SMP/HTT problems with 4.10-BETA

[Pete French]

> It is quite possible that it is my own fault if machdep.hlt_logical_cpus
> should only be set during boot as a loader tunable but that isn't the
> way I read the UPDATING entry.

I set mine during boot and get the same results. I am glad that someone else
can reproduce this anyway! Not sure hhow toset about tracking down useful
debugging data though - a kernel panic is one thing, random processes
existing with bad values is somewhat harder to trace.

-pcf.