Messages in this thread | | | Date | Mon, 11 Jan 1999 12:40:37 +0100 (CET) | From | Andrea Arcangeli <> | Subject | Re: [patch] getname buffer moved to slab cache |
| |
On 11 Jan 1999, Zlatko Calusic wrote:
> be? Do you have a SMP machine. Does 'time' work that way on SMP > machines?
Yes.
> > Also, improvement is so small (if I got it right) that I wouldn't > bother. Benchmarking is a whole science. If I get such a small > improvements in my tests, I regularly dump the idea I had. :)
That's not the worst case. The point of the patch is that it's a waste of resources having to get/put every time a page from the free area struct risking having to defrag/compact some area of memory every time. Having some kind of cache for the so always needed getname buffer make tons of sense to me. Using the slab for implementing the cache is been the strighforward solution to me. If you all think that the slab implementation is not efficient for this kind of work (simply getting 4k memory areas) I can implement a dedicated cache, something like the pagetable-cache.
Andrea Arcangeli
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |