OS 3.1 + Apollo 1260 - Strange Behaviour/Issue

[rabidgerry]

Hi all,

so I got an Apollo 1260 (Apollo 1240 originally converted to a 1260 by Stachu) only recently fired it up. First thing I did was break the god dam fan off! But apparently I can just stick it on again But after this I managed to get it seated and actually close the trap door.

So I figured something was badly wrong as I noticed it loaded up fine, and all seemed well until..............say I tried to double click into a partition drive and it would crash immediatley.

I figured this is timing fix issues or perhaps library issues.

The first thing I did was swapped out what 68060.library I had already which I got from the MMULib. I instead, put in the libs that are on the actual Apollo1260.dms image (which is basically the install disk that would have come with the accelerator to begin with). These were 68040.library (I believe this to be a dummy library as it's tiny), 68o40.library and then a 68060.library plus the CPU60 command into C:

So once I did this I got the same shit. Only by accident I decided to try some other things that were not clicking into a partition drive icon. I tried opening Dopus. This would bring up an error. Then I opened a shell window and I was able to CD my way into the SYSINFO drawer and then RUN sysinfo. SYSINFO would then load. Well that was a minor success I thought! I was able to do the tests in SYSINFO also, so take the speed of the system, test the drive speed etc etc. So this was something at least. However when I closed out of sysinfo I decided lets try opening the drawer where Doom Attack was installed. What do you know? It ran!! Played it for a bit! All good. Then I closed out of that and went to open my workbench partition and I expected it to crash again like before! What do you know??? It allowed me to go in and freely go from drawer to drawer opening everything I wanted. I was also able to open any other partitions as I wanted!

Well this made me think my issue was perhaps solved! Wrong

Basically I shut the Amiga down and left it for a bit. I came back later and tried to open workbench partition again. BAM! Crashed the Amiga and it restarted!

So I booted again and did exactly what I did the first time. Go into shell cd to sysinfo run a few tests and them close out and then try and open some drawers. I did this and what do you know? It started allowing me to open my drawers and partitions again without crashing.

How strange I thought. So it's a work around currently to do this process in order to get things going but I am just wondering does anyone have any ideas as to what might be causing these crash issues and what I could try to resolve them?

Thanks in advance for any help or suggestions

[paul1981]

Personally, I would make a copy of your Workbench 3.1 disk, and then modify it by adding the Apollo 68060.library and the dummy 68040 one to libs (don't forget to rename it to 68040.library) and also make room for the cpu60 command in c:

You'll probably have to delete a few things from you Workbench disk to make room, but once you do this and cold boot the disk then you'll get a clearer picture stability wise. You can run a few things from your hdd then (drives will show up from booting the floppy).

[rabidgerry]

Originally Posted by paul1981

Personally, I would make a copy of your Workbench 3.1 disk, and then modify it by adding the Apollo 68060.library and the dummy 68040 one to libs (don't forget to rename it to 68040.library) and also make room for the cpu60 command in c:

You'll probably have to delete a few things from you Workbench disk to make room, but once you do this and cold boot the disk then you'll get a clearer picture stability wise. You can run a few things from your hdd then (drives will show up from booting the floppy).

Not sure I follow you, so are you suggesting I run of a floppy disk permanently?

There is no need to modify the dummy 68040 library, doesn't say to do that in the manual, just says replace the one that you might already have in there with the tiny little one they supply you with.

I don't really fancy having to use a floppy disk just to use this accelerator. It's actually less hassle to continue as I am where I just go into sysinfo, do a speed test, come out and everything works.

Thanks for the suggestion though.

[fitzsteve]

Haven't had an Apollo 1260 for years but I'm sure you needed to add the CPU060 command to the startup sequence.

[rabidgerry]

Originally Posted by fitzsteve

Haven't had an Apollo 1260 for years but I'm sure you needed to add the CPU060 command to the startup sequence.

Yes I did that

CPU60 B

Which is for best configuration according to the manual.

[NovaCoder]

Oh, fun times

Make sure your power supply is up to the job!

Also, reduce your pain by reading about mine (and others)

http://eab.abime.net/showthread.php?t=63739

remapollo is a must for getting good speed out of your card..

http://eab.abime.net/showthread.php?...ight=remapollo

http://eab.abime.net/showthread.php?...ight=remapollo


This is my old Apollo startup sequence (this is for 3.9 + Boing Bag updates, no reboot needed).

