netdev - Re: ipv6 sysctl

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]

Date:   Thu, 2 Mar 2017 14:23:24 -0800
From:   Ani Sinha <ani@...sta.com>
To:     "netdev@...r.kernel.org" <netdev@...r.kernel.org>,
        David Miller <davem@...emloft.net>
Cc:     fruggeri <fruggeri@...sta.com>, Jeff Chan <jchan@...sta.com>,
        gilligan <gilligan@...sta.com>
Subject: Re: ipv6 sysctl

Hey netdev guys,

Any feedback on this? :-)

thanks
ani


On Tue, Feb 28, 2017 at 11:22 AM, Ani Sinha <ani@...sta.com> wrote:
> Hi guys,
>
> Commit a79ca223e029 ('ipv6: fix bad free of addrconf_init_net')
> introduced in linux 3.9 tries to fix an issue involving free-ing
> statically allocated memory. Additionally, it subtly changes behavior
> of how certain ipv6 sysctl values are inherited from the default net
> namespace to the child namespaces.   Before a79ca223e029, the default
> namespace would directly modify the values in statically allocated
> struct ipv6_devconf for example and all child namespaces would inherit
> these values upon creation (their own private copy was initialized
> using the statically allocated ipv6_devconf). After this change, any
> sysctl value changes in default net namespace is not seen by any new
> child namespaces that are created afterwards. This is because all
> network namespaces, including the default namespace has it's own
> private copy of  struct ipv6_devconf which is initialized by certain
> fixed values. This is in contrast to what we have in ipv4 where child
> namespaces continues to inherit values from the default namespace upon
> creation.
>
> I see that there was a previous discussion here :
> https://patchwork.kernel.org/patch/4639391/
>
> Was the above inconsistency between ipv4 and ipv6 sysctl
> initialization intentional or was it an unintended effect of the above
> change ? It would be nice to have a symmetric behavior between ipv4
> and ipv6. Please share your thoughts on this.
>
> thanks,
> ani