Return-Path: linux-nfs-owner@vger.kernel.org Received: from fieldses.org ([174.143.236.118]:43397 "EHLO fieldses.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752382AbaEVTch (ORCPT ); Thu, 22 May 2014 15:32:37 -0400 From: "J. Bruce Fields" To: linux-nfs@vger.kernel.org Cc: Christoph Hellwig , "J. Bruce Fields" Subject: [PATCH 22/52] nfsd4: reserve space before inlining 0-copy pages Date: Thu, 22 May 2014 15:31:57 -0400 Message-Id: <1400787148-25941-23-git-send-email-bfields@redhat.com> In-Reply-To: <1400787148-25941-1-git-send-email-bfields@redhat.com> References: <1400787148-25941-1-git-send-email-bfields@redhat.com> Sender: linux-nfs-owner@vger.kernel.org List-ID: From: "J. Bruce Fields" Once we've included page-cache pages in the encoding it's difficult to remove them and restart encoding. (xdr_truncate_encode doesn't handle that case.) So, make sure we'll have adequate space to finish the operation first. For now COMPOUND_SLACK_SPACE checks should prevent this case happening, but we want to remove those checks. Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4xdr.c | 11 +++++++---- 1 file changed, 7 insertions(+), 4 deletions(-) diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c index 9c532eb..9fb1b1a 100644 --- a/fs/nfsd/nfs4xdr.c +++ b/fs/nfsd/nfs4xdr.c @@ -3071,6 +3071,10 @@ nfsd4_encode_read(struct nfsd4_compoundres *resp, __be32 nfserr, if (!p) return nfserr_resource; + /* Make sure there will be room for padding if needed: */ + if (xdr->end - xdr->p < 1) + return nfserr_resource; + maxcount = svc_max_payload(resp->rqstp); if (maxcount > read->rd_length) maxcount = read->rd_length; @@ -3122,8 +3126,6 @@ nfsd4_encode_read(struct nfsd4_compoundres *resp, __be32 nfserr, resp->xdr.buf->tail[0].iov_len = 0; if (maxcount&3) { p = xdr_reserve_space(xdr, 4); - if (!p) - return nfserr_resource; WRITE32(0); resp->xdr.buf->tail[0].iov_base += maxcount&3; resp->xdr.buf->tail[0].iov_len = 4 - (maxcount&3); @@ -3156,6 +3158,9 @@ nfsd4_encode_readlink(struct nfsd4_compoundres *resp, __be32 nfserr, struct nfsd if (!p) return nfserr_resource; + if (xdr->end - xdr->p < 1) + return nfserr_resource; + /* * XXX: By default, the ->readlink() VFS op will truncate symlinks * if they would overflow the buffer. Is this kosher in NFSv4? If @@ -3182,8 +3187,6 @@ nfsd4_encode_readlink(struct nfsd4_compoundres *resp, __be32 nfserr, struct nfsd resp->xdr.buf->tail[0].iov_len = 0; if (maxcount&3) { p = xdr_reserve_space(xdr, 4); - if (!p) - return nfserr_resource; WRITE32(0); resp->xdr.buf->tail[0].iov_base += maxcount&3; resp->xdr.buf->tail[0].iov_len = 4 - (maxcount&3); -- 1.9.0