netdev - Re: [net-next 0/9][pull request] Intel Wired LAN Driver Updates

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]

Date:	Thu, 2 Jan 2014 19:06:32 +0000
From:	"Rustad, Mark D" <mark.d.rustad@...el.com>
To:	Ben Hutchings <bhutchings@...arflare.com>
CC:	"Kirsher, Jeffrey T" <jeffrey.t.kirsher@...el.com>,
	David Miller <davem@...emloft.net>,
	"<netdev@...r.kernel.org>" <netdev@...r.kernel.org>,
	"gospo@...hat.com" <gospo@...hat.com>,
	"sassmann@...hat.com" <sassmann@...hat.com>
Subject: Re: [net-next 0/9][pull request] Intel Wired LAN Driver Updates

On Jan 2, 2014, at 10:21 AM, Ben Hutchings <bhutchings@...arflare.com> wrote:

> On Thu, 2014-01-02 at 01:33 -0800, Jeff Kirsher wrote:
>> This series contains updates to ixgbe, e1000e and igb.
>> 
>> Most notably is Mark Rustad's ixgbe patches to add the Live Error
>> Recovery option to ixgbe as well as additional cleanups to the driver.
> [...]
> 
> The Live Error Recovery option is pretty odd; it looks like a
> combination of PCI hotplug and EEH.  Why is it not integrated with
> either of those?

It certainly is related to both of those. The issue being addressed here is specific to this driver in that the driver can be the first to get a read return of all one’s when a device is suddenly removed. This code allows the driver to determine whether such a return is valid, or is a sign of the device having been removed. Without those checks, there are paths where the driver will crash in such a case. This code will race with hotplug, so it allows hotplug to drive the logical removal, but it makes the driver “safe” in the mean time.

Thanks for the comments. You are absolutely right about the need for ACCESS_ONCE and better Kconfig help text.

> Ben.
> 
> -- 
> Ben Hutchings, Staff Engineer, Solarflare
> Not speaking for my employer; that's the marketing department's job.
> They asked us to note that Solarflare product names are trademarked.

-- 
Mark Rustad, Networking Division, Intel Corporation

Download attachment "signature.asc" of type "application/pgp-signature" (842 bytes)