From mboxrd@z Thu Jan 1 00:00:00 1970 From: Thomas Monjalon Subject: Re: [PATCH 0/2] fix performance/cache resource issues with 128-byte cache line targets Date: Mon, 07 Dec 2015 03:40:13 -0800 (PST) Message-ID: <1624480.dTEes5qyS3@xps13> References: <1449417564-29600-1-git-send-email-jerin.jacob@caviumnetworks.com> <1631625.Dg38D5sl3P@xps13> <20151207072644.GA3950@localhost.localdomain> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7Bit Cc: dev@dpdk.org To: Jerin Jacob Return-path: Received: from mail-wm0-f47.google.com (mail-wm0-f47.google.com [74.125.82.47]) by dpdk.org (Postfix) with ESMTP id C3AB795AA for ; Mon, 7 Dec 2015 12:40:13 +0100 (CET) Received: by wmww144 with SMTP id w144so146296445wmw.0 for ; Mon, 07 Dec 2015 03:40:13 -0800 (PST) In-Reply-To: <20151207072644.GA3950@localhost.localdomain> List-Id: patches and discussions about DPDK List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: dev-bounces@dpdk.org Sender: "dev" 2015-12-07 12:56, Jerin Jacob: > On Sun, Dec 06, 2015 at 05:30:50PM +0100, Thomas Monjalon wrote: > > 2015-12-06 21:29, Jerin Jacob: > > > This patchset fixes performance/cache resource issues with 128-byte cache line targets > > > found in mbuf and bitmap DPDK libraries > > > > > > Currently, we have two DPDK targets(ThunderX and ppc_64) which are based on > > > 128-bytes cache line size target. > > > > When introducing IBM Power8, we failed to clean the cache line size definition. > > I promised to not forget this issue in this thread with Neil: > > http://dpdk.org/ml/archives/dev/2014-December/009439.html > > > > It is defined in > > config/defconfig_* > > mk/machine/*/rte.vars.mk > > mk/arch/*/rte.vars.mk > > rte_memory.h > > rte_kni_common.h > > > > It should be defined only in the config files. > > When we will introduce a configure script, we should be able to detect it. > > > > Please Jerin, as ThunderX maintainer, may you help to fix this old mess? > > Yes Thomas, I will takeup this issue when we will have configure script. I thought we could start setting the value in only one place. The detection in configure script would be another step. > apart from that, content of the this patch will be still valid > as the fix going to be generating cache line define from the config file.