Return-Path: Received: from mail-io0-f181.google.com ([209.85.223.181]:34900 "EHLO mail-io0-f181.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751878AbbHSSlN (ORCPT ); Wed, 19 Aug 2015 14:41:13 -0400 Received: by iodt126 with SMTP id t126so19665664iod.2 for ; Wed, 19 Aug 2015 11:41:13 -0700 (PDT) MIME-Version: 1.0 In-Reply-To: References: <1437677341-2943-1-git-send-email-trond.myklebust@primarydata.com> From: Peng Tao Date: Wed, 19 Aug 2015 11:40:53 -0700 Message-ID: Subject: Re: [PATCH 1/5] NFSv4/pnfs: Ensure we don't miss a file extension To: Trond Myklebust Cc: Linux NFS Mailing List Content-Type: text/plain; charset=UTF-8 Sender: linux-nfs-owner@vger.kernel.org List-ID: On Wed, Aug 19, 2015 at 10:50 AM, Peng Tao wrote: > Hi Trond, > > Can we please add following patch to stable as well? It not only fixes > pnfs write, but also fixes in band writes when delegations are > enabled. Without it, I can always get fstest read_write/mctime.t > failure with in band IO. ah, nvm, speaking too soon. I can still reproduce fstest read_write/mctime.t failure on delegation+inband IO with your testing branch... Looking into it... sorry for the noise. > > Thanks, > Tao > > On Thu, Jul 23, 2015 at 11:48 AM, Trond Myklebust > wrote: >> pNFS writes don't return attributes, however that doesn't mean that we >> should ignore the fact that they may be extending the file. This patch >> ensures that if a write is seen to extend the file, then we always set >> an attribute barrier, and update the cached file size. >> >> Signed-off-by: Trond Myklebust >> --- >> fs/nfs/write.c | 15 +++++++++------ >> 1 file changed, 9 insertions(+), 6 deletions(-) >> >> diff --git a/fs/nfs/write.c b/fs/nfs/write.c >> index 359e9ad596c9..0e6a2b8786b4 100644 >> --- a/fs/nfs/write.c >> +++ b/fs/nfs/write.c >> @@ -1378,24 +1378,27 @@ static void nfs_writeback_check_extend(struct nfs_pgio_header *hdr, >> { >> struct nfs_pgio_args *argp = &hdr->args; >> struct nfs_pgio_res *resp = &hdr->res; >> + u64 size = argp->offset + resp->count; >> >> if (!(fattr->valid & NFS_ATTR_FATTR_SIZE)) >> + fattr->size = size; >> + if (nfs_size_to_loff_t(fattr->size) < i_size_read(hdr->inode)) { >> + fattr->valid &= ~NFS_ATTR_FATTR_SIZE; >> return; >> - if (argp->offset + resp->count != fattr->size) >> - return; >> - if (nfs_size_to_loff_t(fattr->size) < i_size_read(hdr->inode)) >> + } >> + if (size != fattr->size) >> return; >> /* Set attribute barrier */ >> nfs_fattr_set_barrier(fattr); >> + /* ...and update size */ >> + fattr->valid |= NFS_ATTR_FATTR_SIZE; >> } >> >> void nfs_writeback_update_inode(struct nfs_pgio_header *hdr) >> { >> - struct nfs_fattr *fattr = hdr->res.fattr; >> + struct nfs_fattr *fattr = &hdr->fattr; >> struct inode *inode = hdr->inode; >> >> - if (fattr == NULL) >> - return; >> spin_lock(&inode->i_lock); >> nfs_writeback_check_extend(hdr, fattr); >> nfs_post_op_update_inode_force_wcc_locked(inode, fattr); >> -- >> 2.4.3 >>