Hi, I wrote a test that showed that thread scheduling had broken. In fact, commit 76eb3b4ea19ded63e3a9192ededc42e27c2dfcc7 broke it. But it should be fixed now; if you rerun your tests you should, I think, see the PDs and TDs on different CPUs.
Eddie Latency Buster wrote: > Thanks.. I shall read it now.. But running `mpstat' in parallel > confirms that somehow MT click I am running is not using all the > cores.. I know this used to work and work well.. so, digging further > .... _______________________________________________ click mailing list [email protected] https://amsterdam.lcs.mit.edu/mailman/listinfo/click
