Messages in this thread | | | Date | Mon, 30 Jan 2006 11:22:09 +0100 | From | Eric Dumazet <> | Subject | Re: Questions about alloc_large_system_hash() and TLB entries |
| |
David S. Miller a écrit : > From: Eric Dumazet <dada1@cosmosbay.com> > Date: Mon, 30 Jan 2006 10:03:40 +0100 > >> What would be the needed changes in the code to get both : >> >> - Allocate ram equally from all the nodes of the machine >> >> - Use large pages (2MB) to lower TLB stress > > These two desires are mutually exclusive, I think. > > If you want an 8MB hash table, for example, with 2MB mappings > you could use memory from a maximum of 4 nodes since the > 2MB chunks have to be physically 2MB aligned and 2MB contiguous.
Yes of course, but as those hash tables are very big (their size is bigger than 2MB * number_of_nodes if you have at least 4GB per node), this could be done ?
Eric
- 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/
| |