linux-kernel - Re: [PATCH 00/26] cred: rework {override,revert}

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: <20241125-mitverantwortlich-bauphase-34b1b97190bc@brauner>
Date: Mon, 25 Nov 2024 12:46:00 +0100
From: Christian Brauner <brauner@...nel.org>
To: Linus Torvalds <torvalds@...ux-foundation.org>
Cc: Amir Goldstein <amir73il@...il.com>, 
	Miklos Szeredi <miklos@...redi.hu>, linux-kernel@...r.kernel.org, linux-fsdevel@...r.kernel.org
Subject: Re: [PATCH 00/26] cred: rework {override,revert}_creds()

On Sun, Nov 24, 2024 at 10:00:24AM -0800, Linus Torvalds wrote:
> On Sun, 24 Nov 2024 at 05:44, Christian Brauner <brauner@...nel.org> wrote:
> >
> > This series does all that. Afaict, most callers can be directly
> > converted over and can avoid the extra reference count completely.
> >
> > Lightly tested.
> 
> Thanks, this looks good to me. I only had two reactions:
> 
>  (a) I was surprised that using get_new_cred() apparently "just worked".

There's only one case and that's io_uring where we can just cast because
we only need it temporarily during the conversion part of the patch
series. Later we don't take any reference count anymore in io_uring.

>  (b) a (slight) reaction was to wish for a short "why" on the
> pointless reference bumps

Yeah, sorry for some of the patches I just quickly jotted down the same
line in the commit message. I updated all those commit messages with
actual explanations why that's safe.

> But not a big deal. Even in this form, I think this is a clear and
> good improvement.

Cool.