linux-kernel - Re: [PATCH v0 00/42] notifiers: Return an error when callback is already registered

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: <20211108141703.GB1666297@rowland.harvard.edu>
Date:   Mon, 8 Nov 2021 09:17:03 -0500
From:   Alan Stern <stern@...land.harvard.edu>
To:     Borislav Petkov <bp@...en8.de>
Cc:     LKML <linux-kernel@...r.kernel.org>, Arnd Bergmann <arnd@...db.de>,
        Ayush Sawal <ayush.sawal@...lsio.com>,
        Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
        Rohit Maheshwari <rohitm@...lsio.com>,
        Steven Rostedt <rostedt@...dmis.org>,
        Vinay Kumar Yadav <vinay.yadav@...lsio.com>,
        alsa-devel@...a-project.org, bcm-kernel-feedback-list@...adcom.com,
        intel-gfx@...ts.freedesktop.org,
        intel-gvt-dev@...ts.freedesktop.org, linux-alpha@...r.kernel.org,
        linux-arm-kernel@...ts.infradead.org, linux-clk@...r.kernel.org,
        linux-crypto@...r.kernel.org, linux-edac@...r.kernel.org,
        linux-fbdev@...r.kernel.org, linux-hyperv@...r.kernel.org,
        linux-iio@...r.kernel.org, linux-leds@...r.kernel.org,
        linux-mips@...r.kernel.org, linux-parisc@...r.kernel.org,
        linux-pm@...r.kernel.org, linuxppc-dev@...ts.ozlabs.org,
        linux-remoteproc@...r.kernel.org,
        linux-renesas-soc@...r.kernel.org, linux-s390@...r.kernel.org,
        linux-scsi@...r.kernel.org, linux-sh@...r.kernel.org,
        linux-staging@...ts.linux.dev, linux-tegra@...r.kernel.org,
        linux-um@...ts.infradead.org, linux-usb@...r.kernel.org,
        linux-xtensa@...ux-xtensa.org, netdev@...r.kernel.org,
        openipmi-developer@...ts.sourceforge.net, rcu@...r.kernel.org,
        sparclinux@...r.kernel.org, x86@...nel.org,
        xen-devel@...ts.xenproject.org
Subject: Re: [PATCH v0 00/42] notifiers: Return an error when callback is
 already registered

On Mon, Nov 08, 2021 at 11:19:24AM +0100, Borislav Petkov wrote:
> From: Borislav Petkov <bp@...e.de>
> 
> Hi all,
> 
> this is a huge patchset for something which is really trivial - it
> changes the notifier registration routines to return an error value
> if a notifier callback is already present on the respective list of
> callbacks. For more details scroll to the last patch.
> 
> Everything before it is converting the callers to check the return value
> of the registration routines and issue a warning, instead of the WARN()
> notifier_chain_register() does now.

What reason is there for moving the check into the callers?  It seems 
like pointless churn.  Why not add the error return code, change the 
WARN to pr_warn, and leave the callers as they are?  Wouldn't that end 
up having exactly the same effect?

For that matter, what sort of remedial action can a caller take if the 
return code is -EEXIST?  Is there any point in forcing callers to check 
the return code if they can't do anything about it?

> Before the last patch has been applied, though, that checking is a
> NOP which would make the application of those patches trivial - every
> maintainer can pick a patch at her/his discretion - only the last one
> enables the build warnings and that one will be queued only after the
> preceding patches have all been merged so that there are no build
> warnings.

Why should there be _any_ build warnings?  The real problem occurs when 
a notifier callback is added twice, not when a caller fails to check the 
return code.  Double-registration is not the sort of thing that can be 
detected at build time.

Alan Stern

> Due to the sheer volume of the patches, I have addressed the respective
> patch and the last one, which enables the warning, with addressees for
> each maintained area so as not to spam people unnecessarily.
> 
> If people prefer I carry some through tip, instead, I'll gladly do so -
> your call.
> 
> And, if you think the warning messages need to be more precise, feel
> free to adjust them before committing.
> 
> Thanks!