linux-kernel - Re: [PATCH] ixgbe/ipsec: use memzero

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]

Message-Id: <20250516150400.512375-1-zilin@seu.edu.cn>
Date: Fri, 16 May 2025 15:04:00 +0000
From: Zilin Guan <zilin@....edu.cn>
To: przemyslaw.kitszel@...el.com
Cc: andrew+netdev@...n.ch,
	anthony.l.nguyen@...el.com,
	davem@...emloft.net,
	dawid.osuchowski@...ux.intel.com,
	edumazet@...gle.com,
	intel-wired-lan@...ts.osuosl.org,
	jianhao.xu@....edu.cn,
	kuba@...nel.org,
	linux-kernel@...r.kernel.org,
	netdev@...r.kernel.org,
	pabeni@...hat.com,
	zilin@....edu.cn
Subject: Re: [PATCH] ixgbe/ipsec: use memzero_explicit() for stack SA structs

On Thu, May 15, 2025 at 11:27:22AM+0200, Przemek Kitszel wrote:
> the general rule is to memzero_explicit() memory that was holding secure
> content
> --
> to have full picture: it is fine to memset() such storage prior to use,
> it is also fine to combine related changes in one commit/one series
> 
> re stated purpose of the patch:
> I see @rsa cleaned in just one exit point of ixgbe_ipsec_add_sa(),
> instead of all of them, so v2 seems warranted

Hi Przemek,

Thank you for your detailed feedback and clarification.

As Dawid pointed out, while @rsa is cleared at one exit point in 
ixgbe_ipsec_add_sa(), another exit path, at which we fail to acquire the 
RX SA table, leaves rsa.key and rsa.salt zeroed. Does this imply there's
no sensitive data to clear in this case? If so, would using memset() on 
the symmetric error path in @tsa be redundant, or am I overlooking 
something?

I'd appreciate your thoughts on this.

Best regards,
Zilin Guan