linux-kernel - Re: [PATCH v3 1/7] mfd: add pruss mfd 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:	Thu, 24 Mar 2011 14:24:51 +0100
From:	Marc Kleine-Budde <mkl@...gutronix.de>
To:	Subhasish Ghosh <subhasish@...tralsolutions.com>
CC:	davinci-linux-open-source@...ux.davincidsp.com,
	sachi@...tralsolutions.com, Samuel Ortiz <sameo@...ux.intel.com>,
	nsekhar@...com, open list <linux-kernel@...r.kernel.org>,
	m-watkins@...com, linux-arm-kernel@...ts.infradead.org,
	Stalin Srinivasan <stalin.s@...tralsolutions.com>
Subject: Re: [PATCH v3 1/7] mfd: add pruss mfd driver.

On 03/24/2011 02:16 PM, Subhasish Ghosh wrote:
> Hello,
> 
>> A device may be de-initialized and another loaded or two different
>> devices
>> can be run simultaneously on the two cores.
>> It's also possible, as in our case, to implement a single device on both
>> the PRU's resulting in improved load sharing.
> 
>> Make you driver compile with sparse "make C=1", you cast way too much
>> when accessing io mem. Use void __iomem * instead of u32 *.
> 
> 
> With make C=1 I am receiving some warnings such as:
> 
> warning: cast removes address space of expression
> drivers/mfd/da8xx_pru.c:61:17: warning: incorrect type in argument 1
> (different base types)
> drivers/mfd/da8xx_pru.c:61:17:    expected void const volatile [noderef]
> <asn:2>*<noident>
> drivers/mfd/da8xx_pru.c:61:17:    got int
> drivers/mfd/da8xx_pru.c:66:28: warning: incorrect type in argument 1
> (different address spaces)

(Re)read the warning carefully:

You have to fix drivers/mfd/da8xx_pru.c, the first argument should be
__void iomem *, not an int. Fix the type of the first argument. Don't
use any casts at all, follow all warnings, eventually you will have void
__iomem * (or struct * instead of void *) and no warnings.

> I can remove all of these by casting to (__force void  __iomem *) but is
> this correct.

No evil casts, please :)

Marc

-- 
Pengutronix e.K.                  | Marc Kleine-Budde           |
Industrial Linux Solutions        | Phone: +49-231-2826-924     |
Vertretung West/Dortmund          | Fax:   +49-5121-206917-5555 |
Amtsgericht Hildesheim, HRA 2686  | http://www.pengutronix.de   |


Download attachment "signature.asc" of type "application/pgp-signature" (263 bytes)