[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <20170305.175544.721573949168827122.davem@davemloft.net>
Date: Sun, 05 Mar 2017 17:55:44 -0800 (PST)
From: David Miller <davem@...emloft.net>
To: ast@...com
Cc: daniel@...earbox.net, daniel@...que.org, dvyukov@...gle.com,
netdev@...r.kernel.org
Subject: Re: [PATCH net] bpf: add get_next_key callback to LPM map
From: Alexei Starovoitov <ast@...com>
Date: Sun, 5 Mar 2017 10:58:00 -0800
> On 3/5/17 10:44 AM, David Miller wrote:
>> From: Alexei Starovoitov <ast@...com>
>> Date: Sun, 5 Mar 2017 09:41:08 -0800
>>
>>> map_get_next_key callback is mandatory. Supply dummy handler.
>>>
>>> Fixes: b95a5c4db09b ("bpf: add a longest prefix match trie map
>>> implementation")
>>> Reported-by: Dmitry Vyukov <dvyukov@...gle.com>
>>> Signed-off-by: Alexei Starovoitov <ast@...nel.org>
>>
>> Wouldn't it be more appropriate to provide a proper DFS ordered
>> traversal rather than return an error code?
>
> yes. of course. This is quick fix for 'net'.
> That's why the error code is ENOTSUPP with meaning that
> it's not supported right now and will be supported in the future.
Ok, fair enough, applied.
Thanks.
Powered by blists - more mailing lists