CompuServe Thread

#AIBB & SYSINFO on A3000?

33 messages in this thread
#81969From: Dave ArreDec 20, 1992 12:23 AM
Steve: I just downloaded AIBB here on the forum, and found that on my A3000, the machine freezes once the message "Testing system.. Please wait" or something like that, comes up. I tried removing all my extra tasks (screen blanker, commodities, etc) to no avail. I was wondering if this is normal behavior. I have: An Amiga 3000 25 mHz 50 Meg with 6 Megs of Ram. No expansion boards. Also, SYSINFO freezes on me when I do the COMPUTE SPEED option, once it gets to the MFLOPS test. Any ideas? Is there something I'm doing wrong, or is something defective in my system? The computer seems to run fine otherwise. Thanks! -Via AutoPilot — Dave
#81970From: Steve AhlstromDec 20, 1992 12:45 AM
Dave, I don't know. I've never run AIBB but have run various versions of SYSINFO with no problems. It might be due to an OS version difference (if you're not running 2.04 or 2.1), it might be hardware problem on your end, or it might be a conflict with something in your wbstartup drawer. If your machine is stable otherwise, I wouldn't be too concerned about it.
#82014From: Dave ArreDec 20, 1992 2:39 PM
Steve: Yes, my machine is stable otherwise. I had thought that since SYSINFO hangs up on the MFLOPS calculations, maybe something was amiss with my FPU, but I have DISTANT SUNS 4.1 in the OS 2.0 – FPU only version, and it works fine, plus I have FRACBLANK_881 that works fine. A possibility is that perhaps I have an "early" or beta version of 2.04… I bought my 3000 used from someone who may have had access to beta version of the OS. How can I tell if I have the correct "release" version? I just brought up a Shell and typed in the VERSION command, and this is what it returned: 1.System2.0:> version Kickstart version 37.175. Workbench version 37.67 1.System2.0:> Does that indicate anything to you? Thanks for your help! — AutoPilot Engaged! — Dave
#82018From: Steve AhlstromDec 20, 1992 2:48 PM
Dave, Those are the correct release version 2.04 numbers.
#82099From: BILL LEACHDec 21, 1992 11:24 AM
As mentioned, your OS release is correct. I suspect that Dean may have hit upon the problem. You did read the docs, yes? It is not exactly considered good form to notify the user of a problem by locking up the machine so LaMonte tries to supply the user with something a bit more friendly. However, while his program is tested with all versions of the OS it is nearly impossible to test a program with all version of the various support libraries that may be around. You could have a corrupt library or even one that is not backwards compatible (that should never be allowed to happen but..). -bill
#82120From: Dave ArreDec 21, 1992 3:50 PM
Bill: I did not have Req.Library, but that did not make a difference after I downloaded it from AmigaTech. I admit I did not read every word of the Docs, but I did read the section on system requirements and usage carefully. Here is a snap from the AIBB docs: AIBB uses a shared library for its file requester functions. For this reason AIBB requires that one of the following shared libraries be within the system LIBS: directory: 1. kd_freq.library ( library version 3.0 or greater ) 2. req.library ( library versiopn 2.0 or greater ) 3. asl.library ( AmigaOS 2.0 systems only ) 4. reqtools.library AIBB will search for these libraries in the order given, stopping when it is either able to open one, or none are found. I have all those now, but it still freezes after displaying "Evaluating System: Please wait." The mouse freezes, but the power light DOES NOT go dim. I have let the thing sit for about 30 minutes. No dice. This is increasingly puzzling. Thanks again for your help…… Merry Christmas! — Dave
#81978From: Erik FlomDec 20, 1992 1:28 AM
AIBB worked fine on my A3000 (before and after adding an 040 accelerator). I had no problems running it under 2.04 and 2.1. Sounds weird. ? Erik Flom
#82015From: Dave ArreDec 20, 1992 2:39 PM
Erik: Yes, I know it is weird. I would suspect the FPU, except I have DISTANT SUNS 4.1 in the OS2.0 FPU only version, and FracBlank_881, both run smoothly without a hitch. I'm puzzled. — Dave
#81992From: BILL LEACHDec 20, 1992 9:34 AM
