You will not be able to post if you are still using Microsoft email addresses such as Hotmail etc
See here for more information viewtopic.php?f=20&t=7296
DO NOT USE MOBILE / CGNAT DEVICES WHERE THE IP CHANGES CONSTANTLY!
At this time, it is unfortunately not possible to whitelist users when your IP changes constantly.
You may inadvertently get banned because a previous attack may have used the IP you are now on.
So I suggest people only use fixed IP address devices until I can think of a solution for this problem!

Fixing my Milan 040 (or not) blog

Blogs & guides and tales of woo by forum members.
Locked
User avatar
alexh
Site sponsor
Site sponsor
Posts: 1200
Joined: Tue Oct 17, 2017 4:51 pm
Location: Oxfordshire

Re: Fixing my Milan 040 (or not) blog

Post by alexh »

Is there an address map for the Milan somewhere?
Principal ASIC Engineer - SystemVerilog, VHDL
Thalion Webshrine - http://thalion.atari.org
STf,STfm,STe,MegaST,MegaSTe,Falcon060
A500+,A600,A4000/060,CD32,CDTV
User avatar
exxos
Site Admin
Site Admin
Posts: 27530
Joined: Wed Aug 16, 2017 11:19 pm
Location: UK
Contact:

Re: Fixing my Milan 040 (or not) blog

Post by exxos »

alexh wrote: Mon Jun 23, 2025 3:28 pm Is there an address map for the Milan somewhere?
No idea. But Uwe did say basically the first outputs from the flash should be all zeros and 90. So the address bus has to be all low to do that. But its not.
User avatar
exxos
Site Admin
Site Admin
Posts: 27530
Joined: Wed Aug 16, 2017 11:19 pm
Location: UK
Contact:

Re: Fixing my Milan 040 (or not) blog

Post by exxos »

So those datalines to the flash are driven here. Uwe did mention earlier these can screw up the flash address.. So I need to see whats going in and out of that chip next :)

1.PNG
1.PNG (156.63 KiB) Viewed 146 times

EDIT:

So both sides of the chip are high :roll: the only thing I could do is remove the chip and retest...

EDIT2:

Nope, same problem :( That all comes from U9 the Intel chip :roll:
User avatar
exxos
Site Admin
Site Admin
Posts: 27530
Joined: Wed Aug 16, 2017 11:19 pm
Location: UK
Contact:

Re: Fixing my Milan 040 (or not) blog

Post by exxos »

Uwe replied again..
maybe the A16,A17,A18 stuck high is ok. The ISA Bridge accesses the ROM usually at 0xF0000 to 0xFFFFF. So it makes total sense that the bootblock is in the last 64k of the flash..
I had to email him back as I'm confused now. As earlier he said the flash start should be 0x00000000 and the second word is the reset vector 0x00000090. Which is at the start of flash. So I'd assume address zero would be first.

I odered the Intel chip as well anyway. Problem is they coming from Poland. So going to be some days.. Problem there is, if I take the Milan off my desk to something else, the Milan will never get back on my desk again :roll:
peters
Posts: 188
Joined: Sat Feb 25, 2023 8:44 pm

Re: Fixing my Milan 040 (or not) blog

Post by peters »

Is the vertical ram socket better than the tilted ones ?

Mine have always been a bit unreliable. If I move the ide cables it usually upsets the simm cards.

I tried to fit 128MB Simms but could never get more than 1 to work. I think it's partly the quality of the sockets and partly because they are so close together.
User avatar
exxos
Site Admin
Site Admin
Posts: 27530
Joined: Wed Aug 16, 2017 11:19 pm
Location: UK
Contact:

Re: Fixing my Milan 040 (or not) blog

Post by exxos »

peters wrote: Mon Jun 23, 2025 9:37 pm Is the vertical ram socket better than the tilted ones ?
I would think no. They are just generic sockets. But also as I been saying for years, a lot of simms are to wide and damage the sockets. I see it all the time on STEs. Once you fit a bad simm, anything after will be likely bad because the socket is now damaged.
SteveBagley
Posts: 29
Joined: Fri Jul 26, 2024 3:53 pm

Re: Fixing my Milan 040 (or not) blog

Post by SteveBagley »

Isn't the Milan basically a 68040 mangled onto the business end of a PC? So the Milan BootBlock ROM being at $F0000 makes sense because that's where the PC hardware will put it. I spent a long flight last summer reading up on the Milan so the follow links might be useful.

