linux-kernel - Re: [Intel-wired-lan] [PATCH] e1000e: add option not to verify NVM checksum

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: <978d1158-c419-4a59-b0dd-ad5be9869991@lunn.ch>
Date: Mon, 31 Mar 2025 16:29:59 +0200
From: Andrew Lunn <andrew@...n.ch>
To: "Lifshits, Vitaly" <vitaly.lifshits@...el.com>
Cc: Jacek Kowalski <jacek@...ekk.info>,
	Tony Nguyen <anthony.l.nguyen@...el.com>,
	Przemek Kitszel <przemyslaw.kitszel@...el.com>,
	Andrew Lunn <andrew+netdev@...n.ch>,
	"David S. Miller" <davem@...emloft.net>,
	Eric Dumazet <edumazet@...gle.com>,
	Jakub Kicinski <kuba@...nel.org>, Paolo Abeni <pabeni@...hat.com>,
	intel-wired-lan@...ts.osuosl.org, netdev@...r.kernel.org,
	linux-kernel@...r.kernel.org
Subject: Re: [Intel-wired-lan] [PATCH] e1000e: add option not to verify NVM
 checksum

> From a technical perspective, your patch looks correct. However, if the
> checksum validation is skipped, there is no way to distinguish between the
> simple checksum error described above, and actual NVM corruption, which may
> result in loss of functionality and undefined behavior. This means, that if
> there is any functional issue with the network adapter on a given system,
> while checksum validation was suspended by the user, we will not be able to
> offer support

We have a similar issue with SFP, which contain a checksum. But a few
vendors are lazy, they set a serial number and don't recalculate the
checksum.

We handle this by adding quirks. We know which vendors/products have
FUBAR checksums, and allow them to be used when the checksum is
FUBAR. You could do something similar here, add a list of vendors with
known FUBAR checksums and allow them to be used, but taint the kernel,
and print a warming that the device is unsupported because the vendor
messed up the CRC.

	Andrew