linux-kernel - Re: [PATCH v5 04/10] xfs: Reflink CoW-based atomic write support

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: <Z9GQDhRn3klzmDpo@infradead.org>
Date: Wed, 12 Mar 2025 06:45:50 -0700
From: Christoph Hellwig <hch@...radead.org>
To: John Garry <john.g.garry@...cle.com>
Cc: Christoph Hellwig <hch@...radead.org>, brauner@...nel.org,
	djwong@...nel.org, cem@...nel.org, linux-xfs@...r.kernel.org,
	linux-fsdevel@...r.kernel.org, linux-kernel@...r.kernel.org,
	ojaswin@...ux.ibm.com, ritesh.list@...il.com,
	martin.petersen@...cle.com
Subject: Re: [PATCH v5 04/10] xfs: Reflink CoW-based atomic write support

On Wed, Mar 12, 2025 at 09:13:45AM +0000, John Garry wrote:
> > > +	if (error || (!*shared && !atomic_sw))
> > 
> > And it's pnly used once.  Basically is is used to force COW, right?
> 
> Yes, we force it. Indeed, I think that is term you used a long time ago in
> your RFC for atomic file updates.
> 
> But that flag is being used to set XFS_BMAPI_EXTSZALIGN, so feels like a bit
> of a disconnect as why we would set XFS_BMAPI_EXTSZALIGN for "forced cow". I
> would need to spell that out.

Maybe use two flags for that even if they currently are set together?
Note that this would go away if we'd always align extsize hinted
allocations, which I suspect is a good idea (even if I'm not 100% sure
about it).