Re: Being more careful about iospace accesses..

[viro@parcelfarce.linux.theplanet.co.uk]

On Wed, Sep 15, 2004 at 04:26:12PM -0700, Linus Torvalds wrote:
 
>    other bitwise type. You'd get a warnign about incompatible types. Makes 
>    sense, no?
>  - you can only do operations that are safe within that byte order. For 
>    example, it is safe to do a bitwise "&" on two __le16 values. Clearly 
>    the result is meaningful.

BTW, so far the most frequent class of endianness bugs had been along the
lines of
	foo->le16_field = cpu_to_le32(12);
and vice versa.  On big-endian it's a guaranteed FUBAR - think carefully about
the value that will end up there.

> Oh, btw, right now you only get the warnings from sparse if you use
> "-Wbitwise" on the command line. Without that, sparse will ignore the
> bitwise attribute.

We probably want __attribute__((opaque)) in addition to bitwise - e.g. for
the handles of all sorts passed in network filesystem protocols.  I'll look
into that when we get the endianness warnings somewhat under control.

For now I'm going to #define __opaque __bitwise and use it for stuff like
	typedef __u32 __opaque cifs_fid;
etc.