Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Message-ID: <a774a513-284c-eb1f-7578-bb6d475b0509@linux.dev>
Date:   Fri, 16 Sep 2022 14:31:25 -0700
MIME-Version: 1.0
Subject: Re: [PATCH bpf-next] bpf: Move nf_conn extern declarations to
 filter.h
Content-Language: en-US
To:     Kumar Kartikeya Dwivedi <memxor@gmail.com>
Cc:     Daniel Xu <dxu@dxuuu.xyz>, pablo@netfilter.org, fw@strlen.de,
        toke@kernel.org, netfilter-devel@vger.kernel.org,
        netdev@vger.kernel.org, linux-kernel@vger.kernel.org,
        bpf@vger.kernel.org, ast@kernel.org, daniel@iogearbox.net,
        andrii@kernel.org
References: <c4cb11c8ffe732b91c175a0fc80d43b2547ca17e.1662920329.git.dxu@dxuuu.xyz>
 <ada17021-83c9-3dad-5992-4885e824ecac@linux.dev>
 <CAP01T74=btUEPDrz0EVm9wNuMmbbqc2wRvtpJ-Qq45OtasMBZQ@mail.gmail.com>
From:   Martin KaFai Lau <martin.lau@linux.dev>
In-Reply-To: <CAP01T74=btUEPDrz0EVm9wNuMmbbqc2wRvtpJ-Qq45OtasMBZQ@mail.gmail.com>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
Precedence: bulk

On 9/16/22 1:35 PM, Kumar Kartikeya Dwivedi wrote:
> On Fri, 16 Sept 2022 at 22:20, Martin KaFai Lau <martin.lau@linux.dev> wrote:
>>
>> On 9/11/22 11:19 AM, Daniel Xu wrote:
>>> We're seeing the following new warnings on netdev/build_32bit and
>>> netdev/build_allmodconfig_warn CI jobs:
>>>
>>>       ../net/core/filter.c:8608:1: warning: symbol
>>>       'nf_conn_btf_access_lock' was not declared. Should it be static?
>>>       ../net/core/filter.c:8611:5: warning: symbol 'nfct_bsa' was not
>>>       declared. Should it be static?
>>>
>>> Fix by ensuring extern declaration is present while compiling filter.o.
>>>
>>> Signed-off-by: Daniel Xu <dxu@dxuuu.xyz>
>>> ---
>>>    include/linux/filter.h                   | 6 ++++++
>>>    include/net/netfilter/nf_conntrack_bpf.h | 7 +------
>>>    2 files changed, 7 insertions(+), 6 deletions(-)
>>>
>>> diff --git a/include/linux/filter.h b/include/linux/filter.h
>>> index 527ae1d64e27..96de256b2c8d 100644
>>> --- a/include/linux/filter.h
>>> +++ b/include/linux/filter.h
>>> @@ -567,6 +567,12 @@ struct sk_filter {
>>>
>>>    DECLARE_STATIC_KEY_FALSE(bpf_stats_enabled_key);
>>>
>>> +extern struct mutex nf_conn_btf_access_lock;
>>> +extern int (*nfct_bsa)(struct bpf_verifier_log *log, const struct btf *btf,
>>> +                    const struct btf_type *t, int off, int size,
>>> +                    enum bpf_access_type atype, u32 *next_btf_id,
>>> +                    enum bpf_type_flag *flag);
>>
>> Can it avoid leaking the nfct specific details like
>> 'nf_conn_btf_access_lock' and the null checking on 'nfct_bsa' to
>> filter.c?  In particular, this code snippet in filter.c:
>>
>>           mutex_lock(&nf_conn_btf_access_lock);
>>           if (nfct_bsa)
>>                   ret = nfct_bsa(log, btf, ....);
>>          mutex_unlock(&nf_conn_btf_access_lock);
>>
>>
>> Can the lock and null check be done as one function (eg.
>> nfct_btf_struct_access()) in nf_conntrack_bpf.c and use it in filter.c
>> instead?
> 
> Don't think so, no. Because we want nf_conntrack to work as a module as well.
Ah, got it.

I don't see nf_conntrack_btf_struct_access() in nf_conntrack_bpf.h is 
used anywhere.  Can be removed?

> I was the one who suggested nf_conn specific names for now. There is
> no other user of such module supplied
> btf_struct_access callbacks yet, when one appears, we should instead
> make registration of such callbacks properly generic (i.e. also
> enforce it is only for module BTF ID etc.).
> But that would be a lot of code without any users right now.

The lock is the only one needed to be in btf.c and 
nfct_btf_struct_access() can be an inline in nf_conntrack_bpf.h instead?