OK so many of you have seen the photos of the A4000T and rare 68060 card that I just got. If you haven't, see here:
http://www.hd-zone.com/amiga/A4000T.htmlAnd a Photo of the Actual CPU. Notice the 68060 50MHz in WHITE PAINT on the top and the number on the side: S23328 002


What you don't know is that I have had all sorts of crashing and problems running software on it and even installing OS 3.9.
Lots of #8000003D errors, #80000003, 4, or 6 errors...amidock crashing, lha thinking files are corrupted, the program "WhichAmiga" by Piru just freezes after clicking on the icon...AmiGod crashes...
I've tried every available 68060 library including various phase-5 060 libs (none work), the GVP libraries from Greg Donner's page, Ralph Babel's 060 library, Thomas Richter's libs from aminet, Carsten Schlote's...the updated 060 library from Ralph modified by Jeff Boyer (the one that ships with the card)...but nothing is able to fix this issue.
OK so I've done alot of researching on this and found some interesting posts:
From this site:
http://eab.abime.net/archive/index.php/t-27664.html"I cannot understand the german CT article but it appears people are making FAKE Rev6's?
05.05.06 R.C. has problems with 25 CPUs for CTs he recently ordered. Apparently they are 060's with falsified inscriptions. They have a white painted MC68060RC50 reference instead of the regular engraved one. It is easily recognized as well by the number on the left side (e.g. S23328-002 instead of -004). These 68060 have a maximum overclock speed of only 75MHz, genuine Rev. 6 overclock to +90 MHz (see also the pictures of Rodolphe Czuba of the wrong ones and the genuine Rev.6)"
And then a thread where a user had a similar problem:
http://www.amiga.org/forums/showthread.php?t=11384Piru writes:
On the #8000003D error:
"This is a "unimplemented interger instruction" exception (vector offset 0x0F4).
This means that 68060 has encountered an instruction it doesn't support in hardware. The unimplemented instructions are:
DIVU.L
,Dr:Dq ; 64 / 32 => 32r,32q
DIVS.L ,Dr:Dq ; 64 / 32 => 32r,32q
MULU.L ,Dh:Dl ; 32 * 32 => 64
MULS.L ,Dh:Dl ; 32 * 32 => 64
MOVEP Dx,(d16,Ay) ; size = W or L
MOVEP (d16,Ay),Dx ; size = W or L
CHK2 ,Rn ; size = B, W or L
CMP2 ,Rn ; size = B, W or L
CAS2 Dc1:Dc2,Du1:Du2,(Rn1):(Rn2) ; size = W or L
CAS Dc:Du, ; size = W or L, misaligned
The exception tells us that MC68060ISP is not installed. See M68060UM/AD Appendix C for details.
In simple terms MC68060ISP provides emulation for the unimplemented integer opcodes. Typically the 68060.library includes this (aswell as other 68060SP parts, emulation for missing FPU things for example).
If no proper 68060.library is installed, the system will not function properly.
Now, it could well be that the provided 68060.library doesn't properly handle 68LC060 and assumes FPU. In that case, contact your dealer and/or the card manufacturer for working drivers.
All the bad things said about Phase5/DCE, at least they had working drivers for their hardware"
----
Ok...so is it possible that this weird 060 is an LC and all the 060 libraries I am using are assuming a FPU is present? And the random crashing I am getting is because the FPU is either disabled or half-working?
I have tried, as I said:
68060.library 1.2 (ralph babel)
68060.library 2.2 (jeff boyer (this is ralphs updated)
all the other libraries on this page:
http://www.gregdonner.org/gvp4060/downloads.html
None of which solve the problem...
I have two other 060 chips here on phase-5 boards (MK3 and CS PPC - with dead PPC! ) that are working fine...but I am too chicken to try and pull them off and put them on this new 060 card...
Any thoughts?