REV 3 - REV 5 - The beginning (ST536)

All about the ST536 030 ST booster.
User avatar
exxos
Site Admin
Site Admin
Posts: 29709
Joined: Wed Aug 16, 2017 11:19 pm
Location: UK
Contact:

Re: REV 3 - REV 5 - The beginning (ST536)

Post by exxos »

Badwolf wrote: Mon Nov 09, 2026 10:58 am Personally I would not trade a changed reset behaviour for cached ROM. We have a cached ROM mechanism already with MAPROM.
MAPROM isn't supported and nobody knows how it works. It's all built into TOS now anyway. MAPROM isn't needed anymore, that was the point.
I don't use that as I don't want cached ROM anyway. If this is an option, then fair enough, but if it's hard wired into the firmware then I think it certainly would annoy some people. Me people! :D
Why not use fast-rom, your just making GEM painfully slow to use for no good reason.
User avatar
Badwolf
Site sponsor
Site sponsor
Posts: 3172
Joined: Tue Nov 19, 2019 12:09 pm

Re: REV 3 - REV 5 - The beginning (ST536)

Post by Badwolf »

exxos wrote: Mon Nov 09, 2026 11:04 am Why not use fast-rom, your just making GEM painfully slow to use for no good reason.
90% of the time if I'm using the desktop the VDI and AES are soft-loaded replacements anyway. I don't want things going on behind the scenes that I don't know about or can't override as I'm usually looking for my own bugs anyway. Introducing more factors into the mix is actively unhelpful.

I'm really after any addons I install to be the lightest touch possible to do their job. If it's a 50MHz 030 with TT-RAM, great! I'll happily run some extra software to get the best out of it -- including declaring the RAM. I don't really want all the bells and whistles on by default as so little is compatible with all the bells and whistles that I like to choose the ones I want at boot time. The ones that *are* compatible with the bells and whistles will probably be run from the desktop after the drivers will have been loaded.

It's why my rev2 DSTB1 went the other way: everything is off until software turns it on. It does actually have a ROM-in-RAM mode (it can even load that ROM from disc), but you have to intentionally turn it on.

BW
DFB1 Open source 50MHz 030 and TT-RAM accelerator for the Falcon
Smalliermouse ST-optimised USB mouse adapter based on SmallyMouse2
FrontBench The Frontier: Elite 2 intro as a benchmark
User avatar
exxos
Site Admin
Site Admin
Posts: 29709
Joined: Wed Aug 16, 2017 11:19 pm
Location: UK
Contact:

Re: REV 3 - REV 5 - The beginning (ST536)

Post by exxos »

Running TOS from TT-RAM is no more “behind the scenes” than loading a replacement desktop into TT-RAM.
User avatar
Badwolf
Site sponsor
Site sponsor
Posts: 3172
Joined: Tue Nov 19, 2019 12:09 pm

Re: REV 3 - REV 5 - The beginning (ST536)

Post by Badwolf »

exxos wrote: Mon Nov 09, 2026 11:22 am Running TOS from TT-RAM is no more “behind the scenes” than loading a replacement desktop into TT-RAM.
Yeah sure. If it's softloaded by the user into GEMDOS memory pool.

If it's built into the hardware with transparent remapping of the memory and no off switch, then I don't agree.

BW
DFB1 Open source 50MHz 030 and TT-RAM accelerator for the Falcon
Smalliermouse ST-optimised USB mouse adapter based on SmallyMouse2
FrontBench The Frontier: Elite 2 intro as a benchmark
User avatar
exxos
Site Admin
Site Admin
Posts: 29709
Joined: Wed Aug 16, 2017 11:19 pm
Location: UK
Contact:

Re: REV 3 - REV 5 - The beginning (ST536)

Post by exxos »

Badwolf wrote: Mon Nov 09, 2026 11:46 am Yeah sure. If it's softloaded by the user into GEMDOS memory pool.

