From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jiri Benc Subject: Re: [PATCH iproute2-next 1/2] ip/tunnel: Be consistent when printing tunnel collect metadata Date: Mon, 22 Jan 2018 17:58:41 +0100 Message-ID: <20180122175841.71400890@redhat.com> References: <1516639614-12960-1-git-send-email-serhe.popovych@gmail.com> <1516639614-12960-2-git-send-email-serhe.popovych@gmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Cc: netdev@vger.kernel.org, u9012063@gmail.com, julien@cumulusnetworks.com To: Serhey Popovych Return-path: Received: from mx1.redhat.com ([209.132.183.28]:47700 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751028AbeAVQ6o (ORCPT ); Mon, 22 Jan 2018 11:58:44 -0500 In-Reply-To: <1516639614-12960-2-git-send-email-serhe.popovych@gmail.com> Sender: netdev-owner@vger.kernel.org List-ID: On Mon, 22 Jan 2018 18:46:53 +0200, Serhey Popovych wrote: > + if (tb[IFLA_GRE_COLLECT_METADATA]) { > + print_bool(PRINT_ANY, "collect_metadata", "external", true); > + return; > + } Nacked-by: Jiri Benc Don't ever use "collect_metadata" for anything visible to the user. collect_metadata is a *horrible* name. It describes the internal implementation of the lwtunneling in the kernel and provides zero explanation to the user about what's that feature good for. The netlink attribute should have never had such name but it's uAPI and we have to live with it. But there's no reason to expose this to the user. Stick with the "external" name. It explains what it is about: instead of the traffic being controlled by the tunnel internal logic (or tunnel control plane, if you want), an external logic needs to be attached to the tunnel in order for the tunneling to work. Jiri