On Sat, Mar 15, 2008 at 06:22:41PM -0800, ron minnich wrote:
On Sat, Mar 15, 2008 at 6:11 PM, joe@smittys.pointclark.net wrote:
coreboot is an alternative. I was asking if anyone knew about interest for a stronger bond between coreboot and Linux.
Ah, gotcha...
Actually, it requires only that somebody start dropping patches in. But what, really, do we need?
As you know my grand plan is to get rid of all things legacy. I do think there is, or will be, a need.
On the way to that 2.6.24 I took a detour into Ulrich Drepper's excellent What Every Programmer Should Know About Memory.
On the topic of HT hypercube information in sysfs:
--8<-- http://people.redhat.com/drepper/cpumemory.pdf page 45 Each processor constitutes its own node as can be seen by the bits set in the value in cpumap file in the node* directories. The distance files in those directories con- tains a set of values, one for each node, which represent a cost of memory accesses at the respective nodes. In this example all local memory accesses have the cost 10, all remote access to any other node has the cost 20. [26] .. [26] This is, by the way, incorrect. The ACPI information is appar- ently wrong since, although the processors used have three coherent HyperTransport links, at least one processor must be connected to a Southbridge. At least one pair of nodes must therefore have a larger distance. -->8--
For people to do a really good job with libNUMA it seems benefitial to replace the wrong information with something that is right.
Any advantage coreboot can offer is good I think.
//Peter