![]() | ||||||||||
Messages in this thread |
On Thu, Nov 10, 2005 at 05:50:09PM +1100, Nick Piggin wrote: > But it isn't *really* constant time lookups? I mean you'll > always have the O(logn) lookup. Amortised I guess that > becomes insignificant? For sequential scan, I get 64*O(1) + 1*O(logn), that's pretty much gain. > Briefly: is there a reason why you couldn't use gang lookups > instead? (Sorry I haven't been able to read and understand your > actual readahead code). Because they have different semantics, one cannot replace the another. > Profile numbers would be great for the cached / non-cached cases. Ok, I'll do it, but at some time later. Currently I have several tasks that need immediate handling, and that will take about a week. See you then :) Regards, Wu - 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/ | |||||||||
| Last update: 2005-11-10 11:04 [from the cache] ©2003-2008 | ||||||||||