Messages in this thread | | | Subject | Re: [PATCH] Introduce nodemask_t ADT [0/7] | From | Matthew Dobson <> | Date | Fri, 19 Mar 2004 14:51:47 -0800 |
| |
On Thu, 2004-03-18 at 17:45, Paul Jackson wrote: > > It's hard to make those types of optimizations on generic masks. > > I would be assuming that by "generic" we meant arrays of unsigned longs > (or one unsigned long or something isomorphic to one or more unsigned > longs ...). > > And I'm assuming that we mean of a size that would allow for putting a > couple of them on a kernel stack ... not _too_ big. Probably NR_CPUS > rough upper limit on the size that was practical to use. > > I wouldn't want to get _too_ generic.
Well, if we're going to make a generic bitmap type, it shouldn't have size limitations, as almost any limit we set will be too small eventually... Supporting arbitrary length bitmaps doesn't mean we can't try to optimize for smaller masks, like less than a couple unsigned longs, as well as single unsigned long optimizations.
-Matt
- 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/
| |