linux-kernel - Re: [RFC PATCH] nvme: add an opt-in to use AWUPF

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: <20250821083537.GB29224@lst.de>
Date: Thu, 21 Aug 2025 10:35:37 +0200
From: Christoph Hellwig <hch@....de>
To: John Garry <john.g.garry@...cle.com>
Cc: kbusch@...nel.org, axboe@...nel.dk, hch@....de, sagi@...mberg.me,
	linux-nvme@...ts.infradead.org, linux-kernel@...r.kernel.org,
	tytso@....edu, nilay@...ux.ibm.com, martin.petersen@...cle.com,
	djwong@...nel.org, mcgrof@...radead.org
Subject: Re: [RFC PATCH] nvme: add an opt-in to use AWUPF

On Wed, Aug 20, 2025 at 03:02:20PM +0000, John Garry wrote:
> As described at [0], many parts of the atomic write specification are
> lacking.

I like your british understatement.

> +	list_for_each_entry(tmp, &subsys->ctrls, subsys_entry)
> +		nvme_queue_scan(tmp);

queueing a full rescan here seems expensive.  What about just keeping
the awupf value in our internal data structures and always use it
for the physical block size calculation, but only apply it to the
atomic limits based on a flag?