linux-kernel - Re: [RFC] [PATCH] v4l2: use unsigned rather than enums in ioctl() structs

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, 11 Apr 2012 17:08:08 -0300
From:	Mauro Carvalho Chehab <mchehab@...hat.com>
To:	Rémi Denis-Courmont <remi@...lab.net>
CC:	mchehab@...radead.org, linux-media@...r.kernel.org,
	linux-kernel@...r.kernel.org, Alan Cox <alan@...rguk.ukuu.org.uk>
Subject: Re: [RFC] [PATCH] v4l2: use unsigned rather than enums in ioctl()
 structs

Em 11-04-2012 15:47, Rémi Denis-Courmont escreveu:
> 	Hello,
> 
> Le mercredi 11 avril 2012 20:02:00 Mauro Carvalho Chehab, vous avez écrit :
>> Using unsigned instead of enum is not a good idea, from API POV, as
>> unsigned has different sizes on 32 bits and 64 bits.
> 
> Fair enough. But then we can do that instead:
> typedef XXX __enum_t;
> where XXX is the unsigned integer with the right number of bits. Since Linux 
> does not use short enums, this ought to work fine.

I forgot to comment about that on the last e-mail. 

A solution close to the above one were already proposed:
	http://www.spinics.net/lists/vfl/msg25707.html

There were also another proposal there that might solve:
	http://www.spinics.net/lists/vfl/msg25702.html


Something like:

#if sizeof(enum) == 1
	typedef u8	__enum_t;
#elif sizeof(enum) == 2
	typedef u16	__enum_t;
#elif sizeof(enum) == 4
	typedef u32	__enum_t;
#elif sizeof(enum) == 8
	typedef u64	__enum_t;
#else
	typedef enum __enum_t;
#endif

Can actually work. Not sure if I really like adding a typedef, but maybe
this is the less dirty way to fix it.

We'll need to properly test the v4l2-compat32 code, as it will need 
to handle a different enum size on userspace. So, there, we'll likely
need to replace every enum with just "u32". Hmm... arm with 64 bits
(if/when added) may be an additional issue for the compat stuff.

Regards,
Mauro
--
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/