DataFlyer SCSI Plus firmware 2.1

[papa_november]

I need version 2.1 of the firmware ROM for the DataFlyer SCSI cards, but I've only been able to locate version 2.0. Anyone have a card with a 2.1 ROM?

[papa_november]

I've managed to locate a card with a 2.1 ROM, but it might be corrupted. The firmware is 16K, and the ROM is 32K. It looks like the firmware was doubled to fill the larger ROM, but the two copies of the firmware differ by a few seemingly random bytes.

I'll have to test it to make sure it works, but if anyone else has a dump of the 2.1 firmware to compare my dump to, I'd greatly appreciate it.

[papa_november]

Yeah, it's corrupted. Instant yellow screen on boot.

[alexh]

How different do you think it would be from a Dataflyer 4000SX?

BTW: My Dataflyer 4000SX sucks and I am pretty sure it's the same across the dataflyer range.

[papa_november]

The 4000SX is just a rebadged DataFlyer Plus, so it should be exactly the same.

It's a pretty slow controller, but at least it works. That and the design is amazingly simple- it's mostly made up of commonly available chips and only has two PALs.

Sucks that there's no ATAPI drivers for the IDE version, though.

[alexh]

Is there any way to dump the ROM in software?

If not then you'll have to wait a few weeks until I can be arsed to unscrew the A4000 and take the card to work to be read on the EPROM programmer.

[papa_november]

Roms on expansion boards are generally not dumpable in software. It makes trying to get the latest ROMs a humongous pain in the ass.

[joeksterz]

Was looking for one as well, literally for years. Picked up an A2000 just the other day and it's dead from battery damage, but had a DataFlyer plus inside. And guess what, an Autoboot 2.1 ROM. Pulled it out, made a copy with my EPROM programmer, and it works great in my other DataFlyer card.

The factory rom was burnt on a 27c256-12 chip although the contents look doubled, so you might be able to burn onto a 27c128. Not sure how fast of a chip you'll need, but usually 150ns works fine (obviously a 120ns chip works)

[emufan]

thanks for sharing - i think it will make it's way into winuae soon

[Toni Wilen]

Thanks. Yes, very soon

Could you also include older version if you have it? Also could you attach output from this program: http://www.winuae.net/files/b/expdump.zip with and without autoboot jumper set.

ROM is confusing, autoconfig data is at the beginning of ROM (which is not unusual) but most expansions with this kind of setup have same ROM image twice with second "bank" having autoconfig diagrom flag cleared but attached ROM contains 100% identical data twice. (Autoconfig jumper can't disable whole ROM because it would break autoconfig. It is only disabled in boards that have autoconfig data inside PAL/GAL chips). Another possibility is that ROM code checks jumper state. Above test program should show some difference between autoconfig on and off.

(btw, software dumping expansion rom contents is usually very easy)

[mark_k]

After taking one half of the 32KB file (since both halves are identical), notice that all nybbles seem to be doubled, i.e. $4E75 (RTS opcode) is $44EE7755.

After processing the file down to an 8KB image (attached), if you look at offset $1000 there is what looks like another set of autoconfig data.

[Toni Wilen]

I should have noticed too obvious "Cx1x" = 64k Z2 without boot rom, "Dx1x" = 64k Z2 with boot rom "identifiers". I just skipped that part and jumped to mid point of ROM file but actual ROM size is only 8k. Duplicate 16k at the end still does not make any sense, unless it was part of the original rom build process for some unknown reason.

No need for expdump outputs anymore

[Toni Wilen]

Emulation implemented, both SCSI and IDE parts.

This is quite different than most other HD controllers. Boot ROM only contains simple and tiny loader ("Caprion_AutoBoot", whatever that means) that loads full driver from RDB blocks, installed by Dataflyer's installer. In other words, it can't autoboot if drive is not initialized with Dataflyer's installer.

[joeksterz]

Attached is the 1.5 ROM that I have. I *think* I might have a 1.7 ROM on another card somewhere, will see if I can't dig it out.

[joeksterz]

And just found the 1.7 ROM (attached) - that's all I have.

Toni: Saw you added support in the latest beta - Thanks! You may want to mention that booting without the "special" driver loaded on the hard drive takes literally a minute or two on real hardware (and assuming that it'll behave the same emulated). Someone might incorrectly think that WinUAE has hung.

[Toni Wilen]

Originally Posted by joeksterz

And just found the 1.7 ROM (attached) - that's all I have.

Toni: Saw you added support in the latest beta - Thanks! You may want to mention that booting without the "special" driver loaded on the hard drive takes literally a minute or two on real hardware (and assuming that it'll behave the same emulated). Someone might incorrectly think that WinUAE has hung.

Thanks.

Yeah, it has very long delay which imho is quite stupid, it just repeats many many times drives & RDB embedded driver detection until it finally decides that there really is no Dataflyer installed drives..

Manual says that driver on disk location was changed in v2.0+. I am not sure what happens when ROM is v2.0 and driver is older or vice versa. This is too weird expansion, even SCSI and IDE have separate drivers and both aren't installed if you don't have both types of drives connected when initializing drives..

EDIT: v1.5 and v1.7 ROMs do work with v2.1 RDB installed driver. ("ExpSys.device (Device Driver, Wayne E. Miller ,(c) 1988-1992")

[TjLaZer]

Does anyone know if this ROM works on the IDE version of the Dataflyer card?

[Toni Wilen]

Originally Posted by TjLaZer

Does anyone know if this ROM works on the IDE version of the Dataflyer card?

I assume it does because IDE and SCSI drivers are separate and only physical difference between boards is installed components.