I'm not using SpeedyIDE here because I'm using the patched scsi.device that already has it built in.

PHP Code:

; $VER: Startup-Sequence_HardDrive 45.2 (19.1.2001)
; Startup-Sequence for AmigaOS 3.9


C:DRAP
C:RemAPollo QUIET MODULES DEVS:Modules/scsi.device mathffp.library mathieeesingbas.library exec_45.20 console.device_44.8 ram-handler_44.23 shell_45.7 bootmenu_44.7 filesystem.resource_45.10 filesystem_45.13 BBlank NoClick


C:SetPatch NOROMUPDATE QUIET




C:Version >NIL:
C:AddBuffers >NIL: DF0: 15
FailAt 21

C:MakeDir RAM:T RAM:Clipboards RAM:ENV RAM:ENV/Sys
C:Copy >NIL: ENVARC: RAM:ENV ALL NOREQ

Resident >NIL: C:Assign PURE
Resident >NIL: C:Execute PURE

Assign >NIL: ENV: RAM:ENV
Assign >NIL: T: RAM:T
Assign >NIL: CLIPS: RAM:Clipboards
Assign >NIL: REXX: S:
Assign >NIL: PRINTERS: DEVS:Printers
Assign >NIL: KEYMAPS: DEVS:Keymaps
Assign >NIL: LOCALE: SYS:Locale
Assign >NIL: LIBS: SYS:Classes ADD
Assign >NIL: HELP: LOCALE:Help DEFER



C:CPU60 B >NIL:
;C:CMQ060
Run >NIL: CopyMem060

C:FBlit
C:FText



BindDrivers
C:Mount >NIL: DEVS:DOSDrivers/~(#?.info)

C:LoadMonDrvs

SetEnv Language "english"
SetEnv Workbench $Workbench
SetEnv Kickstart $Kickstart
UnSet Workbench
UnSet Kickstart

C:AddDataTypes REFRESH QUIET
C:IPrefs

Run > NIL: "System:C/VisualPrefs"
System:C/FPPrefs


C:ConClip

Path >NIL: RAM: C: SYS:Utilities SYS:Rexxc SYS:System S: SYS:Prefs SYS:WBStartup SYS:Tools SYS:Tools/Commodities

SYS:System/REXXMast >NIL:

IF EXISTS S:User-Startup
  Execute S:User-Startup
EndIF

Resident Execute REMOVE
Resident Assign REMOVE

C:LoadWB
EndCLI >NIL: 


Once you get it and running properly I can recommend that you try some of the following games optimized for an 060:

My ports

[paul1981]

Originally Posted by rabidgerry

Not sure I follow you, so are you suggesting I run of a floppy disk permanently?

There is no need to modify the dummy 68040 library, doesn't say to do that in the manual, just says replace the one that you might already have in there with the tiny little one they supply you with.

I don't really fancy having to use a floppy disk just to use this accelerator. It's actually less hassle to continue as I am where I just go into sysinfo, do a speed test, come out and everything works.

Thanks for the suggestion though.

Sorry, I forgot which library was the dummy one, so yes just copy both libs across to the floppy. The cpu60 command needn't be run, if Setpatch has been run (happens every boot).
Anyway, no of course I'm not expecting you to use the floppy as a permanent measure...I have no idea what reset resident software you may have on your hard disk, or other patches etc. or weird library versions and so the best way to quickly know whether you have a real hardware problem is to make a floppy boot disk which bypasses all of those above problems and gives you a clean system. The idea is you turn off you Amiga for 10 seconds, put the disk in and boot the disk and you have a clean system.

[rabidgerry]

Originally Posted by paul1981

Sorry, I forgot which library was the dummy one, so yes just copy both libs across to the floppy. The cpu60 command needn't be run, if Setpatch has been run (happens every boot).
Anyway, no of course I'm not expecting you to use the floppy as a permanent measure...I have no idea what reset resident software you may have on your hard disk, or other patches etc. or weird library versions and so the best way to quickly know whether you have a real hardware problem is to make a floppy boot disk which bypasses all of those above problems and gives you a clean system. The idea is you turn off you Amiga for 10 seconds, put the disk in and boot the disk and you have a clean system.

