netdev - Re: [PATCH iproute2 v2] ip: Extend filter links/addresses

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] [day] [month] [year] [list]

Message-ID: <6421a75c-0341-4813-7c12-5836a440df76@gmail.com>
Date:   Wed, 12 Jan 2022 10:57:49 -0700
From:   David Ahern <dsahern@...il.com>
To:     Anton Danilov <littlesmilingcloud@...il.com>
Cc:     Stephen Hemminger <stephen@...workplumber.org>,
        Linux Kernel Network Developers <netdev@...r.kernel.org>
Subject: Re: [PATCH iproute2 v2] ip: Extend filter links/addresses

On 1/11/22 12:09 PM, Anton Danilov wrote:
> Hello, David.
> 
>> current 'type' filtering is the 'kind' string in the rtnl_link_ops --
>> bridge, veth, vlan, vrf, etc. You are now wanting to add 'exclude_type'
>> and make it based on hardware type. That is a confusing user api.
> 
> The 'exclude_type' options first checks the 'kind' in the
> rtnl_link_ops, then the hardware type.

ok; missed that on the first pass.

Update the commit message to say filtering by "hardware type"

>> On 1/8/22 12:58 PM, Anton Danilov wrote:
>>> @@ -227,6 +227,28 @@ static int match_link_kind(struct rtattr **tb, const char *kind, bool slave)
>>>       return strcmp(parse_link_kind(tb[IFLA_LINKINFO], slave), kind);
>>>  }
>>>
>>> +static int match_if_type_name(unsigned short if_type, const char *type_name)
>>> +{
>>> +
>>> +     char *expected_type_name;
>>> +
>>> +     switch (if_type) {
>>> +     case ARPHRD_ETHER:
>>> +             expected_type_name = "ether";
>>> +             break;
>>> +     case ARPHRD_LOOPBACK:
>>> +             expected_type_name = "loopback";
>>> +             break;
>>> +     case ARPHRD_PPP:
>>> +             expected_type_name = "ppp";

ppp devices have kind set, so ARPHRD_PPP should not be needed.


Also, you have supported hardware types in multiple places - this match
function and the filter.kind check. Make 1 table with supported types
and use that table with helpers or both paths.

Why not allow exclude by "_slave" type? e.g., should all devices but
bridge ports?