netdev - RE: hv_netvsc calls RTM

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: <2b804c0c102e471ebb6a933995f7dba6@DFM-DB3MBX15-06.exchange.corp.microsoft.com>
Date:	Tue, 18 Mar 2014 19:23:42 +0000
From:	Haiyang Zhang <haiyangz@...rosoft.com>
To:	Olaf Hering <olaf@...fle.de>, KY Srinivasan <kys@...rosoft.com>
CC:	"netdev@...r.kernel.org" <netdev@...r.kernel.org>
Subject: RE: hv_netvsc calls RTM_NEWLINK with eth%d



> -----Original Message-----
> From: Olaf Hering [mailto:olaf@...fle.de]
> Sent: Tuesday, March 18, 2014 2:52 PM
> To: Haiyang Zhang; KY Srinivasan
> Cc: netdev@...r.kernel.org
> Subject: hv_netvsc calls RTM_NEWLINK with eth%d
> 
> 
> For some reason 'modprobe -r hv_netvsc;modprobe hv_netvsc' causes a
> RTM_NEWLINK message with ifname "eth%d". This confuses applications
> reacting to RTM_NEWLINK messages. At the time they actually get to
> processing the message eth%d is already renamed to eth0 (I think).
> 
> I added lots of debug to trace RTM_NEWLINK calls, and it turned out that
> the event is issued before register_netdevice calls dev_get_valid_name.
> 
> Is there a chance to rearrange the calls in netvsc_probe, so that
> RTM_NEWLINK messages with ifname "eth%d" do not happen anymore?

Yes, and actually I have already added a condition in netvsc_linkstatus_callback(),
so that netdev_notify_peers() won't be called before netdevice is registered.

The commit:
http://git.kernel.org/cgit/linux/kernel/git/next/linux-next.git/commit/drivers/net/hyperv?id=891de74d693bb4fefe2efcc6432a4a9a9bee561e

Thanks,
- Haiyang