Ok, I totally get ya now. And if I sounded a little "wtf" with my reply apologies, just frustration getting the better of me and I was burning too much midnight oil! I will give this a bash and see what the outcome is. You could be well right, may be a software clash with something on my ClassicWB cf card.

Originally Posted by NovaCoder

Oh, fun times

Make sure your power supply is up to the job!

Also, reduce your pain by reading about mine (and others)

http://eab.abime.net/showthread.php?t=63739

remapollo is a must for getting good speed out of your card..

http://eab.abime.net/showthread.php?...ight=remapollo

http://eab.abime.net/showthread.php?...ight=remapollo


This is my old Apollo startup sequence (this is for 3.9 + Boing Bag updates, no reboot needed).

I'm not using SpeedyIDE here because I'm using the patched scsi.device that already has it built in.

PHP Code:

; $VER: Startup-Sequence_HardDrive 45.2 (19.1.2001)
; Startup-Sequence for AmigaOS 3.9


C:DRAP
C:RemAPollo QUIET MODULES DEVS:Modules/scsi.device mathffp.library mathieeesingbas.library exec_45.20 console.device_44.8 ram-handler_44.23 shell_45.7 bootmenu_44.7 filesystem.resource_45.10 filesystem_45.13 BBlank NoClick


C:SetPatch NOROMUPDATE QUIET




C:Version >NIL:
C:AddBuffers >NIL: DF0: 15
FailAt 21

C:MakeDir RAM:T RAM:Clipboards RAM:ENV RAM:ENV/Sys
C:Copy >NIL: ENVARC: RAM:ENV ALL NOREQ

Resident >NIL: C:Assign PURE
Resident >NIL: C:Execute PURE

Assign >NIL: ENV: RAM:ENV
Assign >NIL: T: RAM:T
Assign >NIL: CLIPS: RAM:Clipboards
Assign >NIL: REXX: S:
Assign >NIL: PRINTERS: DEVS:Printers
Assign >NIL: KEYMAPS: DEVS:Keymaps
Assign >NIL: LOCALE: SYS:Locale
Assign >NIL: LIBS: SYS:Classes ADD
Assign >NIL: HELP: LOCALE:Help DEFER



C:CPU60 B >NIL:
;C:CMQ060
Run >NIL: CopyMem060

C:FBlit
C:FText



BindDrivers
C:Mount >NIL: DEVS:DOSDrivers/~(#?.info)

C:LoadMonDrvs

SetEnv Language "english"
SetEnv Workbench $Workbench
SetEnv Kickstart $Kickstart
UnSet Workbench
UnSet Kickstart

C:AddDataTypes REFRESH QUIET
C:IPrefs

Run > NIL: "System:C/VisualPrefs"
System:C/FPPrefs


C:ConClip

Path >NIL: RAM: C: SYS:Utilities SYS:Rexxc SYS:System S: SYS:Prefs SYS:WBStartup SYS:Tools SYS:Tools/Commodities

SYS:System/REXXMast >NIL:

IF EXISTS S:User-Startup
  Execute S:User-Startup
EndIF

Resident Execute REMOVE
Resident Assign REMOVE

C:LoadWB
EndCLI >NIL: 


Once you get it and running properly I can recommend that you try some of the following games optimized for an 060:

My ports

Thanks Novacoder I will give these a bash. Expect more and more questions as I

Haven't even attempted remapollo yet.

And yes I fully intended to get down with your optimized 060 games I cannot wait actually. Asked you a question on an old one of you Boom youtube videos a few weeks ago but it probably doesn't matter now.

Ok thanks for all trying to help I will be back!

[NovaCoder]

No worries, just replied now

You can use this SysSpeed module to compare your performance with my old Apollo 1260 running at 80 Mhz

http://aminet.net/package/util/moni/...1260_80Mhz_AGA

[paul1981]

Originally Posted by rabidgerry

Ok, I totally get ya now. And if I sounded a little "wtf" with my reply apologies, just frustration getting the better of me and I was burning too much midnight oil!

It's weird, because I felt this too when I read your reply. But as I read it again now, I don't detect any frustration in your reply at all. I must have read your emotional state at the time through the ether, because it has definitely changed now and I don't pick up those vibes at all. Weird eh?

