linux-kernel - Re: [PATCH] spi: remove return value check of debugfs_create

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: <dc807773-2028-4647-aa85-26f08a2810fa@sirena.org.uk>
Date:   Mon, 24 Apr 2023 14:17:27 +0100
From:   Mark Brown <broonie@...nel.org>
To:     Greg Kroah-Hartman <gregkh@...uxfoundation.org>
Cc:     Peter Enderborg <peter.enderborg@...y.com>,
        Yingsha Xu <ysxu@...t.edu.cn>,
        Serge Semin <fancer.lancer@...il.com>,
        linux-spi@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] spi: remove return value check of debugfs_create_dir()

On Mon, Apr 24, 2023 at 03:08:03PM +0200, Greg Kroah-Hartman wrote:
> On Mon, Apr 24, 2023 at 02:00:14PM +0100, Mark Brown wrote:

> > The ones I've seen have been name collisions caused by for example the
> > debugfs structure created being flatter than the device model structure,
> > though obviously something unanticipated could come up.

> Sure name collisions will happen, when people aren't precise about how
> they create their debugfs files (I just had to insist on this type of
> fixups for a USB patch last week.)  But, debugfs failures should never
> stop a driver from working properly, only the debugging functionalities.
> So there's no need to error out from debugfs errors as the only one
> affected is the kernel developer involved, not users.

Developers can be users, and multiple people can work with a single bit
of code.  If these things are being done as a result of looking at the
code and reasonably deciding that actually the error handling wasn't
needed that'd be fine but when it's just silencing a static checker with
no obvious consideration and it's super obviously the right call just
from the diff that's a different thing.

Download attachment "signature.asc" of type "application/pgp-signature" (489 bytes)