netdev - Re: [PATCH net] wireguard: remove peer cache in netns_pre

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:   Thu, 02 Sep 2021 18:26:23 +0200
From:   Toke Høiland-Jørgensen <toke@...hat.com>
To:     "Jason A. Donenfeld" <Jason@...c4.com>,
        Hangbin Liu <liuhangbin@...il.com>
Cc:     netdev@...r.kernel.org, David Miller <davem@...emloft.net>,
        Xiumei Mu <xmu@...hat.com>, Paolo Abeni <pabeni@...hat.com>,
        wireguard@...ts.zx2c4.com
Subject: Re: [PATCH net] wireguard: remove peer cache in netns_pre_exit

"Jason A. Donenfeld" <Jason@...c4.com> writes:

> Hi Hangbin,
>
> Thanks for the patch and especially for the test. While I see that
> you've pointed to a real problem, I don't think that this particular way
> of fixing it is correct, because it will cause issues for userspace that
> expects to be able to read back the list of peers for, for example,
> keeping track of the latest endpoint addresses or rx/tx transfer
> quantities.
>
> I think the real solution here is to simply clear the endpoint src cache
> and consequently the dst_cache. This is slightly complicated by the fact
> that dst_cache releases dsts lazily, so I needed to add a little utility
> function for that, but that was pretty easy to do.
>
> Can you take a look at the below patch and let me know if it works for
> you and passes other testing you and Toke might be doing with it? (Also,
> please CC the wireguard mailing list in addition to netdev next time?)
> If the patch looks good to you and works well, I'll include it in the
> next series of wireguard patches I send back out to netdev. I'm back
> from travels next week and will begin working on the next series then.

Ran this through the same series of tests as the previous patch, and
indeed it also seems to resolve the issue, so feel free to add:

Tested-by: Toke Høiland-Jørgensen <toke@...hat.com>

-Toke