Le 19/10/2020 à 16:16, Morten Rasmussen a écrit : > >>> If there is a provable benefit of having interconnect grouping >>> information, I think it would be better represented by a distance matrix >>> like we have for NUMA. >> There have been some discussions in various forums about how to >> describe the complexity of interconnects well enough to actually be >> useful. Those have mostly floundered on the immense complexity of >> designing such a description in a fashion any normal software would actually >> use. +cc Jerome who raised some of this in the kernel a while back. > I agree that representing interconnect details is hard. I had hoped that > a distance matrix would be better than nothing and more generic than > inserting extra group masks. > The distance matrix is indeed more precise, but would it scale to tens/hundreds of core? When ACPI HMAT latency/bandwidth was added, there were concerns that exposing the full matrix would be an issue for the kernel (that's why only local latency/bandwidth is exposed n sysfs). This was only for NUMA nodes/targets/initiators, you would have significantly more cores than that. Brice
