ST536 STE EDITION

[exxos]

I think it's a ground. Is it connected 'outward' or 'inward', @exxos? It's the inward half that's lifted so may have to bodge wire it if the trace goes that way.

Capture.PNG

2.PNG

[Badwolf]

OK, 288 chip is on.

Firmware 9 giving a 404. Firmware 8 crashing under EmuTOS. Gets to the desktop with TOS, but doesn't autoboot my disc (that may be my config, though).

Went back to Firmware 4: detects TT-RAM but doesn't appear to the system (not sure why -- perhaps failing a more thorough RAM test?). ST-RAM access seems pretty stable. Lots of aliasing-type errors when doing a blind TT-RAM region test.

What firmware should I be using and do I need a clock cap again?

BW

[exxos]

Firmware 9 giving a 404. Firmware 8 crashing under EmuTOS. Gets to the desktop with TOS, but doesn't autoboot my disc (that may be my config, though).

I've not finished firmware9 yet.. You need to use CURRENT8. I've been using my patched 206 for testing currently. So can't say if any errors with EMUTOS yet... will give it a try shortly..

Yes you need that cap, needs to be 33pF or 47pF ideally.

Steves had CURRENT7, I tested it a lot. BUT I found those odd IDE issues after, and not sure if TOS206 blitter stuff was fixed at that point. fast-tos wasn't in that code either yet.

I was fighting this odd corruption ? viewtopic.php?p=128802#p128802 MAybe you mean that ? Seems to stem from when the IDE card is inserted.. *SHOULD* be fixed in current9.. but I only just started testing it..

IDE AB may malfunction on earlier versions...

[exxos]

@Badwolf current9 is up.

You don't need TOSCOPY for it to work (you just wont get fast-rom enabled)

Boots to my 206 and EMUTOS in a quick test...

IMG_3261.JPG

IMG_3260.JPG

Mostly been working with 206 in fixing the blitter ROM issue from the previous firmware.

Will leave it on test all night in 206 doing TTram test to make sure I've not broke anything. Will do STram test in the morning and I'll try copying files around the IDE partitions after that..

[exxos]

Those RAM tests have been passing fine.. Running a lot slower than normal because of the cache issues.

IMG_3263.JPG

IMG_3262.JPG

I tried turning the caches on/off in GB6 but while they seem to toggle fine. I have to turn the blitter off to see any difference in tests. Understandable I suppose.

Desktop menu cache is definitely stuck on off after loading BLTFIX. I loaded SI.CPX , though it wouldn't do anything and now the desktop menu is allowing me to turn the cache on/off again :WTF:

Even more odd is SI.CPX results in the blitter being turned off in GB6.

Then testing TTram speed, it shows a more understandable 826%. (was about 711% before, should be about 833%).
Turned both caches off, TTram speed 767%. So caches working
Turned caches on, back to 826% as expected.
Turned blitter on (GB6) now TTram drops to 723% :WTF:
Turned blitter off (GB6) now TTram still stuck on 723% :WTF:
Toggled the cache off/on in SI.CPX but made no odds.

Dropping back to desktop, the cache is OFF again! Can't turned it on again!! So toggled in SI.CPX, now I can enable the cache in the desktop menu again. TTram speed back up to 826% again !

Its like the cache setting in the desktop menu is doing something else rather than simply turning on/off the caches for the CPU :shrug:

I think this is beyond my pay grade!! :lol: :roll:

[exxos]

I found this little blitter check program in floppy shop PDL.. Useful as the blitter is hidden in the desktop menu :( Really need to figure out how to get that enabled again somehow...

Basically when the blitter is enabled the cache in the desktop menu is forced off. So this must be something TOS is doing. It could make sense, as blitting would invalidate the caches anyway. But with BLTFIX, that problem is solved.. BUT TOS doesn't know about it..

IMG_3265.JPG

@Icky then suppose you can remember anything about the desktop menu stuff when you was fiddling with it a couple years ago ? I cannot even figure out how the menu is generated at this point never mind the blitter menu enabled again!

[agranlund]

The menu option has the same internal ID for blitter and cache control if I remember the TOS206 source code correctly - since there was no machine with both they just reused the same option.
If this means they reused some internal variables(s) too I don't know.

Perhaps it's related to that, and something in TOS206 is going confused.

[exxos]

The menu option has the same internal ID for blitter and cache control if I remember the TOS206 source code correctly - since there was no machine with both they just reused the same option.
If this means they reused some internal variables(s) too I don't know.

Perhaps it's related to that, and something in TOS206 is going confused.

Possibly yes.. I am trying to figure out the blitter and cache relationship.. only found this so far in desktop.c .

Capture.PNG

If I could at least get blitter and cache in the menu then it would be a start. Can't even figure out how the blitter or cache is even enabled or disabled in the first place.


Looks like its either cache or blitter but not both :roll:

Capture.PNG

So maybe the variables were reused somewhere. Cache state opposite of blitter state :roll:

[Steve]

Perhaps it makes more sense to be targetting EmuTOS exclusively for this accelerator? I don't know anyone who runs stock-TOS anymore especially on accelerated machines. At least with EmuTOS there's an active team who work on it, with 2.06 you're relying on ghosts and source code. Any argument people might have around compatibility doesn't hold water due to the fact that an 030 with fastram breaks most old software anyway. (I'm not saying give up on 2.06, but why not focus on EmuTOS 1st then TOS 2nd)

[exxos]

@agranlund If BLTFIX has a branch to use blitter or software blits, maybe the problem could be "solved" if BLTFIX enabled the caches for soft blits, and disabled for blitter ?

It would likely dynamically solve the problem.. then ignore what's in the desktop menu..

Only hiccup is @Badwolf was wanting a register to disable the cache etc.. So not sure yet if CDIS can override the cache register settings ?

BUT IIRC everything in the 536 code is cache inhibited other than TTram access amway.. but not sure if that works as turning the caches on in GB6 make stuff go faster anyway, like its still caching something STram related :shrug: