drs-ano@duvan.nada.kth.se (Arne Nordmark) (12/14/87)
Help! I've just "upgraded" to an A2000. Sadly, I got one of those old ones with the CPU slot occupied by an expansion memory board. The board was only half full (512K) so I just couldn't wait filling it up to one full MEG. So I did, and Workbench proudly announced 1.4 MEG free memory. Nice! Well here's the problem: The ASDG RAM disk just don't like those extra bytes. When I reboot the disk vanishes. But strangely this only happens when I've filled the disk with more than half a MEG or so. "Aha! there must be an error in those memory chips that I added, the ASDG device detects the error and make a new disk" I thought. I found out that the 1 MEG expansion memory occupies addresses C00000-D00000. (That block are "reserved" for future use according to my manuals, but those manuals are for 1.1) I wrote a program that checked the block. And I didn't find a single error! The disk worked perfectly (almost) with a one MEG Starboard occupying addresses 200000-300000. So I just don't know what's wrong with C00000! I really like the ASDG, but I also like the extra memory. What shall I do?
ugmiker@sunybcs.uucp (Michael Reilly) (12/17/87)
In article <244@draken.nada.kth.se> drs-ano@duvan.nada.kth.se (Arne Nordmark) writes: >Help! > >The disk worked perfectly (almost) with a one MEG Starboard occupying >addresses 200000-300000. So I just don't know what's wrong with C00000! > a while ago Perry made it clear that the ASDG RRD would not work with memory in the c$00000 position. So boards that use that memory position , like the insider, or your board in the 2000 will have problems with it. The fact that it only "breaks" when you fill it up past one meg, must have to do with the way the RRD handles the c$XXXXX (where XXXXX is some memory offset from the C$00000 memory spot) memory, not really the c$00000 memory. >I really like the ASDG, but I also like the extra memory. >What shall I do? I think we all like ASDG, and the fact that the c$00000 mem. wont work with the RRD, means I will be (sooner or later) buying a rack and some memory from them.... mike
dedina@silver.bacs.indiana.edu (12/18/87)
drs-ano@duvan writes: > Well here's the problem: The ASDG RAM disk just don't like those extra bytes. > When I reboot the disk vanishes. > But strangely this only happens when I've filled the disk with more than > half a MEG or so. "Aha! there must be an error in those memory chips that I have a similar problem with my A500 with internal memory expansion. When vd0: gets much over half full, it vanishes on reboot. dedina@silver.bacs.indiana.edu
bryce@hoser.berkeley.edu (Bryce Nesbitt) (12/19/87)
In article <7304@sunybcs.UUCP> ugmiker@joey.UUCP (Michael Reilly) writes: > >a while ago Perry made it clear that the ASDG RRD would not work with memory >in the c$00000 position. Hogwash! I've been using and recovering a RRD from 1.5 megs of $C00000 memory for half a year or so. Works fine. Most of that time was *before* I fixed the Kickstart bugs dealing with $C00000 memory. If you have $C00000 memory and don't have the fix, the RRD will loose out on certain types of crashes, but not others. It's not RRDs fault, it is a problem in the auto-sizing Kickstart does. It writes to ram to do the sizing rather than using a trick to eliminate the need for the destructive write. >in the c$00000 position. ^^^^^^^ The proper way to express that is $C00000 or perhaphs C00000H or even 0xC00000, but not C$00000. |\ /| . Ack! (NAK, SOH, EOT) {o O} . bryce@hoser.berkeley.EDU -or- ucbvax!hoser!bryce (or try "cogsci") (") U "Your theory is crazy... but not crazy enought to be true." -Niels Bohr