From owner-svn-src-all@FreeBSD.ORG Fri Sep 7 20:57:37 2012 Return-Path: Delivered-To: svn-src-all@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [69.147.83.52]) by hub.freebsd.org (Postfix) with ESMTP id 64F70106566B; Fri, 7 Sep 2012 20:57:37 +0000 (UTC) (envelope-from rmacklem@FreeBSD.org) Received: from svn.freebsd.org (svn.freebsd.org [IPv6:2001:4f8:fff6::2c]) by mx1.freebsd.org (Postfix) with ESMTP id 35D6C8FC12; Fri, 7 Sep 2012 20:57:37 +0000 (UTC) Received: from svn.freebsd.org (localhost [127.0.0.1]) by svn.freebsd.org (8.14.4/8.14.4) with ESMTP id q87Kvbmc061972; Fri, 7 Sep 2012 20:57:37 GMT (envelope-from rmacklem@svn.freebsd.org) Received: (from rmacklem@localhost) by svn.freebsd.org (8.14.4/8.14.4/Submit) id q87Kvahi061970; Fri, 7 Sep 2012 20:57:36 GMT (envelope-from rmacklem@svn.freebsd.org) Message-Id: <201209072057.q87Kvahi061970@svn.freebsd.org> From: Rick Macklem Date: Fri, 7 Sep 2012 20:57:36 +0000 (UTC) To: src-committers@freebsd.org, svn-src-all@freebsd.org, svn-src-stable@freebsd.org, svn-src-stable-8@freebsd.org X-SVN-Group: stable-8 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Cc: Subject: svn commit: r240213 - stable/8/sys/fs/nfsclient X-BeenThere: svn-src-all@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: "SVN commit messages for the entire src tree \(except for " user" and " projects" \)" List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Fri, 07 Sep 2012 20:57:37 -0000 Author: rmacklem Date: Fri Sep 7 20:57:36 2012 New Revision: 240213 URL: http://svn.freebsd.org/changeset/base/240213 Log: MFC: r237244 Fix the NFSv4 client for the case where mmap'd files are written, but not msync'd by a process. A VOP_PUTPAGES() called when VOP_RECLAIM() happens will usually fail, since the NFSv4 Open has already been closed by VOP_INACTIVE(). Add a vm_object_page_clean() call to the NFSv4 client's VOP_INACTIVE(), so that the write happens before the NFSv4 Open is closed. kib@ suggested using vgone() instead and I will explore this, but this patch fixes things in the meantime. For some reason, the VOP_PUTPAGES() is still attempted in VOP_RECLAIM(), but having this fail doesn't cause any problems except a "stateid0 in write" being logged. Modified: stable/8/sys/fs/nfsclient/nfs_clnode.c Directory Properties: stable/8/sys/ (props changed) stable/8/sys/amd64/include/xen/ (props changed) stable/8/sys/boot/ (props changed) stable/8/sys/cddl/ (props changed) stable/8/sys/cddl/contrib/opensolaris/ (props changed) stable/8/sys/compat/ (props changed) stable/8/sys/contrib/dev/acpica/ (props changed) stable/8/sys/contrib/pf/ (props changed) stable/8/sys/dev/ (props changed) stable/8/sys/dev/e1000/ (props changed) stable/8/sys/dev/sound/ (props changed) stable/8/sys/dev/sound/pci/ (props changed) stable/8/sys/dev/virtio/ (props changed) stable/8/sys/kern/ (props changed) stable/8/sys/sys/ (props changed) stable/8/sys/vm/ (props changed) Modified: stable/8/sys/fs/nfsclient/nfs_clnode.c ============================================================================== --- stable/8/sys/fs/nfsclient/nfs_clnode.c Fri Sep 7 20:57:36 2012 (r240212) +++ stable/8/sys/fs/nfsclient/nfs_clnode.c Fri Sep 7 20:57:36 2012 (r240213) @@ -207,18 +207,28 @@ ncl_inactive(struct vop_inactive_args *a struct nfsnode *np; struct sillyrename *sp; struct vnode *vp = ap->a_vp; + boolean_t retv; np = VTONFS(vp); if (NFS_ISV4(vp) && vp->v_type == VREG) { /* * Since mmap()'d files do I/O after VOP_CLOSE(), the NFSv4 - * Close operations are delayed until now. Any dirty buffers - * must be flushed before the close, so that the stateid is - * available for the writes. + * Close operations are delayed until now. Any dirty + * buffers/pages must be flushed before the close, so that the + * stateid is available for the writes. */ - (void) ncl_flush(vp, MNT_WAIT, NULL, ap->a_td, 1, 0); - (void) nfsrpc_close(vp, 1, ap->a_td); + if (vp->v_object != NULL) { + VM_OBJECT_LOCK(vp->v_object); + retv = vm_object_page_clean(vp->v_object, 0, 0, + OBJPC_SYNC); + VM_OBJECT_UNLOCK(vp->v_object); + } else + retv = TRUE; + if (retv == TRUE) { + (void)ncl_flush(vp, MNT_WAIT, NULL, ap->a_td, 1, 0); + (void)nfsrpc_close(vp, 1, ap->a_td); + } } mtx_lock(&np->n_mtx);