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
| ||
|
Message-ID: <CAM_iQpW0fJJXPS=FoTf-o4f5-mOC=mZYNP8=uqiUOmvX+CDwWw@mail.gmail.com> Date: Thu, 11 May 2017 17:07:15 -0700 From: Cong Wang <xiyou.wangcong@...il.com> To: Julian Anastasov <ja@....bg> Cc: Eric Dumazet <eric.dumazet@...il.com>, David Miller <davem@...emloft.net>, Linux Kernel Network Developers <netdev@...r.kernel.org>, Andrey Konovalov <andreyknvl@...gle.com>, Eric Dumazet <edumazet@...gle.com> Subject: Re: [Patch net] ipv4: restore rt->fi for reference counting On Wed, May 10, 2017 at 12:51 PM, Julian Anastasov <ja@....bg> wrote: > Oh, well, the sockets can hold cached dst. > But if the promise is that rt->fi is used only as > reference to metrics we have to find a way to drop > the dev references at NETDEV_UNREGISTER time. If you > set nh_dev to NULL then all lookups should check it > for != NULL. The sockets will not walk NHs via rt->fi, > i.e. the route lookups will get valid res.fi from trees, > so it may work in this way. > So, if I understand you correctly it is safe to NULL'ing nh_dev in NETDEV_UNREGISTER_FINAL, right? If still not, how about transfer nh_dev's to loopback_dev too in NETDEV_UNREGISTER? Like we transfer dst->dev. I don't want to touch the fast path to check for NULL, as it will change more code and slow down performance. Thanks.
Powered by blists - more mailing lists