Return-Path: linux-nfs-owner@vger.kernel.org Received: from mx11.netapp.com ([216.240.18.76]:43280 "EHLO mx11.netapp.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753758Ab3KYS2S convert rfc822-to-8bit (ORCPT ); Mon, 25 Nov 2013 13:28:18 -0500 From: "Myklebust, Trond" To: "Adamson, Andy" CC: Linux NFS Mailing List Subject: Re: [PATCH 1/1] NFSv4.1 fix a kswap nfs4_state_manger race Date: Mon, 25 Nov 2013 18:28:16 +0000 Message-ID: References: <1385402270-14284-1-git-send-email-andros@netapp.com> <1385402270-14284-2-git-send-email-andros@netapp.com> <5B8C7A9D-CD9E-487A-AC62-B1292649835D@netapp.com> <7E331FE8-EF05-434D-8434-0C35C3EF2F8B@netapp.com> In-Reply-To: <7E331FE8-EF05-434D-8434-0C35C3EF2F8B@netapp.com> Content-Type: text/plain; charset="Windows-1252" MIME-Version: 1.0 Sender: linux-nfs-owner@vger.kernel.org List-ID: On Nov 25, 2013, at 13:17, Adamson, Andy wrote: > > On Nov 25, 2013, at 1:13 PM, "Myklebust, Trond" > wrote: > >> >> On Nov 25, 2013, at 12:57, wrote: >> >>> From: Andy Adamson >>> >>> The state manager is recovering expired state and recovery OPENs are being >>> processed. If kswapd is pruning inodes at the same time, a deadlock can occur >>> when kswapd calls evict_inode on an NFSv4.1 inode with a layout, and the >>> resultant layoutreturn gets an error that the state mangager is to handle, >>> causing the layoutreturn to wait on the (NFS client) cl_rpcwaitq. >>> >>> At the same time an open is waiting for the inode deletion to complete in >>> __wait_on_freeing_inode. >>> >>> If the open is either the open called by the state manager, or an open from >>> the same open owner that is holding the NFSv4.0 sequence id which causes the >>> OPEN from the state manager to wait for the sequence id on the Seqid_waitqueue, >>> then the state is deadlocked with kswapd. >>> >>> Do not handle LAYOUTRETURN errors when called from nfs4_evict_inode. >> >> Why are we waiting for recovery in LAYOUTRETURN at all? Layouts are automatically lost when the server reboots or when the lease is otherwise lost. >> >> IOW: Is there any reason why we need to special-case nfs4_evict_inode? Shouldn?t we just bail out on error in _all_ cases? > > Yeah, I was thinking about this as well - perhaps recovering from session-level errors or grace/delay errors would be useful for the block client. NFS4ERR_DELAY, probably, yes. NFS4ERR_GRACE, no? That?s a reboot situation As for session level errors, I?d say that complicates things too much, since several of those can basically end up masking a NFS4ERR_STALE_CLIENTID error. Either way, all the layout types (including blocks) should be able to continue on even if we miss a layout return or two. The server has to be coded to expect a forgetful client. -- Trond Myklebust Linux NFS client maintainer NetApp Trond.Myklebust@netapp.com www.netapp.com