linux-kernel - Re: [PATCH] drivers: video: Remove useless checks

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, 26 May 2011 07:43:05 -0600
From:	Tim Gardner <tim.gardner@...onical.com>
To:	Maxin B John <maxin.john@...il.com>
CC:	linux-kernel@...r.kernel.org, akpm@...ux-foundation.org,
	airlied@...hat.com, lethal@...ux-sh.org, james@...anarts.com,
	error27@...il.com, randy.dunlap@...cle.com
Subject: Re: [PATCH] drivers: video: Remove useless checks

On 05/26/2011 03:30 AM, Maxin B John wrote:
> Hi,
>
> On Wed, May 25, 2011 at 1:47 PM, Tim Gardner<tim.gardner@...onical.com>  wrote:
>> On 05/19/2011 12:27 PM, Maxin B John wrote:
>>>
>>> Comparing unsigned less than zero will never be true.
>>> Removing similar checks from 'fbmem.c' and 'fbcmap.c'.
>>
>> Looks right to me, though there are other places that suffer from the same
>> issue. See fb_set_cmap() and its use of 'int start' and cmap->start.
>
> Thank you very much for reviewing the patch. As per your suggestion, I
> have checked
> the scenario is fb_set_cmap() and the use of 'int start' and 'cmap->start'.
>
> IMHO, that scenario doesn't fall under the comparison of unsigned int
> <  0. That scenario
> looks similar to the below given code to me:
> /*---------------------------------*/
> #include<stdio.h>
>
> int
> main ()
> {
>    unsigned int u_int = -1;
>    int s_int = 0;
>    s_int = u_int;
>    if (s_int<  0)
>      printf ("s_int is less than 0\n");
>    return 0;
> }
> /*---------------------------------*/
>
> Please let me know your comments.
>
> Best Regards,
> Maxin
>

IMHO mixing signed and unsigned comparisons like this is just wrong. Its 
unnecessarily complicated and misleading, especially for a device 
driver. I'm a firm believer in the KISS philosophy for driver development.

Its likely that the reason the fb code got into this situation is 
because a type was changed from signed to unsigned whence long ago, and 
nobody has bothered to clean it up.

rtg
-- 
Tim Gardner tim.gardner@...onical.com
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/