Dave; LaMonte would be very interested in hearing about your problem. AIBB is a professional level benchmarking program. Many developers have assisted in process required to make it accurate. AIBB gets tested on some of the strangest Amiga systems around. It works with every release of the OS (to the point stated in the docs). It should run without problem on any amiga. If it does not then either you have found a bug, there is something non-standard (in a wicked way) in your system or there is actually something wrong with your system. Would please post all of the relevent information on your machine, do a version command (for workbench and kickstart), do a version on each of the "mathxxxxx.libraries". Also, what HardDrive do you have, how is it partitioned and what filesystem ("version filesystem" or "version wb_2.x:" should work)? -bill
#82016From: Dave ArreDec 20, 1992 2:40 PM
Bill: Yes, I'm very puzzled about this… I'll list all the info on my machine, and perhaps you can identify somthing out of the ordinary. I have to stress that my machine works fine (even with programs specifically written for FPU use) other than SYSINFO and AIBB. Here goes: A3000 25mHz 52 Meg Quantum HD. No Boards. Daisy Wheel printer connected to the Parallel Port; SupraModem 2400 to the serial port. A pair of C= A10's for sound, and a C= 1950 multiscan monitor. One floppy drive. HD has 2 partitions: 6 Meg System2.0: Partition, and 44 Meg Work: partition. I have had no problems with the HD. Here is a listing of Shell inquiries as you requested (via SNAP): 1.System2.0:> cpu System: 68030 68882 FastROM (INST: Cache NoBurst) (DATA: Cache NoBurst) 1.System2.0:> version Kickstart version 37.175. Workbench version 37.67 1.System2.0:> version mathieeedoubbas.library mathieeedoubbas.library version 37.1 1.System2.0:> version mathieeedoubtrans.library mathieeedoubtrans.library version 37.1 1.System2.0:> version mathieeesingtrans.library mathieeesingtrans.library version 37.1 1.System2.0:> version mathtrans.library mathtrans.library version 37.1 1.System2.0:> version filesystem Error: filesystem not found 1.System2.0:> version wb_2.x: fs 37.26 1.System2.0:> ? Does this listing tell you anything? Thanks for your help! AutoPilot Engaged! <continued>
#82091From: BILL LEACHDec 21, 1992 10:48 AM
Thanks Dave; I get back to you after I hear from LaMonte. I don't know if he is doing anything over the holidays or not so don't have any idea how long it will be. Speaking of which…. Merry Christmas! -bill
#82119From: Dave ArreDec 21, 1992 3:49 PM
Thanks for looking into this for me… I really appreciate it. Merry Christmas to you and yours! — Dave
#82130From: Dave ArreDec 21, 1992 5:18 PM
Bill: I just took the cover off of my A3000, and I may have some more news for you. The system has 2Meg Chip ram, and 4 Meg Fast Ram. Fast Ram is 1 x 4 Page mode. Also: I do not have Kickstart in ROM (ROM 0 and ROM 1 are missing). I have known this since I got the machine; I have been waiting for the updates to the OS to happen (2.1, and now 3.0). I noticed, however, that there are no sockets for the ROM chips; they are surface mounted. This means <gulp> that they are not user-installable, and have to be done by a service center, correct? Thanks for your help…. talk to you soon! Merry Christmas! — Dave
#82171From: Erik FlomDec 21, 1992 10:33 PM
I just got 2.0 ROM's with the GVP 040-A3000 accelerator(a FREE bonus!). (But, it fried my Toaster, so buy a bigger Power Supply if you go for this config!) There were 2 empty sets of holes on the PC board, and a set of socketed ROMs in front of them – I just had to pull out the old Bootstrap ROMs and plug the new ones into those sockets. No soldering required! (I WAS using the setup where KICKSTART was loaded from disk, but the new ROMs don't support it. [Gee, I guess I can't use that old 1.3 Kickstart image anymore. Too bad. :^) ]) BTW, I had run v5.0 on AIBB on both the old config, and the new config with the new ROMs – no problem. (I had the same version of 2.x Kickstart as you have in your disk image. [37.175]) Hope you get AIBB to work. Erik Flom P.S. Merry XMas, and a Happy Gnu Year! ,<|:^) (It's a hat, really!)-^
#82212From: Dave ArreDec 22, 1992 3:50 AM
