From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Burakov, Anatoly" Subject: Re: [PATCH 18.05 v4] eal: add function to return number of detected sockets Date: Thu, 8 Mar 2018 14:38:37 +0000 Message-ID: References: <750e30c6dcc7a22a87df9c56fb824042b1db984f.1517848624.git.anatoly.burakov@intel.com> <20180308121229.GA8660@bricha3-MOBL3.ger.corp.intel.com> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8; format=flowed Content-Transfer-Encoding: 7bit Cc: dev@dpdk.org To: Bruce Richardson Return-path: Received: from mga06.intel.com (mga06.intel.com [134.134.136.31]) by dpdk.org (Postfix) with ESMTP id 1E6AC4CA1 for ; Thu, 8 Mar 2018 15:38:39 +0100 (CET) In-Reply-To: <20180308121229.GA8660@bricha3-MOBL3.ger.corp.intel.com> Content-Language: en-US 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 08-Mar-18 12:12 PM, Bruce Richardson wrote: > On Wed, Feb 07, 2018 at 09:58:36AM +0000, Anatoly Burakov wrote: >> During lcore scan, find maximum socket ID and store it. This will >> break the ABI, so bump ABI version. >> >> Signed-off-by: Anatoly Burakov >> --- >> >> Notes: >> v4: >> - Remove backwards ABI compatibility, bump ABI instead >> >> v3: >> - Added ABI compatibility >> >> v2: >> - checkpatch changes >> - check socket before deciding if the core is not to be used >> >> lib/librte_eal/bsdapp/eal/Makefile | 2 +- >> lib/librte_eal/common/eal_common_lcore.c | 37 +++++++++++++++++++++---------- >> lib/librte_eal/common/include/rte_eal.h | 1 + >> lib/librte_eal/common/include/rte_lcore.h | 8 +++++++ >> lib/librte_eal/linuxapp/eal/Makefile | 2 +- >> lib/librte_eal/rte_eal_version.map | 9 +++++++- >> 6 files changed, 44 insertions(+), 15 deletions(-) >> > Breaking the ABI is the best way to implement this change, and given the > deprecation was previously announced I'm ok with that. > > Question: we are ok assuming that the socket numbers are sequential, or > nearly so, and knowing the maximum socket number seen is a good > approximation of the actual physical sockets? I know in terms of cores > on a system, the core id's often jump - are there systems where the > socket numbers do too? > > /Bruce > I am not aware of any system that would jump sockets like that. I'm open to corrections, however :) -- Thanks, Anatoly