netdev - Re: [PATCH net-next] vlan: Invoke driver vlan hooks only if device is present

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: <20151103.111633.1800653969325047017.davem@davemloft.net>
Date:	Tue, 03 Nov 2015 11:16:33 -0500 (EST)
From:	David Miller <davem@...emloft.net>
To:	padmanabh.ratnakar@...gotech.com
Cc:	netdev@...r.kernel.org
Subject: Re: [PATCH net-next] vlan: Invoke driver vlan hooks only if device
 is present

From: Padmanabh Ratnakar <padmanabh.ratnakar@...gotech.com>
Date: Tue,  3 Nov 2015 20:25:59 +0530

> NIC drivers mark device as detached during error recovery.
> It expects no manangement hooks to be invoked in this state.
> Invoke driver vlan hooks only if device is present.
> 
> Signed-off-by: Padmanabh Ratnakar <padmanabh.ratnakar@...gotech.com>

I don't like this.

This approach to solving the issue just peppers checks randomly around
the kernel and the risk for missing cases is very high.  It is also
not clear to me what exactly specifies that these calls are not valid
in such states.

Traditionally we've depended upon the driver to make this kind of check
in it's callback, and that way we don't need to add extra overhead for
everyone at all of these method invokcation points.

Therefore, I would like to see a less dirty, and less error prone
approach to solving this bug.
--
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