mycroft@gnu.ai.mit.edu writes:

> There is a change in the kernel which should fix that.  It was committed
> on January 7.  Are you sure your copy of swap_pager.c is up to date?

My swap_pager.c reads:

 *      $Id: swap_pager.c,v 1.14 1994/01/07 20:33:14 mycroft Exp $

So I think this should be the right version. However, I still need to
do this:

*** swap_pager.c.old    Thu Jan 13 00:42:38 1994
--- swap_pager.c        Thu Jan 13 02:03:35 1994
***************
*** 919,925 ****
         */
  #ifdef notyet /* XXX */
        if (cnt.v_free_count < cnt.v_free_min)
! #else
        /* XXX shouldn't this be the same test as in vm_page_alloc()? */
        if (vm_page_free_count < vm_page_free_min)
  #endif
--- 919,925 ----
         */
  #ifdef notyet /* XXX */
        if (cnt.v_free_count < cnt.v_free_min)
! /* #else */
        /* XXX shouldn't this be the same test as in vm_page_alloc()? */
        if (vm_page_free_count < vm_page_free_min)
  #endif

...to make it work (i.e. take out the 'if' for the thread_wakeup, so it
is always done). This is probably the wrong way to fix it, but it
is about the same as what the earlier mailed patch did, I think.

- Frank

------------------------------------------------------------------------------