netdev - Re: [PATCH net 2/2] geneve: Fix races between socket add and release.

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: <CAEP_g=81tz9UhgCde8h44mCa4GoBQy9t3XezOaQnUbJcjYyWLg@mail.gmail.com>
Date:	Thu, 18 Dec 2014 10:34:39 -0800
From:	Jesse Gross <jesse@...ira.com>
To:	Thomas Graf <tgraf@...g.ch>
Cc:	David Miller <davem@...emloft.net>,
	netdev <netdev@...r.kernel.org>, Andy Zhou <azhou@...ira.com>,
	Stephen Hemminger <stephen@...workplumber.org>
Subject: Re: [PATCH net 2/2] geneve: Fix races between socket add and release.

On Wed, Dec 17, 2014 at 1:15 PM, Thomas Graf <tgraf@...g.ch> wrote:
> On 12/17/14 at 10:48am, Jesse Gross wrote:
>> I generally agree (with the exception of kfree_rcu() - I believe that
>> is still needed since incoming packets reference it using RCU).
>
> I didn't inspect this in full detail but seems like the data path
> should only care about gs->sock which is properly refcnt'ed.

The data path (if you include the consumer in OVS), goes from sk to gs
to gs->rcv_data so it does care about the stuff that is being freed
here.

>> for destroying the socket. This was added by Stephen in "vxlan: listen
>> on multiple ports" but it's not obvious to me what problem it is
>> trying to avoid and I don't see a comment. If possible, it would be
>> nice to simplify this as well if the issue doesn't apply to Geneve.
>
> I don't have an explanation for that either. Each entry on the
> vni_list[] takes a vs->refcnt.

I don't think the workqueue is about refcounting - it looks to me like
the issue that is being avoided is either that RTNL is held and a
lower layer tries to take it again or another lock that nests in the
opposite way with RTNL. However, looking through the code, the problem
wasn't obvious to me.
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html