From owner-svn-src-stable@FreeBSD.ORG Sat Jun 15 01:35:52 2013 Return-Path: Delivered-To: svn-src-stable@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [8.8.178.115]) by hub.freebsd.org (Postfix) with ESMTP id D382FE6C; Sat, 15 Jun 2013 01:35:52 +0000 (UTC) (envelope-from rmacklem@FreeBSD.org) Received: from svn.freebsd.org (svn.freebsd.org [IPv6:2001:1900:2254:2068::e6a:0]) by mx1.freebsd.org (Postfix) with ESMTP id C42521142; Sat, 15 Jun 2013 01:35:52 +0000 (UTC) Received: from svn.freebsd.org ([127.0.1.70]) by svn.freebsd.org (8.14.7/8.14.7) with ESMTP id r5F1Zqa9060661; Sat, 15 Jun 2013 01:35:52 GMT (envelope-from rmacklem@svn.freebsd.org) Received: (from rmacklem@localhost) by svn.freebsd.org (8.14.7/8.14.5/Submit) id r5F1Zqpw060659; Sat, 15 Jun 2013 01:35:52 GMT (envelope-from rmacklem@svn.freebsd.org) Message-Id: <201306150135.r5F1Zqpw060659@svn.freebsd.org> From: Rick Macklem Date: Sat, 15 Jun 2013 01:35:52 +0000 (UTC) To: src-committers@freebsd.org, svn-src-all@freebsd.org, svn-src-stable@freebsd.org, svn-src-stable-8@freebsd.org Subject: svn commit: r251768 - in stable/8/sys: fs/nfsclient nfsclient X-SVN-Group: stable-8 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit X-BeenThere: svn-src-stable@freebsd.org X-Mailman-Version: 2.1.14 Precedence: list List-Id: SVN commit messages for all the -stable branches of the src tree List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sat, 15 Jun 2013 01:35:52 -0000 Author: rmacklem Date: Sat Jun 15 01:35:52 2013 New Revision: 251768 URL: http://svnweb.freebsd.org/changeset/base/251768 Log: MFC: r249623 Both NFS clients can deadlock when using the "rdirplus" mount option. This can occur when an nfsiod thread that already holds a buffer lock attempts to acquire a vnode lock on an entry in the directory (a LOR) when another thread holding the vnode lock is waiting on an nfsiod thread. This patch avoids the deadlock by disabling readahead for this case, so the nfsiod threads never do readdirplus. Since readaheads for directories need the directory offset cookie from the previous read, they cannot normally happen in parallel. As such, testing by jhb@ and myself didn't find any performance degredation when this patch is applied. If there is a case where this results in a significant performance degradation, mounting without the "rdirplus" option can be done to re-enable readahead for directories. Modified: stable/8/sys/fs/nfsclient/nfs_clbio.c stable/8/sys/nfsclient/nfs_bio.c Directory Properties: stable/8/sys/ (props changed) stable/8/sys/fs/ (props changed) stable/8/sys/nfsclient/ (props changed) Modified: stable/8/sys/fs/nfsclient/nfs_clbio.c ============================================================================== --- stable/8/sys/fs/nfsclient/nfs_clbio.c Fri Jun 14 23:43:44 2013 (r251767) +++ stable/8/sys/fs/nfsclient/nfs_clbio.c Sat Jun 15 01:35:52 2013 (r251768) @@ -1368,10 +1368,18 @@ ncl_asyncio(struct nfsmount *nmp, struct * Commits are usually short and sweet so lets save some cpu and * leave the async daemons for more important rpc's (such as reads * and writes). + * + * Readdirplus RPCs do vget()s to acquire the vnodes for entries + * in the directory in order to update attributes. This can deadlock + * with another thread that is waiting for async I/O to be done by + * an nfsiod thread while holding a lock on one of these vnodes. + * To avoid this deadlock, don't allow the async nfsiod threads to + * perform Readdirplus RPCs. */ mtx_lock(&ncl_iod_mutex); - if (bp->b_iocmd == BIO_WRITE && (bp->b_flags & B_NEEDCOMMIT) && - (nmp->nm_bufqiods > ncl_numasync / 2)) { + if ((bp->b_iocmd == BIO_WRITE && (bp->b_flags & B_NEEDCOMMIT) && + (nmp->nm_bufqiods > ncl_numasync / 2)) || + (bp->b_vp->v_type == VDIR && (nmp->nm_flag & NFSMNT_RDIRPLUS))) { mtx_unlock(&ncl_iod_mutex); return(EIO); } Modified: stable/8/sys/nfsclient/nfs_bio.c ============================================================================== --- stable/8/sys/nfsclient/nfs_bio.c Fri Jun 14 23:43:44 2013 (r251767) +++ stable/8/sys/nfsclient/nfs_bio.c Sat Jun 15 01:35:52 2013 (r251768) @@ -1360,10 +1360,18 @@ nfs_asyncio(struct nfsmount *nmp, struct * Commits are usually short and sweet so lets save some cpu and * leave the async daemons for more important rpc's (such as reads * and writes). + * + * Readdirplus RPCs do vget()s to acquire the vnodes for entries + * in the directory in order to update attributes. This can deadlock + * with another thread that is waiting for async I/O to be done by + * an nfsiod thread while holding a lock on one of these vnodes. + * To avoid this deadlock, don't allow the async nfsiod threads to + * perform Readdirplus RPCs. */ mtx_lock(&nfs_iod_mtx); - if (bp->b_iocmd == BIO_WRITE && (bp->b_flags & B_NEEDCOMMIT) && - (nmp->nm_bufqiods > nfs_numasync / 2)) { + if ((bp->b_iocmd == BIO_WRITE && (bp->b_flags & B_NEEDCOMMIT) && + (nmp->nm_bufqiods > nfs_numasync / 2)) || + (bp->b_vp->v_type == VDIR && (nmp->nm_flag & NFSMNT_RDIRPLUS))) { mtx_unlock(&nfs_iod_mtx); return(EIO); }