Return-Path: Received: from userp1040.oracle.com ([156.151.31.81]:22071 "EHLO userp1040.oracle.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750919AbbGMQk2 convert rfc822-to-8bit (ORCPT ); Mon, 13 Jul 2015 12:40:28 -0400 Content-Type: text/plain; charset=windows-1252 Mime-Version: 1.0 (Mac OS X Mail 7.3 \(1878.6\)) Subject: Re: [PATCH v1 5/5] svcrdma: Boost NFS READ/WRITE payload size maximum From: Chuck Lever In-Reply-To: <20150710160506.GE31121@fieldses.org> Date: Mon, 13 Jul 2015 12:40:25 -0400 Cc: Linux NFS Mailing List , linux-rdma@vger.kernel.org Message-Id: References: <20150709204230.8481.45457.stgit@klimt.1015granger.net> <20150709204546.8481.65857.stgit@klimt.1015granger.net> <20150710141814.GA31121@fieldses.org> <11D08769-F6EB-4696-8910-8B207B20CC10@oracle.com> <20150710155424.GD31121@fieldses.org> <24367FE9-7C2B-4B2D-A48A-90329D92296E@oracle.com> <20150710160506.GE31121@fieldses.org> To: "J. Bruce Fields" Sender: linux-nfs-owner@vger.kernel.org List-ID: On Jul 10, 2015, at 12:05 PM, J. Bruce Fields wrote: > On Fri, Jul 10, 2015 at 11:59:20AM -0400, Chuck Lever wrote: >> >> On Jul 10, 2015, at 11:54 AM, J. Bruce Fields wrote: >> >>> On Fri, Jul 10, 2015 at 10:59:48AM -0400, Chuck Lever wrote: >>>> >>>> On Jul 10, 2015, at 10:18 AM, bfields@fieldses.org wrote: >>>> >>>>> On Thu, Jul 09, 2015 at 04:45:47PM -0400, Chuck Lever wrote: >>>>>> Increased to 1 megabyte. >>>>> >>>>> Why not more or less? >>>>> >>>>> Why do we even have this constant, why shouldn't we just use >>>>> RPCSVC_MAXPAYLOAD? >>>> >>>> The payload size maximum for RDMA is based on RPCRDMA_MAX_SVC_SEGS. >>>> We could reverse the relationship and make RPCRDMA_MAX_SVC_SEGS >>>> based on RPCSVC_MAXPAYLOAD divided by the platform?s page size. >>> >>> But there'd be no reason to do that, because we're not using >>> RPCRDMA_MAX_SVC_SEGS anywhere. Should we be? >> >> Let me try using RPCSVC_MAXPAYLOAD. That is based on RPCSVC_MAXPAGES, >> which is actually used in svc_rdma_*.c > > OK, thanks, that sounds like that would make more sense. The change to use RPCSVC_MAXPAYLOAD is clean enough, but it may re-expose a PPC64-x86_64 interop bug. We need to dig up some testing resources to see if that bug is still a problem. Stay tuned. Since you?ve taken the other patches in this series, I?ll hold off on reposting a v2 until I have this straightened out. > --b. > >> >> >>> --b. >>> >>>> >>>> >>>>> --b. >>>>> >>>>>> >>>>>> Signed-off-by: Chuck Lever >>>>>> --- >>>>>> >>>>>> include/linux/sunrpc/svc_rdma.h | 2 +- >>>>>> 1 files changed, 1 insertions(+), 1 deletions(-) >>>>>> >>>>>> diff --git a/include/linux/sunrpc/svc_rdma.h b/include/linux/sunrpc/svc_rdma.h >>>>>> index 13af61b..1bca6dd 100644 >>>>>> --- a/include/linux/sunrpc/svc_rdma.h >>>>>> +++ b/include/linux/sunrpc/svc_rdma.h >>>>>> @@ -172,7 +172,7 @@ struct svcxprt_rdma { >>>>>> #define RDMAXPRT_SQ_PENDING 2 >>>>>> #define RDMAXPRT_CONN_PENDING 3 >>>>>> >>>>>> -#define RPCRDMA_MAX_SVC_SEGS (64) /* server max scatter/gather */ >>>>>> +#define RPCRDMA_MAX_SVC_SEGS (256) /* server max scatter/gather */ >>>>>> #if RPCSVC_MAXPAYLOAD < (RPCRDMA_MAX_SVC_SEGS << PAGE_SHIFT) >>>>>> #define RPCRDMA_MAXPAYLOAD RPCSVC_MAXPAYLOAD >>>>>> #else >>>>>> >>>>>> -- >>>>>> To unsubscribe from this list: send the line "unsubscribe linux-nfs" in >>>>>> the body of a message to majordomo@vger.kernel.org >>>>>> More majordomo info at http://vger.kernel.org/majordomo-info.html >>>> >>>> -- >>>> Chuck Lever >>>> >>>> >> >> -- >> Chuck Lever >> >> > -- > To unsubscribe from this list: send the line "unsubscribe linux-nfs" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html -- Chuck Lever