Hmmm… The version of AIBB I've been using is 4.1. Maybe I need to find a new version somewhere….. I got it from either here or in the AmigaTech forum. Interesting about the ROMs. The empty places I saw (no sockets) were under the equipment platform, on the left and almost all the way to the back past the mouse & joystick plugs and in front of the fan. The empty spaces were labeled, on the circuit board, as ROM0 and ROM1, and they looked like huge, long rectangular chips. Perhaps this is not the place where the OS ROM is installed? Merry Christmas! I like the hat! ,<|:^) — Dave
#82259From: Erik FlomDec 22, 1992 1:28 PM
I just got v5 of AIBB from CIS recently, if I remember correctly. (Has worked just fine for me.) Yes, those 2 empty spaces are in the right area, but there should have been 2 socketed chips in front of them (towards the RAM chips). Those are the ones you would swap out (U180 & U181 on my MB). (You do have an A3000, and not a tower 3000? They only have 2 sockets.) Noel, Noel. ,<|:^) Erik Flom
#82342From: Dave ArreDec 23, 1992 12:44 AM
Erik, yes I have the A3000. I must have missed the two socketed ROMS when I was peaking through the side. At least that's good news! I wouldn't want to try to install a 40-pin surface mount chip (never mind 2!) if my life depended on it. I will do a catalog update for the Amiga Forums to see if AIBB 5.0 has been posted more recently than I think… then if I find it, I'll let you know what happens. Thanks & Buon Natale! — via AutoPilot. — Dave
#82349From: Erik FlomDec 23, 1992 1:09 AM
The socketed ROMs ARE easy to miss. Especially if you haven't pulled off the equipment bay. (I HAVE removed 40 pin IC's & all I can say is that it requires much patience [& a very low-power soldering iron].) Good Luck w/ AIBB v5.0! And a big Ho-Ho-Ho to you too! Erik Flom ,<|:^)
#82518From: Dave ArreDec 24, 1992 10:52 AM
Erik: AIBB 5.0 is not here on CIS. Bill just uploaded 5.5, though, and that worked fine on my system! It asked me on startup what CPU I have (68030 or 68EC030), but after that ran beautifully. My machine held its own against the A3000-25 (but darn that A4000 just SPOILS everything! g>) Thanks for all your help! Feliz Navidad! — Dave
#82386From: BILL LEACHDec 23, 1992 8:58 AM
Dave; I just pushed AIBB_55.lha up this morning. Again, I'm reminded of a feature that AP badly needs and that is the ability to specify the name that the file should have on the CIS. I suspect that the file will be named AIBB_5.lha which will no doubt be very misleading. Of course the description says that it is 5.5 and not 5.0 version. Happy Holidays -bill
#82517From: Dave ArreDec 24, 1992 10:52 AM
Bill: Just downloaded your AIBB 5.5 and…. It Works! It came up with a requester saying it could not firmly determine if I had a 68030 or 68EC030 CPU. I clicked on the 68030 button, and all the tests worked fine. I placed rather favorably vs the A3000-25 w/ SCRAM. *Whew!* that was a relief. I had been using AIBB 4.1… perhaps it had a bug! Happy Holidays! & Thanks for your help! — Dave
#82231From: BILL LEACHDec 22, 1992 8:47 AM
Surface mounts Gawk! I have never hear of such a thing. You get a "loading kickstart" message briefly on the screen when you power up the machine? If the boot roms are actually surface mount, then I suspect that, as a practical matter, they can not be replaced but rather the whole motherboard would be replaced. -bill
#82340From: Dave ArreDec 23, 1992 12:44 AM
Bill: Yes, I get a "Loading Kickstart 2.x" and "Rebooting System" message very quickly when I power up. Perhaps I misled you about the ROMs: The ROM spaces I mentioned have no chips in them. They are instead blank areas on the motherboard, with an outline of a chip, and pads for mounting the chip. The two are labled ROM0 and ROM1. I was under the impression that it was in this place that the OS ROMs would go. However, after reading Erik's message, I checked the schematics of the A3000, and it appears that there may be socketed ROMs already (the boot roms) that I overlooked (I was peaking though the side while blowing the dust away…. I didn't take off the equipment platform), so it looks like there are actually *4* ROM places, two occupied by the boot ROMs which I didn't see, and the other two, with just "pads waiting for pins" on the motherboard. Could that be the place for the "A2000" ROMS that the schematics list? The "missing" ROMS seem to be U182 and U183. — AutoPilot Engaged! — Dave
