Any additional information you would like? Any other testing you want
performed? The Apple developer tools allow me to turn off cores on
the fly, if we think this might be related to SMP.
An interesting thought, but nothing has pointed that way yet.
You can see if smaller delays work. What I do is run CLOSE_DELAY up
high enough, and then find the smallest WRITE_PAD that works, and then
shrink CLOSE_DELAY. Basically, CLOSE_DELAY is about gpsd not reading
fast enough at the end, and WRITE_PAD is about overrunning the buffer
in the middle, except that it's not that cleanly separated.