[rabidgerry]

Originally Posted by paul1981

It's weird, because I felt this too when I read your reply. But as I read it again now, I don't detect any frustration in your reply at all. I must have read your emotional state at the time through the ether, because it has definitely changed now and I don't pick up those vibes at all. Weird eh?

That is weird I think when I was typing it I felt I had a bit of frustration coming through but I guess it didn't even though some psychic link passed the feelings to you regardless

Originally Posted by NovaCoder

No worries, just replied now

You can use this SysSpeed module to compare your performance with my old Apollo 1260 running at 80 Mhz

http://aminet.net/package/util/moni/...1260_80Mhz_AGA

I will give this a go! Thankyou

Still digesting previous responses so will report back with findings.

[rabidgerry]

Originally Posted by paul1981

Personally, I would make a copy of your Workbench 3.1 disk, and then modify it by adding the Apollo 68060.library and the dummy 68040 one to libs (don't forget to rename it to 68040.library) and also make room for the cpu60 command in c:

You'll probably have to delete a few things from you Workbench disk to make room, but once you do this and cold boot the disk then you'll get a clearer picture stability wise. You can run a few things from your hdd then (drives will show up from booting the floppy).

Ok so I did this. Stripped a few things out, added the exact libraries I got from the Apollo060.dms disk.

Made a disk out of the ADF

rebooted using the 3.1 install disk with added libraries

Clicked on partition drives

ALL WORKING FINE

So we conclude something in my 3.1 build is causing the issue I am having.

Question is now finding out what that is.

Originally Posted by NovaCoder

Oh, fun times

Make sure your power supply is up to the job!

Also, reduce your pain by reading about mine (and others)

http://eab.abime.net/showthread.php?t=63739

remapollo is a must for getting good speed out of your card..

http://eab.abime.net/showthread.php?...ight=remapollo

http://eab.abime.net/showthread.php?...ight=remapollo


This is my old Apollo startup sequence (this is for 3.9 + Boing Bag updates, no reboot needed).

I'm not using SpeedyIDE here because I'm using the patched scsi.device that already has it built in.

Ok so I started going through the threads you sent me, I downloaded remapollo also. Got nowhere with is. Tried following the guide and nowhere really. First thing it asks you to do is run the "patch" it includes. This failed for me and through out the error saying only commodore 04 libraries supported.

The remapollo I was trying was v1.7

Upon reading your first thread you recommend I see someone links to remapollo v1.8. So I have downloaded this and yet to try it.


I still want to fix this issue but I have an important development.

I built a ClassisWB 3.1.4 cf card on Friday night. It worked with the Apollo right away. No having to do anything. However (this is something that will need improved over all my builds using the Apollo) the file transfer rate is laughable. For some reason I am getting about 1,000,046 on the 3.1 build and just under 999,989 in 3.1.4 weird eh? Also in SYSinfo I get 62.6mflops and about 5966 in drystones in 3.1 but in 3.1.4 I'm getting around 5824 or around that in drystones. All a bit weird.

Still to go through completely those threads novacoder but I have to admit a know dam all about startup-sequence so a lot of that is going over my head.

Sure I added C:CPU60 B in there and it appears to work but I took a gable with that and hoped it would. Basic knowledge of the startup-sequence really is all I have.

Somethings I have done in both 3.1 and 3.1.4 is play Doom Attack (the only version I have at the mo but I will bet getting your versions once we are up and running better) for a couple of hours at a time. It flies!!! Flies to the point that even I'm getting a little ill playing it

Defo room to get this card to perform better I think.

Can't live with those transfer speeds. I was using my ACA1233n-4- before and I was getting about 3,500,000 using the IDE speeder on DH0: and about 6,700,000 on other partitions when I tested them. That's a lot more noticeable. All using sysinfo 4.3 btw.

what I was getting on the 3.1 CF card, still had the old sysinfo then though.

[paul1981]

The Apollo doesn't have the IDE speeder, so I wouldn't expect it to be as fast, but you'll not notice this whilst navigating Workbench as the 060 will more than make up for it. What MHz is your 060 by the way?

Under 1MB/s does seem pretty slow though, it's supposed to be 2.5MB/s with the A1200 IDE or there abouts. Mind you that's a real HDD, not sure about CF's as they come in a variety of speeds.

