Re: [PATCH 2.6.6] bootmem.c cleanup

[H. Peter Anvin]

Followup to:  <20040514133353.236acf3a.akpm@xxxxxxxx>
By author:    Andrew Morton <akpm@xxxxxxxx>
In newsgroup: linux.dev.kernel
>
> Michael Buesch <mbuesch@xxxxxxxxxx> wrote:
> >
> >  -		if (!test_and_clear_bit(i, bdata->node_bootmem_map))
> >  -			BUG();
> >  +		BUG_ON(!test_and_clear_bit(i, bdata->node_bootmem_map));
> 
> Please don't put expressions whihc actually change state inside BUG_ON(). 
> Someone may decide to make BUG_ON() a no-op to save space.
> 
> I'm not aware of anyone actually trying that, but it's a good objective.
> 

If so they should make it:

#define	BUG_ON(x)	((void)(x))

.. which preserves side effects while generating no object code for a
side-effect-free expression.

The only reason for the cast to (void) at all is to keep gcc from
complaining about an ignored expression.

	-hpa
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/