Messages in this thread | | | Date | Tue, 15 Mar 2005 08:51:04 -0800 (PST) | From | Christoph Lameter <> | Subject | Re: [PATCH] Per cpu irq stat |
| |
On Tue, 15 Mar 2005, Christoph Hellwig wrote:
> On Mon, Mar 14, 2005 at 10:32:34PM -0800, Christoph Lameter wrote: > > The definition of the irq_stat as an array means that the individual > > elements of the irq_stat array are located on one NUMA node requiring > > internode traffic to access irq_stat from other nodes. This patch makes > > irq_stat a per_cpu variable which allows most accesses to be local. > > There's architectures accessing it from assemly.
The patch is i386 specific!
> But furthermore there's absolutely not point for the irq_stat structure > at all anymore now that we have the per_cpu infrastructure. so kill it > completely and let every architecture just provide a local_softirq_pending > macro.
That would require reworking a lot of architectures.
- 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/
| |