Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP Path: utzoo!watmath!clyde!rutgers!mit-eddie!husc6!seismo!rochester!ur-tut!ur-valhalla!badri From: badri@ur-valhalla.UUCP Newsgroups: comp.unix.wizards Subject: Not enough core: how to cure? Message-ID: <1010@ur-valhalla.UUCP> Date: Fri, 13-Feb-87 16:00:54 EST Article-I.D.: ur-valha.1010 Posted: Fri Feb 13 16:00:54 1987 Date-Received: Sat, 14-Feb-87 04:40:51 EST Organization: EE VLSI Graduate Research Facility, UR, Rochester NY 14627 Lines: 36 Keywords: BSD4.2, VAX 11/750, core I posted this question sometime back but did not get a single reply, so either something got screwed up or this newsgroup is read by a bunch of heartless twerps! We run BSD 4.2 on a VAX 11/750 with 2M core and 16M swap space. We have a some software written in fortran77, that has hardwired array sizes in it. We do not want to rewrite the code for dynamic memory management right now, though it will be done eventually. The problem is that when the array sizes exceed some value, we get a message "Not enough core". By playing around with array sizes, I have the following details from /bin/size: Wire1 (Works) text data bss dec hex 384000 66560 6071296 6521856 638400 Wire2 (Does not work) text data bss dec hex 384000 66560 6988076 7438636 71812c By putting debug messages in the kernel (auuugh!), I have narrowed the problem down to the following: ctod(ds) = 12906 exceeds maxdmap = 12256 in /sys/machine/vm_machdep.c Does someone know if there are kernel parameters that can be tweaked to get around this problem? I know that I am not directly exceeding the hardlimits MAXTSIZE, MAXDSIZE or MAXSSIZE. If there are any, I would like to know the consequent side effects of changing them. Thank you, Badri Lokanathan -- "We will fight for the right to be free {) ur-valhalla!badri@rochester.arpa We will build our own society //\\ {ames,caip,cmcl2,columbia,cornell, And we will sing, we will sing ///\\\ harvard,ll-xn,rutgers,seismo, We will sing our own song." -UB40 _||_ topaz}!rochester!ur-valhalla!badri