From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jakub Kicinski Subject: Re: [PATCH bpf-next 0/2] libbpf: adding af_xdp support Date: Mon, 10 Dec 2018 16:10:49 -0800 Message-ID: <20181210161049.7b66cc86@cakuba.netronome.com> References: <1544456085-14258-1-git-send-email-magnus.karlsson@intel.com> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Cc: bjorn.topel@intel.com, ast@kernel.org, daniel@iogearbox.net, netdev@vger.kernel.org, bjorn.topel@gmail.com, qi.z.zhang@intel.com, brouer@redhat.com To: Magnus Karlsson Return-path: Received: from mail-qt1-f195.google.com ([209.85.160.195]:42237 "EHLO mail-qt1-f195.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1728707AbeLKAKy (ORCPT ); Mon, 10 Dec 2018 19:10:54 -0500 Received: by mail-qt1-f195.google.com with SMTP id d19so14513171qtq.9 for ; Mon, 10 Dec 2018 16:10:53 -0800 (PST) In-Reply-To: <1544456085-14258-1-git-send-email-magnus.karlsson@intel.com> Sender: netdev-owner@vger.kernel.org List-ID: On Mon, 10 Dec 2018 16:34:43 +0100, Magnus Karlsson wrote: > * Is it ok to have static inline functions in this library? They are > currently static inline for performance reasons, but maybe -flto > could fix this but not everyone uses this. I'd think so, we have few static inlines in kernel uapi headers. > * I have included 3 more header files compared to libbpf without AF_XDP > functionality: barrier.h (for the memory barriers used for correctly > ordered accesses to the rings) compiler.h (for one unliekly and one > likely with a tiny performance impact, if any) and list.h (some > extra functions). What to do with these (and the header files they > include) as they need to be dual licensed for libbpf.so? Indeed, but I think we already have that problem, we include barrier.h indirectly for the perf ring helper which in turn includes compiler.h and list.h is included directly. Thanks for the work!