Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP Posting-Version: version B 2.10.1 6/24/83; site watcgl.UUCP Path: utzoo!watmath!watcgl!dmmartindale From: dmmartindale@watcgl.UUCP (Dave Martindale) Newsgroups: net.unix-wizards Subject: Re: Lost Memory (4.*BSD) Message-ID: <2128@watcgl.UUCP> Date: Thu, 16-Feb-84 00:27:50 EST Article-I.D.: watcgl.2128 Posted: Thu Feb 16 00:27:50 1984 Date-Received: Thu, 16-Feb-84 04:38:27 EST References: <321@flairvax.UUCP> Organization: U of Waterloo, Ontario Lines: 24 The "mcr1 at tr2" is printed out by the autoconfigure code which goes probing through the NEXUSes. It does nothing about looking at or adjusting the starting address of that memory - it just configures it so error reporting gets done. The actual determination of how much memory is present on the machine is done by simply scanning up through memory a page at a time until the CPU takes a machine check from referencing non-existent memory. If the second block isn't configured to be contiguous with the first, UNIX won't find it. Someone may have botched the starting-address jumpering, or the address may have been clobbered since the last powerup. Try powering the memory completely down (using the breaker on its power supply) and reboot. If UNIX doesn't find it then, the starting address is likely bad. VMS is probably smarter about changing the starting address of a second controller to make it contiguous with the first. If you have two controllers with equal amounts of memory on each, you will want to interleave them. Add the lines; DEPOSIT 20002000 101 ! ENABLE INTERLEAVE FOR TR #1 MEMORY CONTROLLER DEPOSIT 20002004 4000 ! FORCE 0 STARTING ADDRESS DEPOSIT 20004000 101 ! ENABLE INTERLEAVE FOR TR #2 MEMORY CONTROLLER DEPOSIT 20004004 4000 ! FORCE 0 STARTING ADDRESS to your xxSBOO.CMD and xxMBOO.CMD command files just before the "load boot" line. (This should go in RESTAR.CMD too, I think.)