If it's built into the hardware with transparent remapping of the memory and no off switch, then I don't agree.
I understand your point, but that is really your own use case as a developer, not how the vast majority will use it. You said yourself you have very little time for anything these days, so I doubt you would be developing for the ST536 at that level, and I seriously doubt anyone else would either. So it is a bit of a moot point really.

As for it being built into the hardware, it is not the hardware doing it by itself. TOS copies itself into TT-RAM at boot and then tells the PLD to switch over, so it is effectively soft loaded by TOS, just at boot rather than from the desktop. I actually wrote a module which turns off TT-RAM and fast ROM from software, but I disabled it because I really did not see the point.

EmuTOS does not have the copy built in, so it never triggers fast ROM at all. As you tend to use EmuTOS, it already acts the way you want anyway. EmuTOS would need a program to switch fast ROM on, which I have not written yet.

Even from a development point of view, I don't see any use case where TOS running fast has any downside at all. During development there are endless reboots, and having TOS running fast is a huge time saver. Having it running slow while developing would absolutely drive me nuts!

Most people just want to plug it in and have it work, with no software to install or configure. That is exactly why I built fast ROM into TOS and got rid of the need for MAPROM altogether, as it kept things a lot simpler for everyone.

MAPROM does a bunch of strange things by itself, and even Anders himself doesn't really know how it works any more. It is not supported, nobody knows how it works, and nobody is going to maintain it, that much is clear. Running MAPROM adds a lot more layers of complication and issues than simply running TOS fast, and people using it may well run into compatibility issues because of the tricks it does to speed up the system. So I bypassed the whole issue by not requiring it at all, and not using it has far more advantages than using it.

As for the warm reset, I would argue a cold boot every time is an enormous advantage. I have seen plenty of times where switching between TOS and EmuTOS does not play nicely unless a cold boot happens. The user flips the ROM jumper with the machine powered on, presses reset, and one or the other screws up because memory has not been reset properly, which then leaves them confused and frustrated because it has all gone nuts.

That is exactly why I made Flashy Clock do a cold boot, as switching operating systems can corrupt the other one otherwise. With the ST536 it also means that after a reset the ROM copy is taken fresh from the ROM chips, so a ROM bank change takes effect straight away instead of the old TOS carrying on from TT-RAM. Of course nobody should change jumpers with the power on, but I find myself doing it repeatedly and no doubt other people will do exactly the same. A cold boot on every reset rules out a flood of weird issues and complaints before they happen, simply because people do not power down for 30 seconds after changing the ROM bank. It only takes a few seconds longer anyway.

The whole point of an accelerator is to go fast, and deliberately having something run slow defeats the object of having an accelerator in the first place. I am not going to make 99% of users put up with GEM moving like it is loading from a C15 tape, or force them to run a bunch of programs just to enable features. That just gets way too confusing for people, and every extra program adds another layer of confusion and its own issues on top.
User avatar
JezC
Posts: 2902
Joined: Mon Aug 28, 2017 11:44 pm

Re: REV 3 - REV 5 - The beginning (ST536)

Post by JezC »

Would an option be to provide a program that can be run to disable the TOS in TT RAM for those who feel the need for it? This could be done after the new designs are released even

Sounds like you were already developing something like this before, so is it much work for someone (else) to finish it?

From my perspective I think your default choice is sensible but I too can see some cases where I might need to disable it temporarily for certain applications...

Just having a quick think about how this new ST536 will compare to the original ones I have already too ...
User avatar
stephen_usher
Site sponsor
Site sponsor
Posts: 7690
Joined: Mon Nov 13, 2017 7:19 pm
Location: Oxford, UK.
Contact:

Re: REV 3 - REV 5 - The beginning (ST536)

Post by stephen_usher »

exxos wrote: Mon Nov 09, 2026 11:22 am Running TOS from TT-RAM is no more “behind the scenes” than loading a replacement desktop into TT-RAM.
If it's RAM that the CPU can't access and you're not using 68030 relocation tables then I would agree. i.e. if the hardware is redirecting ROM address reads to a read-only section of high RAM which isn't addressable by the CPU otherwise.

