From mboxrd@z Thu Jan 1 00:00:00 1970 From: Thomas Monjalon Subject: Re: [PATCH] ethdev: add siblings iterator Date: Tue, 11 Dec 2018 19:19:52 +0100 Message-ID: <1653320.uaj56YXoz4@xps> References: <20181130002716.27325-1-thomas@monjalon.net> <3b5368bf-400d-66e0-d6a8-19bdb33351dc@intel.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7Bit Cc: arybchenko@solarflare.com, dev@dpdk.org To: Ferruh Yigit Return-path: Received: from out1-smtp.messagingengine.com (out1-smtp.messagingengine.com [66.111.4.25]) by dpdk.org (Postfix) with ESMTP id E5C602C2F for ; Tue, 11 Dec 2018 19:19:58 +0100 (CET) In-Reply-To: <3b5368bf-400d-66e0-d6a8-19bdb33351dc@intel.com> List-Id: DPDK patches and discussions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: dev-bounces@dpdk.org Sender: "dev" 11/12/2018 17:31, Ferruh Yigit: > On 11/30/2018 12:27 AM, Thomas Monjalon wrote: > > If multiple ports share the same hardware device (rte_device), > > they are siblings and can be found thanks to the new function > > and loop macro. > > > > The ownership is not checked because siblings may have > > different owners. > > Looks good on its own, but I think now we require an implementation of any new > API, so it can be good to have: > - a sample implementation of this new API and the macro > - an unit test for the API and the macro Yes sure. I should have added "RFC" in the title. v2 will have some usage of this API. About the unit test, I'm really not sure whether we should test the ehtdev API in test/test/ or just inside testpmd. We used to implement ethdev tests only in testpmd. Opinions?