[xemacs-hg @ 2002-06-28 14:24:07 by michaels] 2002-06-27 Mike Sperber <mike@xemacs.org> * data.c (prune_weak_boxes): Rewrite for better readability. 2002-06-23 Martin Köbele <martin@mkoebele.de> Jens Müller <jmueller@informatik.uni-tuebingen.de> Mike Sperber <mike@xemacs.org> * lrecord.h (lrecord_type): add lrecord_type_ephemeron to lrecord_type enumeration. * lisp.h (XEPHEMERON): (XEPHEMERON_REF): (XEPHEMERON_NEXT): (XEPHEMERON_FINALIZER): (XSET_EPHEMERON_NEXT): (XSET_EPHEMERON_VALUE): (XSET_EPHEMERON_KEY): (wrap_ephemeron): (EPHEMERONP): (CHECK_EPHEMERON): (CONCHECK_EPHEMERON): (struct ephemeron): Add representation of ephemerons. * alloc.c (garbage_collect_1): (finish_marking_ephemerons): (prune_ephemerons): Call. * data.c: (finish_marking_ephemerons): (prune_ephemerons): (mark_ephemeron): (print_ephemeron): (ephemeron_equal) (ephemeron_hash):: (make_ephemeron): (Fmake_ephemeron): (Fephemeronp): (Fephemeron_ref): (syms_of_data): (vars_of_data): Add implementation of ephemerons

/* Synched up with: FSF 19.31. */

#include "sunos4-1.h"

#if 0
/* XEmacs: FSF 19.31 removes this.  Let's just comment it out. */
/* TERMIOS is broken under SunOS??

   Someone says: This causes failure in process_send_signal (tcgetattr
   loses) and may also cause hanging at Emacs startup when parent is
   not a job control shell.  */
/* murray@chemical-eng.edinburgh.ac.uk says this works, and avoids
   the problem of spurious ^M in subprocess output.  */
#undef HAVE_TERMIOS
#endif

#if 0
/* XEmacs: FSF 19.31 mistakenly reenables this. */
/* jik@gza.com says this works now.  */
/* The bug that corrupts GNU malloc's memory pool is fixed in SunOS 4.1.3. */

#undef SYSTEM_MALLOC
#endif /* 0 */

/* barrie@calvin.demon.co.uk says memmove is missing.  */
#ifndef SYSTEM_MALLOC
#define MEMMOVE_MISSING
#endif

/* A reliable source says this is broken through SunOS 4.1.3 */
/* but not SunOS 4.1.4 */
#ifdef BROKEN_SIGIO
#undef BROKEN_SIGIO
#endif