If it's using 68030 relocation tables (as MAPROM does) then any software which changes the tables *WILL* break things.
Intro retro computers since before they were retro...
ZX81->Spectrum->Memotech MTX->Sinclair QL->520STM->BBC Micro->TT030->PCs & Sun Workstations.
Added code to the MiNT kernel (still there the last time I checked) + put together MiNTOS.
Collection now with added Macs, Amigas, Suns and Acorns.
User avatar
Badwolf
Site sponsor
Site sponsor
Posts: 3172
Joined: Tue Nov 19, 2019 12:09 pm

Re: REV 3 - REV 5 - The beginning (ST536)

Post by Badwolf »

exxos wrote: Mon Nov 09, 2026 12:01 pm
Badwolf wrote: Mon Nov 09, 2026 11:46 am Yeah sure. If it's softloaded by the user into GEMDOS memory pool.

If it's built into the hardware with transparent remapping of the memory and no off switch, then I don't agree.
I understand your point, but that is really your own use case as a developer, not how the vast majority will use it.
Well I still tend to think "most people" exposed to an accelerator that turns on everything with no input from the user will end up just going 'this, that, the other doesn't work with ST536...' and so it'll be a reputation for it being massively incompatible.

OTOH, if you can selectively turn things on if you want them, that incompatibility level will drop.

It's just my opinion at the end of the day, but if I were in the market it would probably be enough to put me off buying one. If your judgement is that mine's a minority view, that's fair enough. But I think there's no harm in having the contrary position kicked around.

Mind you, if not having it enabled is just a case of popping out the ROM then I suppose it could be a non-issue. That said when I last ran into debugging problems with the 536 there wasn't an easy way to avoid using the fast-access ROM either.

Cheers,

BW
DFB1 Open source 50MHz 030 and TT-RAM accelerator for the Falcon
Smalliermouse ST-optimised USB mouse adapter based on SmallyMouse2
FrontBench The Frontier: Elite 2 intro as a benchmark
User avatar
exxos
Site Admin
Site Admin
Posts: 29709
Joined: Wed Aug 16, 2017 11:19 pm
Location: UK
Contact:

Re: REV 3 - REV 5 - The beginning (ST536)

Post by exxos »

JezC wrote:Would an option be to provide a program that can be run to disable the TOS in TT RAM for those who feel the need for it?
I did write a module for it, but I disabled it. Every extra feature in the PLD adds more work and hassle, and as we have seen many times, even small changes upset the SDRAM timings. That is the main reason it went. Also, people would still have to boot to the desktop and run a program to turn it off, by which point TOS has already booted and they are sitting at the desktop in fast mode anyway. No doubt people would then complain that turning it off at that point is too late, so it seems a whole lot of hassle for basically no valid reason.

If you use EmuTOS, fast ROM is never switched on in the first place, as EmuTOS does not have the copy built in.

Features can always be added later if enough people actually want them. But I am not going to develop and spend time on stuff which I would assume hardly anyone will use, on a product where I would assume hardly any will sell in the first place. If people want me to spend more time on stuff, then things have to be more worth me spending the time on. Unfortunately my spare time does not come with an infinite lives poke.
stephen_usher wrote:If it's RAM that the CPU can't access and you're not using 68030 relocation tables then I would agree.
No MMU tables are involved at all. The PLD simply redirects the ROM reads in hardware to a reserved area at the top of the SDRAM, above the TT-RAM that TOS hands out, so nothing gets allocated there. Software changing the MMU tables cannot break it, because it does not use them. That is one of the reasons I did it this way instead of relying on MAPROM.
Badwolf wrote:Well I still tend to think "most people" exposed to an accelerator that turns on everything with no input from the user will end up just going 'this, that, the other doesn't work with ST536...'
Incompatible with what exactly though? All fast ROM does is make GEM and TOS run a bit faster, and I cannot think of a single reason why that would cause anyone a problem. If it was running at 1GHz then maybe, but a few hundred percent faster is very unlikely to upset anything.

