprtrdma: Replace outdated comment for rpcrdma_ep_post Chuck Lever
2018-12-17 16:41 ` [PATCH v4 28/30] xprtrdma: Add documenting comment for rpcrdma_buffer_destroy Chuck Lever
2018-12-17 16:41 ` [PATCH v4 29/30] xprtrdma: Clarify comments in rpcrdma_ia_remove Chuck Lever
2018-12-17 16:42 ` [PATCH v4 30/30] xprtrdma: Don't leak freed MRs Chuck Lever

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=c639c203e1c9a5c0016a46e95ceb690744480583.camel@hammerspace.com \
    --to=trondmy@hammerspace.com \
    --cc=chuck.lever@oracle.com \
    --cc=linux-nfs@vger.kernel.org \
    --cc=linux-rdma@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).