Re: [2.6 patch][1/3] ipc/ BUG -> BUG_ON conversions

[Adrian Bunk]

On Sat, Aug 28, 2004 at 12:58:16PM -0700, Andrew Morton wrote:
> Adrian Bunk <bunk@xxxxxxxxx> wrote:
> >
> >  > Anything you put in BUG_ON() must *NOT* have side effects.
> >  >...
> > 
> >  I'd have said exactly the same some time ago, but I was convinced by 
> >  Arjan that if done correctly, a BUG_ON() with side effects is possible  
> >  with no extra cost even if you want to make BUG configurably do nothing.
> 
> Nevertheless, I think I'd prefer that we not move code which has
> side-effects into BUG_ONs.  For some reason it seems neater that way.
> 
> Plus one would like to be able to do
> 
> 	BUG_ON(strlen(str) > 22);
> 
> and have strlen() not be evaluated if BUG_ON is disabled.
> 
> A minor distinction, but one which it would be nice to preserve.

OTOH, only very few people use the disabled BUG_ON, and a statement 
with a side effect might stay there unnoticed for some time.

If it's a
  #define BUG_ON(x)
and in one place something with a side effect slipped into the BUG_ON, 
you have a classical example for a heisenbug...

cu
Adrian

-- 

       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed

-
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/