Re: [BUG] Bad #define, nonportable C, missing {}

[ebiederm@xmission.com (Eric W. Biederman)]

"Richard B. Johnson" <root@chaos.analogic.com> writes:

> On Wed, 21 Nov 2001, Jan Hudec wrote:
> 
> > > >     *a++ = byte_rev[*a]
> > > It looks perferctly okay to me. Anyway, whenever would you listen to a
> > > C++ book talking about good C coding :p
> > 
> 
> It's simple. If any object is modified twice without an intervening
> sequence point, the results are undefined. The sequence-point in
> 
> 	*a++ = byte_rev[*a];
> 
> ... is the ';'.
> 
> So, we look at 'a' and see if it's modified twice. It isn't. It
> gets modified once with '++'. Now we look at the object to which
> 'a' points. Is it modified twice? No, it's read once in [*a], and
> written once in "*a++ =".
> 
> So, it's perfectly good code with a well defined behavior as far as
> 'C' is concerned. 

Nope.  In particular it isn't defined if [*a] is evaluated before
or after a++ is evaluated.  

> I think it is ugly, however, the writer probably
> thought it was beautiful. If somebody went around and fixed all
> the ugly code, it would still be ugly in someone else's eyes.

True, but not significant.  You can find a set of people whose opinion
matters (say core kernel maintainers) and not have that looks ugly in
their eyes.   In any case broken code is broken.

Eric