Return-Path: Received: from mx1.redhat.com ([209.132.183.28]:60150 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750759AbdI1TVu (ORCPT ); Thu, 28 Sep 2017 15:21:50 -0400 Date: Thu, 28 Sep 2017 15:21:49 -0400 From: "J. Bruce Fields" To: Olga Kornievskaia Cc: linux-nfs@vger.kernel.org Subject: Re: [PATCH v4 10/10] NFSD stop queued async copies on client shutdown Message-ID: <20170928192148.GM10182@parsley.fieldses.org> References: <20170928172945.50780-1-kolga@netapp.com> <20170928172945.50780-11-kolga@netapp.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii In-Reply-To: <20170928172945.50780-11-kolga@netapp.com> Sender: linux-nfs-owner@vger.kernel.org List-ID: On Thu, Sep 28, 2017 at 01:29:45PM -0400, Olga Kornievskaia wrote: > If client is shutting down and there are still async copies going > on, then stop queued async copies. > > Signed-off-by: Olga Kornievskaia > --- > fs/nfsd/nfs4proc.c | 9 +++++++++ > fs/nfsd/nfs4state.c | 1 + > fs/nfsd/state.h | 2 ++ > 3 files changed, 12 insertions(+) > > diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c > index 74bfa4b..8d0c87f 100644 > --- a/fs/nfsd/nfs4proc.c > +++ b/fs/nfsd/nfs4proc.c > @@ -91,6 +91,12 @@ void nfsd4_destroy_copy_queue(void) > destroy_workqueue(copy_wq); > } > > +void nfsd4_shutdown_copy(struct nfs4_client *clp) > +{ > + set_bit(NFSD4_CLIENT_COPY_KILL, &clp->cl_flags); > + flush_workqueue(copy_wq); Looks like that will just stop new copies and wait on any ongoing ones. That could be a long wait. A signal should help there too. If you still need a "this client is dying" bit in cl_flags, you could probably just use CB_KILL instead of this new one. --b. > +} > + > #define NFSDDBG_FACILITY NFSDDBG_PROC > > static u32 nfsd_attrmask[] = { > @@ -1205,6 +1211,9 @@ static void nfsd4_do_async_copy(struct work_struct *work) > container_of(work, struct nfsd4_copy, cp_work); > struct nfsd4_copy *cb_copy; > > + if (test_bit(NFSD4_CLIENT_COPY_KILL, ©->cp_clp->cl_flags)) > + goto out; > + > copy->nfserr = nfsd4_do_copy(copy, 0); > > if (copy->cps->cp_cancelled) > diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c > index 97ab3f8..f4b193e 100644 > --- a/fs/nfsd/nfs4state.c > +++ b/fs/nfsd/nfs4state.c > @@ -1961,6 +1961,7 @@ static __be32 mark_client_expired_locked(struct nfs4_client *clp) > } > nfsd4_return_all_client_layouts(clp); > nfsd4_shutdown_callback(clp); > + nfsd4_shutdown_copy(clp); > if (clp->cl_cb_conn.cb_xprt) > svc_xprt_put(clp->cl_cb_conn.cb_xprt); > free_client(clp); > diff --git a/fs/nfsd/state.h b/fs/nfsd/state.h > index c31025c..55f09a9 100644 > --- a/fs/nfsd/state.h > +++ b/fs/nfsd/state.h > @@ -336,6 +336,7 @@ struct nfs4_client { > #define NFSD4_CLIENT_RECLAIM_COMPLETE (3) /* reclaim_complete done */ > #define NFSD4_CLIENT_CONFIRMED (4) /* client is confirmed */ > #define NFSD4_CLIENT_UPCALL_LOCK (5) /* upcall serialization */ > +#define NFSD4_CLIENT_COPY_KILL (6) /* stop copy workqueue */ > #define NFSD4_CLIENT_CB_FLAG_MASK (1 << NFSD4_CLIENT_CB_UPDATE | \ > 1 << NFSD4_CLIENT_CB_KILL) > unsigned long cl_flags; > @@ -644,6 +645,7 @@ extern void nfsd4_init_cb(struct nfsd4_callback *cb, struct nfs4_client *clp, > extern int nfsd4_create_callback_queue(void); > extern void nfsd4_destroy_callback_queue(void); > extern void nfsd4_shutdown_callback(struct nfs4_client *); > +extern void nfsd4_shutdown_copy(struct nfs4_client *clp); > extern void nfsd4_prepare_cb_recall(struct nfs4_delegation *dp); > extern struct nfs4_client_reclaim *nfs4_client_to_reclaim(const char *name, > struct nfsd_net *nn); > -- > 1.8.3.1 >