The difference speed ratings between your different boot setups will depend on a few things, exec being in Fast RAM is a big one, also whether ROM is in Fast RAM or not is another big one. There's other patches and things too, but one setup isn't naturally slower than the other, it's just a matter of the correct patches being applied.

Anyway, back to this boot disk. I'm not sure whether the Install3.1 disk runs SetPatch or not, that's why I said do it with the Workbench disk. You can copy SetPatch over to the C drawer and add it to the top of the Install3.1 startup-sequence if it's not already there - just thought I'd mention this as the 68060.library won't be loaded without it, and it needs to be loaded in order to test your Apollo card.

[rabidgerry]

Originally Posted by paul1981

The Apollo doesn't have the IDE speeder, so I wouldn't expect it to be as fast, but you'll not notice this whilst navigating Workbench as the 060 will more than make up for it. What MHz is your 060 by the way?

Under 1MB/s does seem pretty slow though, it's supposed to be 2.5MB/s with the A1200 IDE or there abouts. Mind you that's a real HDD, not sure about CF's as they come in a variety of speeds.

The difference speed ratings between your different boot setups will depend on a few things, exec being in Fast RAM is a big one, also whether ROM is in Fast RAM or not is another big one. There's other patches and things too, but one setup isn't naturally slower than the other, it's just a matter of the correct patches being applied.

Anyway, back to this boot disk. I'm not sure whether the Install3.1 disk runs SetPatch or not, that's why I said do it with the Workbench disk. You can copy SetPatch over to the C drawer and add it to the top of the Install3.1 startup-sequence if it's not already there - just thought I'd mention this as the 68060.library won't be loaded without it, and it needs to be loaded in order to test your Apollo card.

Setpatch is in the startup sequence of the 3.1 builds. I will get a copy and show it here. I know bugger all about it though. I'd love to be able to utilize these things a bit more.

My Apollo is 80mhz.

About the speed thing, I think on my ACA1233n-40 with the IDE speeder off I get about 2mb. So this is dam slow if you ask me. On one of Novacoders threads someone mentions that their transfer rate is dam slow also and wasn't even great after they got a Fast IDE kit which is something I was thinking off. Better off getting the best config now though before any extras.

[NovaCoder]

yep it can be a bit tricky to setup.

Best thing to test performance is to compare yours to mine in SysSpeed, they should be very similar if you have it setup properly.

From memory my HD transfer was about 3 MB/s using PFS 060 file management.

MIPS should be up around 106 in SysInfo if you have the right version of SysInfo that understands 060's.

[rabidgerry]

Originally Posted by NovaCoder

yep it can be a bit tricky to setup.

Best thing to test performance is to compare yours to mine in SysSpeed, they should be very similar if you have it setup properly.

From memory my HD transfer was about 3 MB/s using PFS 060 file management.

MIPS should be up around 106 in SysInfo if you have the right version of SysInfo that understands 060's.

I did this (I stumbled through a comparison btw as I have never used sysspeed before). From what I gather I was behind your all the time.

I can try run this again and is there a way for me to save the report?

Off the topic question, did you leave your trap door open? I have mine inside and trap door on but it's bulging. Also my keyboard on the right side doesn't come down properly on the little rail for it to rest on. The Apollo is squeezed in there but it doesn't look like it is causing this. I shawn a torch in there and there still seems a tiny little gap. No idea why this is.

[NovaCoder]

Nope I removed the trap door to improve airflow.

Use some hot glue along the frame rail to hold it securely otherwise the weight will all be on the expansion bus connector

[rabidgerry]

Originally Posted by NovaCoder

Nope I removed the trap door to improve airflow.

Use some hot glue along the frame rail to hold it securely otherwise the weight will all be on the expansion bus connector

Yeah I noticed that if I leave it free hanging it is putting weight on that edge.

I might try and modify the new trap door I got which is mega perforated to allow airflow. Have the Amiga sitting on stilts now. Might fashion together some kind of platform/frame to keep it off which ever surface it sits on.

The f-ing fan came off also and it actually resting freely over the CPU. I was chatting to Stachu about this and he said just glue it back on. I'd worry it would come off again as there is old glue residue on it already.

[crazyc]

If you look on thingiverse I think there are risers for the a1200 to lift it up.

[crazyc]

When I had my b1260 and scsi kit in the wedge I basically cut away almost all the trapdoor cover but left enough to support the card and a fan