Attached is a file from the MagiC describing how the hardware appears from the Motorola side (which gives some indication to how it is mapped into the PC hardware) and he source code to the boot block is on github: https://github.com/ps68060/MilanBootblock

The bootblock indicates that the as far as the Motorola is concerned, at reset the bootblock is at $f0000, but it immediately configures the PLX chip so it's mapped to $40F00000 and the code continues execution from there:

Code: Select all

romstart:
; Resetvektor
            dc.l    0                       ; Berr-Emulation
            dc.l    reset-romstart
            dc.l    reset-romstart
            dc.l    reset-romstart
and

Code: Select all

reset:      MOVE    #$2700,SR
            reset

; Transparent Translation fr Schutz des IO-Bereiches ab 2GB
; Bit 8 - U0 gesetzt -> Little endian
            move.l  #$807F0000 | %1110000101000000,d0
;           movec.l d0,pmmu_dttr0
            dc.w    $4e7b,$0006             ; movec D0,DTTR0

; PLX init
            move.l  #jmp1-romstart+$f0000,a0 ; abs. Adr fuer ROM 000F0000
            move.l  #$C0000000,PLX_DMRR.w   ; $c0000000 = 1GB Range
            jmp     (a0)                    ; Funktioniert wegen Prefetch
jmp1:
            move.l  #jmp2-romstart+$f0000+$40000000,a0 ; abs. Adr fuer ROM 400F0000
            move.l  #$40000000,PLX_DMLBAM.w ; $40000000
            jmp     (a0)
            
Is it worth seeing if you get 4 on A32-A28 (on the Motorola side) soon after boot as it switches the addresses?

Steve
Attachments
milan_hw.txt
(2.86 KiB) Downloaded 9 times
User avatar
exxos
Site Admin
Site Admin
Posts: 27530
Joined: Wed Aug 16, 2017 11:19 pm
Location: UK
Contact:

Re: Fixing my Milan 040 (or not) blog

Post by exxos »

@SteveBagley Thanks. Yeah I need to test the CPU address as I don't even know if the CPU is good at this point.

The flash may live higher up. But I think it's confusing as the flash address bus doesn't seem to be connected to the CPU address bus. It's driven by the Intel chip. But even so, the flash address simply has to be zero to access the data at all. The cpu address could anything at that point as I didn't know what's decoding what exactly.

I'd assume the first CPU cycles start from address zero. Then the flash address should (I assume) be decoded for address zero as well.

The flash could like high up in the memory map, but that would mean the cpu would have to be at that address from power up.. AFAIK, the CPU always starts from address zero.

Normal STFM ROM acts like that. Then does a jump to higher ROM address, which is decided as FC0000 or whatever. But I don't think it's getting that far.

After reset the MILAN sets the flash CS/OE low, but there is nothing in the Flash higher up. I don't know how much is used offhand, but I don't think it was much. Those high adress lines set high must mean it's decoding flash more like in the upper half where there is nothing there. It's just outputting FF and that's it. Nothing to run.

Its why I think the flash address should be all zeros after reset as I can't see it working another way. But this is the flash address bus not the cpu address bus. It's a bit confusing to follow. But looks like the whole ISA address bus isn't the same address bus as the CPU. Flash seems to be wired into the ISA address bus.

Like I thought earlier, might be easier if I just hardwire the cpu address bus to all zeros as it should result in all zeros on the flash address. The flash should output all zeros on the first address set by the CPU.
peters
Posts: 188
Joined: Sat Feb 25, 2023 8:44 pm

Re: Fixing my Milan 040 (or not) blog

Post by peters »

exxos wrote: Mon Jun 23, 2025 9:44 pm
peters wrote: Mon Jun 23, 2025 9:37 pm Is the vertical ram socket better than the tilted ones ?
I would think no. They are just generic sockets. But also as I been saying for years, a lot of simms are to wide and damage the sockets. I see it all the time on STEs. Once you fit a bad simm, anything after will be likely bad because the socket is now damaged.
Do they have the bigger end catches rather than those flimsy clips that hook over the sides of the PCB ?

That would be one upgrade I would consider if it means I can install all 512MB
User avatar
exxos
Site Admin
Site Admin
Posts: 27530
Joined: Wed Aug 16, 2017 11:19 pm
Location: UK
Contact:

Re: Fixing my Milan 040 (or not) blog

Post by exxos »

peters wrote: Mon Jun 23, 2025 10:29 pm Do they have the bigger end catches rather than those flimsy clips that hook over the sides of the PCB ?
That would be one upgrade I would consider if it means I can install all 512MB
I use the sockets with the metal end tabs but I no idea where you get 72pin versions from anymore.

There's a few in my store but don't know if they metal or not. But they are proper brand ones maybe Tyco.

https://www.exxosforum.co.uk/atari/store2/#0072

I'll look tomorrow as I need to change the socket on my Milan anyway..
Locked

Return to “MEMBER BLOGS”