From mboxrd@z Thu Jan  1 00:00:00 1970
From: Jakub Kicinski <jakub.kicinski@netronome.com>
Subject: Re: libbpf backward compatibility (was: [PATCH bpf-next v5 09/10]
 bpf: btf: Add BTF support to libbpf)
Date: Wed, 9 May 2018 15:20:57 -0700
Message-ID: <20180509152057.2aeae80f@cakuba.netronome.com>
References: <20180418225606.2771620-1-kafai@fb.com>
        <20180418225606.2771620-10-kafai@fb.com>
        <20180509151712.7d6826f7@cakuba.netronome.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Cc: <netdev@vger.kernel.org>, <kernel-team@fb.com>,
        David Beckett <david.beckett@netronome.com>
To: Martin KaFai Lau <kafai@fb.com>, Alexei Starovoitov <ast@fb.com>,
        Daniel Borkmann <daniel@iogearbox.net>,
        Andrey Ignatov <rdna@fb.com>
Return-path: <netdev-owner@vger.kernel.org>
Received: from mail-qt0-f193.google.com ([209.85.216.193]:45504 "EHLO
        mail-qt0-f193.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org
        with ESMTP id S964941AbeEIWVB (ORCPT
        <rfc822;netdev@vger.kernel.org>); Wed, 9 May 2018 18:21:01 -0400
Received: by mail-qt0-f193.google.com with SMTP id j42-v6so128706qtj.12
        for <netdev@vger.kernel.org>; Wed, 09 May 2018 15:21:01 -0700 (PDT)
In-Reply-To: <20180509151712.7d6826f7@cakuba.netronome.com>
Sender: netdev-owner@vger.kernel.org
List-ID: <netdev.vger.kernel.org>

On Wed, 9 May 2018 15:17:12 -0700, Jakub Kicinski wrote:
> On Wed, 18 Apr 2018 15:56:05 -0700, Martin KaFai Lau wrote:
> > diff --git a/tools/lib/bpf/bpf.h b/tools/lib/bpf/bpf.h
> > index 39f6a0d64a3b..01bda076310f 100644
> > --- a/tools/lib/bpf/bpf.h
> > +++ b/tools/lib/bpf/bpf.h
> > @@ -26,6 +26,20 @@
> >  #include <linux/bpf.h>
> >  #include <stddef.h>
> >  
> > +struct bpf_create_map_attr {
> > +	const char *name;
> > +	enum bpf_map_type map_type;
> > +	__u32 map_flags;
> > +	__u32 key_size;
> > +	__u32 value_size;
> > +	__u32 max_entries;
> > +	__u32 numa_node;
> > +	__u32 btf_fd;
> > +	__u32 btf_key_id;
> > +	__u32 btf_value_id;
> > +};
> > +
> > +int bpf_create_map_xattr(const struct bpf_create_map_attr *create_attr);
> >  int bpf_create_map_node(enum bpf_map_type map_type, const char *name,
> >  			int key_size, int value_size, int max_entries,
> >  			__u32 map_flags, int node);
> > @@ -87,4 +101,6 @@ int bpf_obj_get_info_by_fd(int prog_fd, void *info, __u32 *info_len);
> >  int bpf_prog_query(int target_fd, enum bpf_attach_type type, __u32 query_flags,
> >  		   __u32 *attach_flags, __u32 *prog_ids, __u32 *prog_cnt);
> >  int bpf_raw_tracepoint_open(const char *name, int prog_fd);
> > +int bpf_load_btf(void *btf, __u32 btf_size, char *log_buf, __u32 log_buf_size,
> > +		 bool do_log);
> >  #endif  
> 
> Does libbpf have to provide backward compatibility?  Are the function
> prototypes not supposed to change?
> 
> Recently a number of *_xattr functions were added, these are nice for
> limiting the number of parameters passed to functions, but they don't
> buy us anything in terms of extensibility unless we can grow the
> structures.  Should structure size be passed in as one of parameters?

Or is the backward compatibility not at the binary level so we can grow
the structures at will?