[net.bugs.4bsd] possible bug in 4.1BSD paging algorithms

haight (02/11/83)

an 11/750: A 1MB system running a fairly high DZ load and a very high
load of nested shell scripts would
  panic: dup page unlock...
about once a day in an obvious load-dependant way.  After a couple of
weeks we "fixed" it by adding 1MB (thereby eliminating paging).
Anyone else seen this one?