Messages in this thread | | | From | Daniel Phillips <> | Subject | Re: [RFC][PATCH] Faster generic_fls | Date | Wed, 30 Apr 2003 16:07:59 +0200 |
| |
On Wed 30 Apr 03 15:17, Falk Hueffner wrote: > Daniel Phillips <dphillips@sistina.com> writes: > > It's somewhat annoying that __builtin_clz leaves the all-ones case > > dangling instead of returning -1. > > I assume you mean all-zero...
:-)
> Well, the semantics of the corresponding > instructions simply differ too much. clz(0) is 31 on some, 32 on some > others, and undefined on some more architectures. So nailing it down > would incur some performance penalty.
It would only penalize stupidly broken architectures; it would help the ones that got it right. It's hard to imagine how a sane person could think the result of clz((u32)0) should be other than 32.
Sigh.
Regards,
Daniel
- 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/
| |