From mboxrd@z Thu Jan 1 00:00:00 1970 From: Prashant Sreedharan Subject: Re: Question regarding {G,S}CHANNELS API Date: Fri, 8 Jan 2016 10:17:13 -0800 Message-ID: <1452277033.13470.43.camel@prashant> References: <20160107154324.4081743c@jkicinski-Precision-T1700> <1452201864.13470.27.camel@prashant> <20160108114139.74a35521@jkicinski-Precision-T1700> Mime-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 7bit Cc: "netdev@vger.kernel.org" , David Miller , Michael Chan , Ben Hutchings , To: Jakub Kicinski Return-path: Received: from mail-gw3-out.broadcom.com ([216.31.210.64]:54122 "EHLO mail-gw3-out.broadcom.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754891AbcAHSk1 (ORCPT ); Fri, 8 Jan 2016 13:40:27 -0500 In-Reply-To: <20160108114139.74a35521@jkicinski-Precision-T1700> Sender: netdev-owner@vger.kernel.org List-ID: On Fri, 2016-01-08 at 11:41 +0000, Jakub Kicinski wrote: > The point of my email was that it's a bit cumbersome to express such > configuration with the current ethtool API. Therefore I'm asking if > it's preferred to use: > ethtool ethX -L rx N tx M For tg3 driver this is the preferred way > in this case, or: > ethtool ethX -L combined N tx/rx abs(M-N) tg3 will ignore the combined option and only consider separate tx / rx values here. Combined will always be displayed as 0 to the user. We understand your concern, and agree that the internal implementation does not assign separate irq's to each tx, rx rings when tx/rx option is used by user. With limited number of status blocks (max 4) and a HW bug due to which not more than 1 TX ring is recommended to be enabled, it was implemented in this manner.