Re: Reachable killed buffers

[martin rudalics]

> If part of the list is shared, but the GC doesn't discover
> this until after that code runs, it'll remove dead
> buffers from that part.  From the user's point of
> view, whether dead buffers are removed would depend
> on the phase of the moon.

I'm too silly to understand.  If a buffer on one of these lists is dead,
it simply doesn't get marked from the list, so it's possible that it's
reclaimed with the next sweep.  In addition, the cons is removed from
the list which should not harm anyone.  If there's some other live
reference to the cons around, the cons gets marked and the dead buffer
as well, and mayby some other former member of the list.  So can some
kind soul please tell me what's wrong?

martin