[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <acd099f4-afc5-0b35-b6a1-0bf9a17bb245@iogearbox.net>
Date: Tue, 20 Feb 2018 23:12:27 +0100
From: Daniel Borkmann <daniel@...earbox.net>
To: Arnd Bergmann <arnd@...db.de>
Cc: Alexei Starovoitov <ast@...nel.org>,
"David S. Miller" <davem@...emloft.net>,
John Fastabend <john.fastabend@...il.com>,
Lawrence Brakmo <brakmo@...com>,
Jesper Dangaard Brouer <brouer@...hat.com>,
Jakub Kicinski <jakub.kicinski@...ronome.com>,
Networking <netdev@...r.kernel.org>,
Linux Kernel Mailing List <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] bpf: hide a possibly unused variable
On 02/20/2018 11:08 PM, Arnd Bergmann wrote:
> On Tue, Feb 20, 2018 at 10:44 PM, Daniel Borkmann <daniel@...earbox.net> wrote:
>> Hi Arnd,
>>
>> On 02/20/2018 10:16 PM, Arnd Bergmann wrote:
>>> The only user of this variable is inside of an #ifdef, causing
>>> a warning without CONFIG_INET:
>>>
>>> net/core/filter.c: In function '____bpf_sock_ops_cb_flags_set':
>>> net/core/filter.c:3382:6: error: unused variable 'val' [-Werror=unused-variable]
>>> int val = argval & BPF_SOCK_OPS_ALL_CB_FLAGS;
>>>
>>> This adds the same #ifdef around the declaration.
>>>
>>> Fixes: b13d88072172 ("bpf: Adds field bpf_sock_ops_cb_flags to tcp_sock")
>>> Signed-off-by: Arnd Bergmann <arnd@...db.de>
>>> ---
>>> net/core/filter.c | 2 ++
>>> 1 file changed, 2 insertions(+)
>>>
>>> diff --git a/net/core/filter.c b/net/core/filter.c
>>> index 08ab4c65a998..c3dc6d60b4bb 100644
>>> --- a/net/core/filter.c
>>> +++ b/net/core/filter.c
>>> @@ -3379,7 +3379,9 @@ BPF_CALL_2(bpf_sock_ops_cb_flags_set, struct bpf_sock_ops_kern *, bpf_sock,
>>> int, argval)
>>> {
>>> struct sock *sk = bpf_sock->sk;
>>> +#ifdef CONFIG_INET
>>> int val = argval & BPF_SOCK_OPS_ALL_CB_FLAGS;
>>> +#endif
>>
>> Looks good, thanks for the fix!
>>
>> Could you move the existing '#ifdef CONFIG_INET' to the beginning of
>> the function given the only error in case of !CONFIG_INET is -EINVAL
>> anyway? That would at least not increase the ifdef ugliness further.
>
> Sure, sent a new version now. I decided to clean up that #ifdef
> check as well, since a IS_ENABLED() check is nicer anway.
Looks great, thanks! I'll get it into bpf once the pending pr got pulled.
Powered by blists - more mailing lists