netdev - Re: [PATCH iproute2] devlink: Ignore unknown attributes

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Fri, 19 Jan 2018 14:27:06 -0800
From:   David Ahern <dsahern@...il.com>
To:     Stephen Hemminger <stephen@...workplumber.org>
Cc:     Arkadi Sharshevsky <arkadis@...lanox.com>, netdev@...r.kernel.org,
        davem@...emloft.net, mlxsw@...lanox.com
Subject: Re: [PATCH iproute2] devlink: Ignore unknown attributes

On 1/19/18 2:02 PM, Stephen Hemminger wrote:
> On Thu, 18 Jan 2018 20:42:44 -0800
> David Ahern <dsahern@...il.com> wrote:
> 
>> On 1/17/18 5:28 AM, Arkadi Sharshevsky wrote:
>>> In case of extending the UAPI old packages would break.
>>>
>>> Signed-off-by: Arkadi Sharshevsky <arkadis@...lanox.com>
>>> ---
>>>  devlink/devlink.c | 2 +-
>>>  1 file changed, 1 insertion(+), 1 deletion(-)
>>>
>>> diff --git a/devlink/devlink.c b/devlink/devlink.c
>>> index 39cda06..c9d1838 100644
>>> --- a/devlink/devlink.c
>>> +++ b/devlink/devlink.c
>>> @@ -343,7 +343,7 @@ static int attr_cb(const struct nlattr *attr, void *data)
>>>  	int type;
>>>  
>>>  	if (mnl_attr_type_valid(attr, DEVLINK_ATTR_MAX) < 0)
>>> -		return MNL_CB_ERROR;
>>> +		return MNL_CB_OK;
>>>  
>>>  	type = mnl_attr_get_type(attr);
>>>  	if (mnl_attr_validate(attr, devlink_policy[type]) < 0)
>>>   
>>
>> What's the point of calling mnl_attr_type_valid if you disregard a
>> failure? you might as well not call mnl_attr_type_valid at all.
> 
> The way mnl handles attributes, you have to have a callback and it is up
> to the callback to copy the values it wants.  The idea is that old code
> running against a newer kernel will have a smaller array of attributes
> it wants, and only copy those.
> 

mnl_attr_type_valid calls mnl_attr_get_type which does attr->nla_type &
NLA_TYPE_MASK. Since you are no longer acknowledging the return code of
mnl_attr_type_valid, you don't care about its checks so you might as
well not call it. I don't see anything in libmnl that checks that
mnl_attr_type_valid is invoked on an attr, so hence my question -- given
the change above why call it all?