linux-kernel - Re: [PATCHv3 5/5] misc: gehc-achc: new driver

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]

Date:   Wed, 9 Jun 2021 13:42:08 +0100
From:   Mark Brown <broonie@...nel.org>
To:     Sebastian Reichel <sebastian.reichel@...labora.com>
Cc:     Rob Herring <robh+dt@...nel.org>, Arnd Bergmann <arnd@...db.de>,
        Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
        Shawn Guo <shawnguo@...nel.org>,
        Sascha Hauer <kernel@...gutronix.de>,
        Fabio Estevam <festevam@...il.com>, Ian Ray <ian.ray@...com>,
        linux-kernel@...r.kernel.org, linux-spi@...r.kernel.org,
        devicetree@...r.kernel.org, kernel@...labora.com
Subject: Re: [PATCHv3 5/5] misc: gehc-achc: new driver

On Fri, May 28, 2021 at 01:33:47PM +0200, Sebastian Reichel wrote:

>  drivers/misc/Kconfig                     |  15 +
>  drivers/misc/Makefile                    |   2 +
>  drivers/misc/gehc-achc.c                 | 160 ++++++++
>  drivers/misc/nxp-ezport.c                | 476 +++++++++++++++++++++++
>  drivers/spi/spidev.c                     |   7 +-
>  include/linux/platform_data/nxp-ezport.h |   9 +
>  include/linux/spi/spi.h                  |   5 +

This clearly seems to be something that should be split into separate
patches, even without getting as far as the code it's hard to see how
changes to spidev.c wouldn't be standalone changes that should be split
out from what looks like adding two new drivers.

> +static int gehc_achc_probe(struct spi_device *spi)
> +{

...

> +       ret = spidev_probe(spi);
> +       if (ret)
> +               return ret;

This isn't integrated at all with the flashing code, userspace might be
trying to do something with the device while flashing happens.  I would
expect at a very minimum that we'd be hot unplugging the spidev while
the flashing is in progress.  

> -static int spidev_probe(struct spi_device *spi)
> +int spidev_probe(struct spi_device *spi)
>  {
>  	struct spidev_data	*spidev;
>  	int			status;
> @@ -789,8 +788,9 @@ static int spidev_probe(struct spi_device *spi)
>  
>  	return status;
>  }
> +EXPORT_SYMBOL_GPL(spidev_probe);

This really looks like a very, very bad idea in general - attempting to
combine the unmediated userspace access that spidev offers with an in
kernel driver seems like it's not going to go well, as with the flashing
issue.  Really this looks like two marginally related devices that are
getting glued together which is abot the best case for this being
supportable and even then it'd need more work.  I'm not keen about
providing the interface at all as it's the sort of thing people are
going to get wrong a bunch.

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