> > Seems all these bug checks are fairly expensive. I can get 1%
> > back on system time for kernel compiles by changing BUG to
> > "do {} while (0)" to make them all compile away. Profiles aren't
> > very revealing though ... seems to be within experimental error ;-(
> >
> > I was pondering CONFIG_RUN_WILD_NAKED_AND_FREE
>
> The sort of folks who would worry about that very last 1% are the
> sort of people that would more than likely hit these BUGs as they're
> really stressing things.
>
> Losing a bunch of potential reports (and possibly doing bad things),
> in the name of a 1% performance boost doesn't sound too productive to me.
True - however I should have included some more info ... Andrew worked
out that some of the hottest ones lead to a null ptr dereference
immediately afterwards anyways, so they're actually pointless.
I wasn't seriously suggesting just removing all of them, was just a point
of interest for some things that would be worth looking at ;-)
I'd agree with you that an unreliable system is 100% slower than a working
one ;-)
M.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Tue Apr 15 2003 - 22:00:33 EST