Return-Path: Received: from userp1040.oracle.com ([156.151.31.81]:20223 "EHLO userp1040.oracle.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752451AbbLMToY convert rfc822-to-8bit (ORCPT ); Sun, 13 Dec 2015 14:44:24 -0500 Content-Type: text/plain; charset=us-ascii Mime-Version: 1.0 (Mac OS X Mail 8.2 \(2104\)) Subject: Re: [PATCH v3 1/6] svcrdma: Do not send XDR roundup bytes for a write chunk From: Chuck Lever In-Reply-To: <566CE288.6000808@talpey.com> Date: Sun, 13 Dec 2015 14:44:21 -0500 Cc: linux-rdma@vger.kernel.org, Linux NFS Mailing List Message-Id: References: <20151207203851.12988.97804.stgit@klimt.1015granger.net> <20151207204231.12988.59287.stgit@klimt.1015granger.net> <566CE288.6000808@talpey.com> To: Tom Talpey Sender: linux-nfs-owner@vger.kernel.org List-ID: Hi Tom- > On Dec 12, 2015, at 10:14 PM, Tom Talpey wrote: > > Two small comments. > > On 12/7/2015 12:42 PM, Chuck Lever wrote: >> Minor optimization: when dealing with write chunk XDR roundup, do >> not post a Write WR for the zero bytes in the pad. Simply update >> the write segment in the RPC-over-RDMA header to reflect the extra >> pad bytes. >> >> The Reply chunk is also a write chunk, but the server does not use >> send_write_chunks() to send the Reply chunk. That's OK in this case: >> the server Upper Layer typically marshals the Reply chunk contents >> in a single contiguous buffer, without a separate tail for the XDR >> pad. >> >> The comments and the variable naming refer to "chunks" but what is >> really meant is "segments." The existing code sends only one >> xdr_write_chunk per RPC reply. >> >> The fix assumes this as well. When the XDR pad in the first write >> chunk is reached, the assumption is the Write list is complete and >> send_write_chunks() returns. >> >> That will remain a valid assumption until the server Upper Layer can >> support multiple bulk payload results per RPC. >> >> Signed-off-by: Chuck Lever >> --- >> net/sunrpc/xprtrdma/svc_rdma_sendto.c | 7 +++++++ >> 1 file changed, 7 insertions(+) >> >> diff --git a/net/sunrpc/xprtrdma/svc_rdma_sendto.c b/net/sunrpc/xprtrdma/svc_rdma_sendto.c >> index 969a1ab..bad5eaa 100644 >> --- a/net/sunrpc/xprtrdma/svc_rdma_sendto.c >> +++ b/net/sunrpc/xprtrdma/svc_rdma_sendto.c >> @@ -342,6 +342,13 @@ static int send_write_chunks(struct svcxprt_rdma *xprt, >> arg_ch->rs_handle, >> arg_ch->rs_offset, >> write_len); >> + >> + /* Do not send XDR pad bytes */ > > It might be clearer to say "marshal" instead of "send". Marshaling each segment happens unconditionally in the svc_rdma_xdr_encode_array_chunk() call just before this comment. I really do mean "Do not send" here: the patch is intended to squelch the RDMA Write of the XDR pad for this chunk. Perhaps "Do not write" would be more precise, but Bruce has already committed this patch, IIRC. >> + if (chunk_no && write_len < 4) { > > Why is it necessary to check for chunk_no == 0? It is not > possible for leading data to ever be padding, nor is a leading > data element ever less than 4 bytes long. Right? I'm checking for chunk_no != 0, for exactly the reasons you mentioned. > Tom. > >> + chunk_no++; >> + break; >> + } >> + >> chunk_off = 0; >> while (write_len) { >> ret = send_write(xprt, rqstp, -- Chuck Lever