linux-kernel - Re: [PATCH] floppy: hide invalid floppy disk types

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] [day] [month] [year] [list]

Date:   Mon, 09 Dec 2019 18:30:18 +0100
From:   philip@...pmail.net (Philip K.)
To:     Denis Efremov <efremov@...ux.com>
Cc:     moritzm.mueller@...teo.de, linux-kernel@...r.kernel.org,
        linux-block@...r.kernel.org, linux-kernel@...cs.fau.de
Subject: Re: [PATCH] floppy: hide invalid floppy disk types

> Hmm, I would say that driver blacklisting is a more proper solution in
> this case. I doubt there are people with this issue and real floppy drives
> in their setup. Altering the default driver's initialization scheme seems
> superfluous to me.

As long as major distributions like Ubuntu ship the floppy module, there
are enough people who could be affected by this peculiar behaviour.
While I agree that blacklisting the module would be more elegant, I
still think that a patch that goes in this direction could help more
people, especially those who don't want or cannot solve kernel-related
issues.

> This will force users (if there are ones) who depends on this behavior
> to rebuild the kernel. blacklisting doesn't require kernel rebuild.

Are there floppy disks of unknown types? Our patch is intentionally
conservative: We won't hide false negatives. If the motherboard reports
an non-existent disk

If you're ready to think about it, we could consider extending the patch
to un-register the device if it can recognise that it's (probably) not
real. In our case, for example, fdisk reported that fd0 had a size of
4k, something think is a strong indicator that something's not right.

Alternatively, we could look into what the comment

	/* FIXME: additional physical CMOS drive detection should go here */

would imply. This particular bug can only affect fd0 and fd1, so if we
spent some more time, we could find something.

-- 
	With kind regards,
	Philip K.

Download attachment "signature.asc" of type "application/pgp-signature" (488 bytes)