netdev - Re: rtnetlink problems with Cisco enic and VFs

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]

Message-Id: <20140423092458.d9b5ee99d389e30e87b86376@redhat.com>
Date:	Wed, 23 Apr 2014 09:24:58 +1000
From:	David Gibson <dgibson@...hat.com>
To:	"Christian Benvenuti (benve)" <benve@...co.com>
Cc:	"netdev@...r.kernel.org" <netdev@...r.kernel.org>,
	"Sujith Sankar (ssujith)" <ssujith@...co.com>,
	Govindarajulu Varadarajan <govindarajulu90@...il.com>,
	"Neel Patel (neepatel)" <neepatel@...co.com>,
	Nishank Trivedi <nistrive@...co.com>
Subject: Re: rtnetlink problems with Cisco enic and VFs

On Tue, 22 Apr 2014 19:14:04 +0000
"Christian Benvenuti (benve)" <benve@...co.com> wrote:

> David,
> 
> > -----Original Message-----
> > From: David Gibson [mailto:dgibson@...hat.com]
> > Sent: Monday, April 21, 2014 9:14 PM
> > To: netdev@...r.kernel.org
> > Cc: Christian Benvenuti (benve); Sujith Sankar (ssujith); Govindarajulu
> > Varadarajan; Neel Patel (neepatel); Nishank Trivedi
> > Subject: RFC: rtnetlink problems with Cisco enic and VFs
> > 
> > I believe I've found a problem with netlink handling which can be triggered
> > on Cisco enic devices with a large number (30-40) of virtual functions.  I
> > believe this is the cause of a real customer problem we've seen.
> > 
> >  * When requesting a list of interfaces with RTM_GETLINK, enic devices
> >    (and currently, _only_ enic devices) report IFLA_VF_PORTS
> >    information
> 
> Is the fact that Enic is the only driver implementing ndo_get_vf_port [1]
> the root cause of the problem and the reason why this happens only with Enic?

Yes, or at least it's one of the factors.

> 
> /Chris
> 
> [1]
> This is what makes rtnl_port_size to account for vf_port_size*dev_num_vf(...)
> and rtnl_port_fill to add IFLA_VF_PORTS.

-- 
David Gibson <dgibson@...hat.com>

Content of type "application/pgp-signature" skipped