Greetings! Yes, this is what I feared in general -- the hook is recursively calling the gc in this case, due to our placement of the call in GBC itself for the purposes of simplicity and clarity..
I've put in some simple code in a preliminary attempt to fix this, which is achieved at least for this example. Please test and let me know how it works for you. Take care, Robert Boyer <[EMAIL PROTECTED]> writes: > The following transcript causes 2.7.0 to go into an infinite loop on > > [GC for 33951 RELOCATABLE-BLOCKS pages..(T=15).GC finished] > > However, if you omit the setting of si::*after-gbc-hook*, there's no problem. > > Furthermore, if you instead omit the static array declarations, there is no > problem. > > Bob > > ------------------------------------------------------------------------------- > > (setq si::*notify-gbc* t) > > (setq *mht* (make-array 15000000 > :initial-element 0 > :static t) bar nil) > > (setq *cht* (make-array 20 > :element-type 'fixnum > :initial-element 0 > :static t)) > > (setq si::*after-gbc-hook* (function (lambda (x) nil))) > > > (setq foo (loop for i below 100000 collect > (progn (intern (format nil "A~a" i)))) bar nil) > > > ------------------------------------------------------------------------------- > > I got this 2.7.0 a day or two ago and built it on elgin with > > ./configure --quiet --silent --enable-ansi \ > --disable-statsysbfd --enable-locbfd \ > --enable-ihssize=2097110 \ > --enable-vssize=2097110 \ > --enable-maxpage=524288 \ > --enable-holediv=100 \ > --enable-static > > > -- Camm Maguire [EMAIL PROTECTED] ========================================================================== "The earth is but one country, and mankind its citizens." -- Baha'u'llah _______________________________________________ Gcl-devel mailing list [email protected] http://lists.gnu.org/mailman/listinfo/gcl-devel