I would argue that running a 030 is the most incompatible layer in the first place. The caches, self-modifying code, different exception stack frames and timing loops written for a 68000 will all break far more software than TOS running a bit faster ever will. That comes with any 030 accelerator, fast ROM or not. If people want it totally compatible, then they should not be running a 030 at all. You cannot have it all ways.

Fast ROM has been standard in my stuff for about two decades now, and in all that time nobody has ever complained about TOS running faster. The same goes for the PAK030, which ran TOS from 32-bit ROMs on the board, easily twice as fast as stock, and I never heard a single complaint about that either.

I honestly do not see how the general user looking for a plug and play accelerator would want the acceleration turned off by default, and then have to depend on software to enable it. By the time that software runs, they have already had to sit through the extra wait to get to the desktop in slow mode. Personally I think that is just messy, when we can accelerate the whole boot process simply by having it on by default. It would be like fitting a turbo to a Sinclair C5 that only kicks in once you have parked.

In fact, Frank Lucas was "informing" us that the PAK booted faster, because the PAK had fast ROM turned on by default. We went from about 4 minutes to about 30 seconds to match the PAK. So there has already been one person "complaining" about it NOT having fast ROM on at boot.

Like I keep saying, you cannot have it all ways by default. People also need to realise that whether I do option A or option B, 50% of people will likely complain either way. So I do things which benefit my own workflow and simply make sense, and my workflow runs a lot better with things running fast by default, fast ROM included.

People looking for a plug and play experience just want to plug it in and have everything work out of the box. I always aim to make things as easy and simple as possible, and in my opinion at least, that is the way that makes sense, as it is how I use everything myself. Having to run stuff from AUTO just makes things more complicated and more hassle. Think about games or anything else which autoboots from floppy. With everything off by default, you would get no acceleration at all from the booster, because the "switch it on" programs never got loaded. It all just becomes a huge nightmare.

Sorry if this upsets some people, but this was all decided over a year ago and nobody said a word at the time. It is a bit late now to start complaining, after I have put tons of work into making it what I think is the best and simplest way for everybody to use it, without needing annoying software to load.

I may consider adding a proper off option at some point, but first I would like to see some mainstream proof of software people actually use, where TOS running faster causes a provable problem by itself. I do not recall anyone ever complaining about the PAK030 or any other booster doing the same thing over the past 30 years either. Maybe those scenarios do exist, but obviously not mainstream enough for anyone to have ever mentioned it, in my recollection anyway.
Badwolf wrote:Mind you, if not having it enabled is just a case of popping out the ROM then I suppose it could be a non-issue. That said when I last ran into debugging problems with the 536 there wasn't an easy way to avoid using the fast-access ROM either.
Yes, like I said, EmuTOS will not enable fast ROM, so a ROM without the copy patch simply never turns it on. It can also be turned off in the firmware if a developer wants that, so that is not a problem either. What I am thinking about is the end user's install, their expectations and their experience. Knowing the hassle with multiple versions of MAPROM, which is still largely unproven in a lot of scenarios, simply not having it running at all saves me a huge support nightmare on its own.
coonsgm
Posts: 464
Joined: Sat Jan 30, 2021 1:30 am

Re: REV 3 - REV 5 - The beginning (ST536)

Post by coonsgm »

All three TT Ram tests freeze.
The desktop is clearly accelerated with the rapid animations suggesting that TOS on RAM is working

I will continue to see if I can have find any hardware issues (bad soldering by myself).

Thank you again for all the help!
Post Reply

Return to “ST536 030 ST ACCELERATOR”