#82379From: BILL LEACHDec 23, 1992 7:42 AM
Dave; Yep, that is correct. The two long chips next to the blanks (toward the front of the machine) are the boot loader ROMs. Did you ever determine what version of AIBB you were running? I have a new one to post now also. Happy Holidays -bill
#82017From: Dave ArreDec 20, 1992 2:40 PM
— Dave
#82078From: DKB SoftwareDec 21, 1992 7:39 AM
Dave, I believe the apparent lockup is due to AIBB looking for req.library. You should be able to find that here in the Libs somewhere. Dean DKB Software
#82118From: Dave ArreDec 21, 1992 3:49 PM
Dean: Nope, that's not it. Here's a Snap from the docs: AIBB uses a shared library for its file requester functions. For this reason AIBB requires that one of the following shared libraries be within the system LIBS: directory: 1. kd_freq.library ( library version 3.0 or greater ) 2. req.library ( library versiopn 2.0 or greater ) 3. asl.library ( AmigaOS 2.0 systems only ) 4. reqtools.library AIBB will search for these libraries in the order given, stopping when it is either able to open one, or none are found. I have kd_freq.library 3.0, asl.library and reqtools.library. I went ahead and downloaded the req.library (2.2) from AmigaTech, but it still froze up. I really don't understand this. — Dave
#82237From: DKB SoftwareDec 22, 1992 9:04 AM
Dave, Hmmm, can you do a L-Amiga M and cycle the screens? I've had AIBB do this to me on a floppy based system where it was looking for 'something' on the boot disk. I'd get a blank screen that I couldn't get rid of by any way other than rebooting. My solution for this is to copy AIBB into the RAM Disk, insert the boot disk into df0: and then run AIBB out of RAM. If this isn't what is happening to you, then I'm stumped! Dean DKB Software
#82341From: Dave ArreDec 23, 1992 12:44 AM
Dean: I haven't tried L-Amiga M, but I CAN tell you that the mouse freezes, and the machine seems to become a doorstop. I had been running AIBB (Ver. 4.1, BTW) out of RAM, but I tried running it off of a floppy as well. Same thing. I even re-booted the machine using a copy of the original Workbench 2.04 floppy, so that NONE of my "extra add-ons" were loaded. Same thing happens. AIBB doesn't even look at the drives… as soon as the message "Evaluating System. Please Wait." fades up, the mouse freezes, and the computer becomes a veritable doorstop. I even went away and came back 30 minutes later… the message was still on the screen, and my screen blanker didn't even take it away. I'm very puzzled. Did I mention that the program SYSINFO will load and perform everything perfectly…. until it gets to "Calculating MFLOPS" in the speed section. The computer then freezes and acts just as it does when I try to load AIBB. Only a CTRL-A-A (reboot) works. I thought it was a bug in the software, but when I downloaded SYSINFO 3.1 and AIBB recently, I began to worry. Is there a problem with my FPU? My copies of DISTANT SUNS / 2.0-FPU and FracBlank_881 work fine….. I'm really puzzled. — Via AutoPilot — Dave
#82381From: DKB SoftwareDec 23, 1992 8:38 AM
Dave, AIBB will freeze the system at that point, although it should only be for at most 10 seconds. Since what AIBB is attempting to do at that point is determine the clock rate of the CPU and FPU, I'd suspect the problem may be hardware. The first thing I'd do is to check the 8520's. Dean DKB Software
#82516From: Dave ArreDec 24, 1992 10:52 AM
Dean: Good News! I just downloaded AIBB 5.5 this morning, and it works! It came up with a requester saying that it could not firmly determine whether I was using a 68030 or 68EC030 CPU. I clicked on the correct button, and all the tests performed flawlessly. I was right on-par (roughly) with the A3000-25 given for comparison. Perhaps 4.1 had a bug? — Dave
#82524From: DKB SoftwareDec 24, 1992 12:25 PM
Dave, From all apperances it would seem so! Glad you finally determined what the problem was. Dean DKB Software
#82607From: Dave ArreDec 25, 1992 12:38 AM
Dean: I'm certainly glad it is not a hardware problem. Thanks for taking the time to help me out, and may you have a joyous Holiday season! All the best for the New Year— — via AutoPilot — Dave