Re: [RFC][PATCH] sched: Cache aware load-balancing

From: Chen, Yu C
Date: Thu Mar 27 2025 - 04:10:29 EST


On 3/26/2025 5:42 PM, Peter Zijlstra wrote:
On Wed, Mar 26, 2025 at 05:15:24PM +0800, Chen, Yu C wrote:
Thanks for running the test. I think hackbenc/schbench would be the good
benchmarks to start with. I remember that you and Gautham mentioned that
schbench prefers to be aggregated in a single LLC in LPC2021 or 2022. I ran
a schbench test using mmtests on a Xeon server which has 4 NUMA nodes. Each
node has 80 cores (with SMT disabled). The numa=off option was appended to
the boot commandline, so there are 4 "LLCs" within each node.

We really should look at getting the SnC topology even without SnC being
in use.



I agree, unfortunately it seems that with SNC disabled there is no information exposed to the OS that can be used to divide the large LLC into smaller pieces.

thanks,
Chenyu

The sheer size of these LLCs is untenable.