arii@tut.fi (Ari Ikonen) (01/05/88)
I faintly remember someone mentioned something similar some time ago in this news group. We are running magic (Berkeley VLSI Tools) version 4.1 on Suns (3/50,3/160). Lately we have started to suspect magic has some kind of memory deallocation problem. During the editing session the size of magic monotonically grows until magic dies when it can't gain more memory (if the session is long enough). Magic grows even though the layout would not suggest such a growth. Has anyone experienced the same or have I just imagined the previous discussions about the matter? Does anyone have any ideas on what's going on or have bug fixes if this really is a bug? We haven't yet done full-proof tests but maybe we are forced to it, if we don't find the reason for magic's growth some other way. -- Ari Ikonen # Tampere University of Technology arii@tut.fi (Internet) # Microelectronics Laboratory arii@tut.UUCP # PO box 527, SF-33101 Tampere, Finland ...mcvax!tut!arii # tel: +358 31 162619 fax: 162766