Homebuild IDE interface that autoboots?

[fafler]

Ultimately, i want to build a MMC interface for my A500+ (KS 2.0). Right now, i'm looking for scematics for a homebuild autobooting IDE controller.

[amipal]

@fafler

Quick search on Aminet gave me this:
http://www.aminet.net/package.php?package=hard/hack/hddmem.lha

There may be more from here:
http://www.aminet.net/search.php?name=IDE&path[]=hard&q_desc=OR&desc=IDE

;-)

[fafler]

Thanks. I did a search for "ide autoboot", and it didn't come up with anything. Anyway, it's not very well documented, so other links will be appreciated.

[KThunder]

an autobooting device needs a rom to send mount info and memory footprint info to the kickstart on startup you would need to know how this all works to build one and searched for a long time on this and never found an autobooting one.

non-autobooting interfaces work well and need only a bootable disk a small startup-sequence and boot hd mount

[fafler]

Have you looked at the link amipal posted? It looks alot more complicated than the other A500 IDE controllers i've seen. But it doens't seem to have other ROM's than the ones used for the kickstart switcher.

[Doobrey]

I was thinking of the same thing over Xmas, came up with 2 ideas.
 1. Build a simple interface but hack the driver into the Kickstart ROM.
 2. Rip out Gary and hack in the A600's Gayle chip..the driver is already in the Kickstart (for v37.300+ anyway) plus the added bonus of a PCMCIA slot too..assuming it all works (cos they aren't pin compatible, and Gayle needs a few more signals that aren't available on the Gary socket)

[Ohforf]

The latest non-autoboot IDE interface Version
can be found at http://www.students.tut.fi/~leinone3/ide/

a quote from the readme File :

I have built a new interface which will eventually support Autoconfig(tm) and autoboot. The interface uses a Lattice Semiconducor ispLSI1032E in PLCC-84 package. The autoboot ROM will be a 24C128 or 24C256 serial EEPROM in 8 pin DIP or SOIC package.

It's from 9. December 2005.

I will build the simple non Autoboot Version first
and see what happens.

[fafler]

Doobrey wrote:
I was thinking of the same thing over Xmas, came up with 2 ideas.
 1. Build a simple interface but hack the driver into the Kickstart ROM.

That would be a good solution, but i have no idea how to do that.

 2. Rip out Gary and hack in the A600's Gayle chip..the driver is already in the Kickstart (for v37.300+ anyway) plus the added bonus of a PCMCIA slot too..assuming it all works (cos they aren't pin compatible, and Gayle needs a few more signals that aren't available on the Gary socket)

Wouldn't that give the A500 all the A600's memory problems? I don't want to give up my 8 mb FastRAM for PCMCIA.

[fafler]

Ohforf wrote:
The latest non-autoboot IDE interface Version
can be found at http://www.students.tut.fi/~leinone3/ide/

a quote from the readme File :

I have built a new interface which will eventually support Autoconfig(tm) and autoboot. The interface uses a Lattice Semiconducor ispLSI1032E in PLCC-84 package. The autoboot ROM will be a 24C128 or 24C256 serial EEPROM in 8 pin DIP or SOIC package.

It's from 9. December 2005.

I will build the simple non Autoboot Version first
and see what happens.

I couldn't find anything about his new interface. Maybe a email would help on that matter.

[Doobrey]

fafler wrote:
That would be a good solution, but i have no idea how to do that.

The simplest way would be to just overwrite scsi.device with the new driver code compiled to ORG at the same ROM address,then burn a new kickstart (Or even better, build Redskull's KF500 flashrom board on Aminet, and set the second half of the flash to be an 'extended' rom and write the ide driver in there)
 The downside to overwriting scsi.device is that the new code can't be any larger, otherwise you'll start overwriting other residents too!

Wouldn't that give the A500 all the A600's memory problems? I don't want to give up my 8 mb FastRAM for PCMCIA.

Yeah that's true. It didn't even cross my mind when I thought about it, cos I don't have 8mb to worry about, lucky git.
 I guess the neatest way would be to use a PLD to emulate the IDE registers of Gayle and let Gary do the rest

[fafler]

The simplest way would be to just overwrite scsi.device with the new driver code compiled to ORG at the same ROM address,then burn a new kickstart (Or even better, build Redskull's KF500 flashrom board on Aminet, and set the second half of the flash to be an 'extended' rom and write the ide driver in there)
 The downside to overwriting scsi.device is that the new code can't be any larger, otherwise you'll start overwriting other residents too!

It still sounds a bit complicated. I don't have any knowledge on 68000 assembler and the needed system calls.

Yeah that's true. It didn't even cross my mind when I thought about it, cos I don't have 8mb to worry about, lucky git.

Well, the plan was to include a 72 pin SIMM slot for a 16 mb FPM SIMM and use the first 16 bit of the 32 bit bus.

I guess the neatest way would be to use a PLD to emulate the IDE registers of Gayle and let Gary do the rest

That might work. Then i would just need to upgrade to kickstart 2.05. Where can i find detailed info on Gayle?

[mrmkl]

I have now the user-nick for this site, so I can post some info here.
I haven't done yet the autoconfig(tm) part, but I have made the board look like the A600 gayle, so it autoboots with KS40.063 3.1 ROM. I use it with a IDE->CF adapter, like this: http://www.iki.fi/mkl/ide2cf/