Return-Path: Received: from mail-la0-f47.google.com ([209.85.215.47]:33094 "EHLO mail-la0-f47.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752501AbbEHPxs (ORCPT ); Fri, 8 May 2015 11:53:48 -0400 Received: by layy10 with SMTP id y10so55668347lay.0 for ; Fri, 08 May 2015 08:53:46 -0700 (PDT) MIME-Version: 1.0 In-Reply-To: <554B45E7.4040902@dev.mellanox.co.il> References: <20150504174626.3483.97639.stgit@manet.1015granger.net> <20150504175856.3483.57373.stgit@manet.1015granger.net> <554B45E7.4040902@dev.mellanox.co.il> Date: Fri, 8 May 2015 21:23:46 +0530 Message-ID: Subject: Re: [PATCH v1 14/14] xprtrmda: Reduce per-transport MR allocation From: Devesh Sharma To: Sagi Grimberg Cc: Chuck Lever , linux-rdma@vger.kernel.org, Linux NFS Mailing List Content-Type: text/plain; charset=UTF-8 Sender: linux-nfs-owner@vger.kernel.org List-ID: Reviewed-by: Devesh Sharma On Thu, May 7, 2015 at 4:30 PM, Sagi Grimberg wrote: > On 5/4/2015 8:58 PM, Chuck Lever wrote: >> >> Reduce resource consumption per-transport to make way for increasing >> the credit limit and maximum r/wsize. Pre-allocate fewer MRs. >> >> Signed-off-by: Chuck Lever >> --- >> net/sunrpc/xprtrdma/fmr_ops.c | 6 ++++-- >> net/sunrpc/xprtrdma/frwr_ops.c | 6 ++++-- >> 2 files changed, 8 insertions(+), 4 deletions(-) >> >> diff --git a/net/sunrpc/xprtrdma/fmr_ops.c b/net/sunrpc/xprtrdma/fmr_ops.c >> index 4a53ad5..f1e8daf 100644 >> --- a/net/sunrpc/xprtrdma/fmr_ops.c >> +++ b/net/sunrpc/xprtrdma/fmr_ops.c >> @@ -69,8 +69,10 @@ fmr_op_init(struct rpcrdma_xprt *r_xprt) >> INIT_LIST_HEAD(&buf->rb_mws); >> INIT_LIST_HEAD(&buf->rb_all); >> >> - i = (buf->rb_max_requests + 1) * RPCRDMA_MAX_SEGS; >> - dprintk("RPC: %s: initializing %d FMRs\n", __func__, i); >> + i = max_t(int, RPCRDMA_MAX_DATA_SEGS / RPCRDMA_MAX_FMR_SGES, 1); >> + i += 2; /* head + tail */ >> + i *= buf->rb_max_requests; /* one set for each RPC slot */ >> + dprintk("RPC: %s: initalizing %d FMRs\n", __func__, i); >> >> rc = -ENOMEM; >> while (i--) { >> diff --git a/net/sunrpc/xprtrdma/frwr_ops.c >> b/net/sunrpc/xprtrdma/frwr_ops.c >> index edc10ba..fc2d0c6 100644 >> --- a/net/sunrpc/xprtrdma/frwr_ops.c >> +++ b/net/sunrpc/xprtrdma/frwr_ops.c >> @@ -270,8 +270,10 @@ frwr_op_init(struct rpcrdma_xprt *r_xprt) >> INIT_LIST_HEAD(&buf->rb_mws); >> INIT_LIST_HEAD(&buf->rb_all); >> >> - i = (buf->rb_max_requests + 1) * RPCRDMA_MAX_SEGS; >> - dprintk("RPC: %s: initializing %d FRMRs\n", __func__, i); >> + i = max_t(int, RPCRDMA_MAX_DATA_SEGS / depth, 1); >> + i += 2; /* head + tail */ >> + i *= buf->rb_max_requests; /* one set for each RPC slot */ >> + dprintk("RPC: %s: initalizing %d FRMRs\n", __func__, i); >> >> while (i--) { >> struct rpcrdma_mw *r; >> > > Looks good. > > Reviewed-by: Sagi Grimberg > > -- > To unsubscribe from this list: send the line "unsubscribe linux-rdma" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html -- -Regards Devesh