Return-Path: linux-nfs-owner@vger.kernel.org Received: from mail-qg0-f65.google.com ([209.85.192.65]:49572 "EHLO mail-qg0-f65.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752832AbaG2Qlq (ORCPT ); Tue, 29 Jul 2014 12:41:46 -0400 Received: by mail-qg0-f65.google.com with SMTP id z60so3515507qgd.4 for ; Tue, 29 Jul 2014 09:41:46 -0700 (PDT) From: Jeff Layton Date: Tue, 29 Jul 2014 12:41:43 -0400 To: "J. Bruce Fields" Cc: linux-nfs@vger.kernel.org, hch@infradead.org Subject: Re: [PATCH v6 2/9] nfsd: fully unhash delegations when revoking them Message-ID: <20140729124143.7e3bae67@tlielax.poochiereds.net> In-Reply-To: <20140729145331.GB19210@fieldses.org> References: <1406288067-20663-1-git-send-email-jlayton@primarydata.com> <1406288067-20663-3-git-send-email-jlayton@primarydata.com> <20140729145331.GB19210@fieldses.org> MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Sender: linux-nfs-owner@vger.kernel.org List-ID: On Tue, 29 Jul 2014 10:53:31 -0400 "J. Bruce Fields" wrote: > On Fri, Jul 25, 2014 at 07:34:20AM -0400, Jeff Layton wrote: > > @@ -698,11 +699,10 @@ static void revoke_delegation(struct nfs4_delegation *dp) > > struct nfs4_client *clp = dp->dl_stid.sc_client; > > > > if (clp->cl_minorversion == 0) > > - destroy_delegation(dp); > > + destroy_revoked_delegation(dp); > > I don't understand this one--in the NFSv4.0 case, when is the delegation > unhashed? > > (And, apologies if that's a repeat question. I remember wondering about > this before but can't find mail in the archives....) > > --b. > It's unhashed by the laundromat (or the fault injection code). The deltas that change that are later in this patch, but it doesn't have much context so it's a little tricky to spot. Here's the relevant code chunk from nfs4_laundromat: ---------------[snip]----------------- spin_lock(&state_lock); list_for_each_safe(pos, next, &nn->del_recall_lru) { dp = list_entry (pos, struct nfs4_delegation, dl_recall_lru); if (net_generic(dp->dl_stid.sc_client->net, nfsd_net_id) != nn) continue; if (time_after((unsigned long)dp->dl_time, (unsigned long)cutoff)) { t = dp->dl_time - cutoff; new_timeo = min(new_timeo, t); break; } unhash_delegation_locked(dp); list_add(&dp->dl_recall_lru, &reaplist); } spin_unlock(&state_lock); list_for_each_safe(pos, next, &reaplist) { dp = list_entry (pos, struct nfs4_delegation, dl_recall_lru); revoke_delegation(dp); } ---------------[snip]----------------- ...so the laundromat walks the nn->del_recall_lru list, unhashes any delegation that has expired and gathers them up onto the list. Then, it calls revoke_delegation on each outside of the lock. As a side note, that net_generic comparison there looks bogus to me. How could you ever end up with an entry from a different namespace on this list? I left it in for now since it didn't seem relevant (or terribly expensive), but I suspect that it can be removed (or turned into a WARN or something). > > else { > > - unhash_delegation(dp); > > dp->dl_stid.sc_type = NFS4_REVOKED_DELEG_STID; > > - list_add(&dp->dl_recall_lru, &clp->cl_revoked); > > + list_move(&dp->dl_recall_lru, &clp->cl_revoked); > > } > > } > > > > @@ -1458,15 +1458,14 @@ destroy_client(struct nfs4_client *clp) > > spin_lock(&state_lock); > > while (!list_empty(&clp->cl_delegations)) { > > dp = list_entry(clp->cl_delegations.next, struct > > nfs4_delegation, dl_perclnt); > > - list_del_init(&dp->dl_perclnt); > > - /* Ensure that deleg break won't try to requeue it > > */ > > - ++dp->dl_time; > > - list_move(&dp->dl_recall_lru, &reaplist); > > + unhash_delegation_locked(dp); > > + list_add(&dp->dl_recall_lru, &reaplist); > > } > > spin_unlock(&state_lock); > > while (!list_empty(&reaplist)) { > > dp = list_entry(reaplist.next, struct > > nfs4_delegation, dl_recall_lru); > > - destroy_delegation(dp); > > + list_del_init(&dp->dl_recall_lru); > > + nfs4_put_delegation(dp); > > } > > list_splice_init(&clp->cl_revoked, &reaplist); > > while (!list_empty(&reaplist)) { > > @@ -3662,7 +3661,7 @@ nfs4_open_delegation(struct net *net, struct > > svc_fh *fh, open->op_delegate_type = NFS4_OPEN_DELEGATE_READ; > > return; > > out_free: > > - destroy_delegation(dp); > > + nfs4_put_delegation(dp); > > out_no_deleg: > > open->op_delegate_type = NFS4_OPEN_DELEGATE_NONE; > > if (open->op_claim_type == NFS4_OPEN_CLAIM_PREVIOUS && > > @@ -3900,7 +3899,8 @@ nfs4_laundromat(struct nfsd_net *nn) > > new_timeo = min(new_timeo, t); > > break; > > } > > - list_move(&dp->dl_recall_lru, &reaplist); > > + unhash_delegation_locked(dp); > > + list_add(&dp->dl_recall_lru, &reaplist); > > } > > spin_unlock(&state_lock); > > list_for_each_safe(pos, next, &reaplist) { > > @@ -5382,12 +5382,8 @@ static u64 nfsd_find_all_delegations(struct > > nfs4_client *clp, u64 max, if (dp->dl_time != 0) > > continue; > > > > - /* > > - * Increment dl_time to ensure that > > delegation breaks > > - * don't monkey with it now that we are. > > - */ > > - ++dp->dl_time; > > - list_move(&dp->dl_recall_lru, victims); > > + unhash_delegation_locked(dp); > > + list_add(&dp->dl_recall_lru, victims); > > } > > if (++count == max) > > break; > > @@ -5642,12 +5638,14 @@ nfs4_state_shutdown_net(struct net *net) > > spin_lock(&state_lock); > > list_for_each_safe(pos, next, &nn->del_recall_lru) { > > dp = list_entry (pos, struct nfs4_delegation, > > dl_recall_lru); > > - list_move(&dp->dl_recall_lru, &reaplist); > > + unhash_delegation_locked(dp); > > + list_add(&dp->dl_recall_lru, &reaplist); > > } > > spin_unlock(&state_lock); > > list_for_each_safe(pos, next, &reaplist) { > > dp = list_entry (pos, struct nfs4_delegation, > > dl_recall_lru); > > - destroy_delegation(dp); > > + list_del_init(&dp->dl_recall_lru); > > + nfs4_put_delegation(dp); > > } > > > > nfsd4_client_tracking_exit(net); > > -- > > 1.9.3 > > -- Jeff Layton