From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from dpdk.org (dpdk.org [92.243.14.124]) by inbox.dpdk.org (Postfix) with ESMTP id 4C524A04C7; Tue, 15 Sep 2020 00:03:50 +0200 (CEST) Received: from [92.243.14.124] (localhost [127.0.0.1]) by dpdk.org (Postfix) with ESMTP id DE980FFA; Tue, 15 Sep 2020 00:03:49 +0200 (CEST) Received: from mga06.intel.com (mga06.intel.com [134.134.136.31]) by dpdk.org (Postfix) with ESMTP id E942AE07 for ; Tue, 15 Sep 2020 00:03:47 +0200 (CEST) IronPort-SDR: 3iKsk8k6dUbRoSAcd3ing0MxyX6RWQFDUoe+c+8I1XDPKfYUunoMj5nbYsCk4YzA0PA19JvWkM NkadcDVwqJiA== X-IronPort-AV: E=McAfee;i="6000,8403,9744"; a="220723159" X-IronPort-AV: E=Sophos;i="5.76,427,1592895600"; d="scan'208";a="220723159" X-Amp-Result: SKIPPED(no attachment in message) X-Amp-File-Uploaded: False Received: from fmsmga005.fm.intel.com ([10.253.24.32]) by orsmga104.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 14 Sep 2020 15:03:46 -0700 IronPort-SDR: CD6/qkhOcMdSau/3OGbxGrT3CMFb+ESquH1nlyBG6pfNUh0CZ0vDtU4Gv4Oe/+u4iHV8MN5oSP PLDxv7Cu5znA== X-IronPort-AV: E=Sophos;i="5.76,427,1592895600"; d="scan'208";a="507304649" Received: from fyigit-mobl1.ger.corp.intel.com (HELO [10.213.247.225]) ([10.213.247.225]) by fmsmga005-auth.fm.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 14 Sep 2020 15:03:45 -0700 To: Lance Richardson , Ajit Khaparde , Somnath Kotur Cc: dev@dpdk.org, rasland@nvidia.com References: <20200909155302.28656-1-lance.richardson@broadcom.com> <20200909155302.28656-6-lance.richardson@broadcom.com> From: Ferruh Yigit Message-ID: <4152848a-c2d9-6e96-d931-32f774dbc122@intel.com> Date: Mon, 14 Sep 2020 23:03:41 +0100 User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:78.0) Gecko/20100101 Thunderbird/78.2.2 MIME-Version: 1.0 In-Reply-To: <20200909155302.28656-6-lance.richardson@broadcom.com> Content-Type: text/plain; charset=utf-8 Content-Language: en-US Content-Transfer-Encoding: 8bit Subject: Re: [dpdk-dev] [PATCH 05/12] net/bnxt: improve support for small ring sizes X-BeenThere: dev@dpdk.org X-Mailman-Version: 2.1.15 Precedence: list List-Id: DPDK patches and discussions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: dev-bounces@dpdk.org Sender: "dev" On 9/9/2020 4:52 PM, Lance Richardson wrote: > Improve support for small ring sizes: > - Ensure that transmit free threshold is no more than 1/4 ring size. > - Ensure that receive free threshold is no more than 1/4 ring size. > - Validate requested ring sizes against minimum supported size. > - Use rxq receive free threshold instead of fixed maximum burst > size to trigger bulk receive buffer allocation. > > Reviewed-by: Ajit Kumar Khaparde > Signed-off-by: Lance Richardson > --- > drivers/net/bnxt/bnxt_rxq.c | 6 +++-- > drivers/net/bnxt/bnxt_rxtx_vec_common.h | 10 +++++---- > drivers/net/bnxt/bnxt_rxtx_vec_neon.c | 29 +++++++++++++------------ > drivers/net/bnxt/bnxt_rxtx_vec_sse.c | 29 +++++++++++++------------ > drivers/net/bnxt/bnxt_txq.c | 7 ++++-- > 5 files changed, 45 insertions(+), 36 deletions(-) > > diff --git a/drivers/net/bnxt/bnxt_rxq.c b/drivers/net/bnxt/bnxt_rxq.c > index db9aa1f3ed..4ef3b5cb5c 100644 > --- a/drivers/net/bnxt/bnxt_rxq.c > +++ b/drivers/net/bnxt/bnxt_rxq.c > @@ -14,6 +14,7 @@ > #include "bnxt_rxq.h" > #include "bnxt_rxr.h" > #include "bnxt_vnic.h" > +#include "bnxt_rxtx_vec_common.h" > #include "hsi_struct_def_dpdk.h" > Hi Lance, Ajit, Somnath, Raslan reported a build error for PPC architecture, this happens because static inline functions in 'bnxt_rxtx_vec_common.h' are using 'rxq->rxrearm_nb' & 'rxq->rxrearm_start' which are not defined for PPC. As far as I can see the 'bnxt_rxtx_vec_common.h' is included because of some macros. A quick fix can be to wrap all static inline functions with !PPC checks but from the name of the header file, it looks like it shouldn't be included by scalar datapath .c files at first place. Instead it can be possible to extract those macros into another header and both these .c file and 'bnxt_rxtx_vec_common.h' can include it. Or 'bnxt_txq.h' & 'bnxt_txr.h' can be used to hold those macros, and those headers were already included by 'bnxt_rxtx_vec_common.h' and .c files. Anyway can you please provide the fix as incremental patches on top of latest head, so I can squash them into original patches. If the patches can't be done as incremental fixes or if they will delay, I will need to drop the patchset from next-net, to not block any possible pull from main repo.