netdev - Re: [PATCH net 2/2] net: ipv6: fix dst ref loops in rpl, seg6 and ioam6 lwtunnels

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: <4dd86c71-b6b7-483d-abe4-0fc40c51bc5e@uliege.be>
Date: Thu, 30 Jan 2025 01:24:00 +0100
From: Justin Iurman <justin.iurman@...ege.be>
To: Jakub Kicinski <kuba@...nel.org>
Cc: davem@...emloft.net, netdev@...r.kernel.org, edumazet@...gle.com,
 pabeni@...hat.com, andrew+netdev@...n.ch, horms@...nel.org,
 dsahern@...nel.org
Subject: Re: [PATCH net 2/2] net: ipv6: fix dst ref loops in rpl, seg6 and
 ioam6 lwtunnels

On 1/29/25 21:14, Jakub Kicinski wrote:
> On Wed, 29 Jan 2025 17:50:14 +0100 Justin Iurman wrote:
>>> +		if (dst->lwtstate != cache_dst->lwtstate) {
>>> +			local_bh_disable();
>>> +			dst_cache_set_ip6(&ilwt->cache, cache_dst, &fl6.saddr);
>>> +			local_bh_enable();
>>> +		}
>>
>> I agree the above patch fixes what kmemleak reported. However, I think
>> it'd bring the double-reallocation issue back when the packet
>> destination did not change (i.e., cache will always be empty). I'll try
>> to come up with a solution...
> 
> True, dunno enough about use cases so I may be missing the point.

Possible use cases: (i) with inline mode, or (ii) with encap mode using 
the same destination address. For (ii), the egress node of the IOAM 
domain also happens to be the actual destination of the packet, but it's 
not "your" packet... so you use a tunnel to stay compliant with RFC8200.

> But the naive solution would be to remember that the tunnel "doesn't
> re-route" and use dst directly, instead of cache_dst?

Correct, that'd work.