Testers needed

All 680x0 related coding posts in this section please.

Moderators: simonsunnyboy, Mug UK, Zorro 2, Moderator Team

User avatar
Foxie
Captain Atari
Captain Atari
Posts: 354
Joined: Wed Feb 03, 2016 7:12 pm

Testers needed

Postby Foxie » Thu Feb 04, 2016 2:40 am

Greetings, ST users ^.^

New to the forum, not new to the ST (got my first 520 in 1989).

I'm returning to coding for the ST... it's been a while, I'm a bit rusty. What would really help is if someone could try some routines on their hardware. I've got a test program here:

https://www.dropbox.com/s/9ggmi6sd920c0 ... r.zip?dl=0
(it needs all three files in the same folder to run, it takes a while to load after the pic appears)

Hopefully it works on all ST/STE/Falcon systems (except VGA).

I'm interested in how it behaves on a range of systems -

1) Does it bomb out? For some reason it's bombing on Steem but not Hatari.
2) Is the sound quality good and noise-free? (It sounds like a scratchy 78 record on Hatari)
3) Do the raster bars in the side borders remain stable even when you wiggle the mouse? I think I finally nailed that one...

Click the mouse to exit (should exit cleanly and restore desktop resolution)

Any help is very much appreciated ^.^

AtariZoll
Fuji Shaped Bastard
Fuji Shaped Bastard
Posts: 2978
Joined: Mon Feb 20, 2012 4:42 pm
Contact:

Re: Testers needed

Postby AtariZoll » Thu Feb 04, 2016 7:41 am

Hi and welcome to the forum. It's always pleasure to see someone doing creative work here .
Unfortunately, my testing of your player SW was pretty much short. I tried it with Steem Debugger - and first and most important what I can say to you is that use it for your Atari SW development - it is really well made SW.
Player crashed immediately after start - illegal instruction at $F368 (code start at $F196 - TOS 1.04) . So, there is error in assembling or compiling it. Really no point to try in real HW before you make it work in Steem. I use Steem Debugger a lot, and it is very reliable and accurate emulator+debugging, tracing SW - 99% Atari ST(E) programs work well in it. And is still updated by Steven Seagal. For development is ideal, really.
Famous Schrodinger's cat hypothetical experiment says that cat is dead or alive until we open box and see condition of poor animal, which deserved better logic. Cat is always in some certain state - regardless from is observer able or not to see what the state is.

User avatar
LaceySnr
Obsessive compulsive Atari behavior
Obsessive compulsive Atari behavior
Posts: 120
Joined: Wed Jun 26, 2013 5:00 am
Contact:

Re: Testers needed

Postby LaceySnr » Thu Feb 04, 2016 10:55 am

I gave this a quick spin on my Falcon (despite reading AtariZoll's post) just to see if it'd help provide any more information. I got a monochrome screen with a bunch of garbage, then hitting the space bar tried to take me back to GEM but failed to restore the video mode correctly. Saw two bombs though which I think indicates a bus error.

Let us know when you've got another build to try!

User avatar
Foxie
Captain Atari
Captain Atari
Posts: 354
Joined: Wed Feb 03, 2016 7:12 pm

Re: Testers needed

Postby Foxie » Fri Feb 05, 2016 3:06 am

Thanks for the help! I've been doing some debugging, I've got a new version that stands a better chance of working on the STE, but I'm still puzzled why it crashes on the Falcon:

https://www.dropbox.com/s/tkc3hnw1t5pnu ... 3.zip?dl=0

LaceySnr - thanks for testing it anyway! I doubt this new version will be worth testing - I didn't change anything Falcon-related. Did you try testing with a VGA monitor? It should detect VGA, print an error and exit cleanly, but I get the feeling something is badly wrong. Difficult to actually debug without a Falcon!

I've found what was causing it to crash on Steem - apparently I'd accidentally left the CPU type in Devpac set to 68020, and then I used cmpi.w #...,variable(pc). I've removed the (pc) and it's working fine now.

What's puzzling me is this: I could have sworn program counter relative addressing was allowed for cmpi, even on the base 68000. Hatari also seems to think so! A couple of 68000 instruction references I've got to hand seem to suggest it's valid, so I'm puzzled. It can't be a bug in Steem because Devpac also disallows this mode on the 68000.

User avatar
LaceySnr
Obsessive compulsive Atari behavior
Obsessive compulsive Atari behavior
Posts: 120
Joined: Wed Jun 26, 2013 5:00 am
Contact:

Re: Testers needed

Postby LaceySnr » Fri Feb 05, 2016 10:10 pm

Ha! I tried it in Hatari, and the image worked but the sound didn't, so I was saved ;)

User avatar
AtariCrypt
Captain Atari
Captain Atari
Posts: 359
Joined: Fri Mar 14, 2014 5:04 pm
Location: Lancashire, England
Contact:

Re: Testers needed

Postby AtariCrypt » Fri Feb 05, 2016 10:37 pm

Hi there :)
I also see an image.. but no sound. then after about 20 seconds (?) or so I hear something briefly and then I get 4 bombs.
Using a Mega STe at 16mhz / TOS 2.06

User avatar
Foxie
Captain Atari
Captain Atari
Posts: 354
Joined: Wed Feb 03, 2016 7:12 pm

Re: Testers needed

Postby Foxie » Sun Feb 07, 2016 2:58 am

Thanks for testing, I tested in Hatari at 16MHz, and much to my surprise it crashed also!

I finally pinned down the problem - it depended which order two simultaneous interrupts got executed, and it was such a corner case the CPU speed actually determined the outcome.

New version here -

https://www.dropbox.com/s/ml6x3cazgvepx ... 4.zip?dl=0

I haven't done much work on the Falcon side, although it now produces sound in Hatari's Falcon emulation mode. Sound is wrong though, it's pitch shifted upward. I added some printouts to show the current detected machine etc.

Maybe one day I'll actually succeed in getting this thing to run half reliably!

AtariZoll
Fuji Shaped Bastard
Fuji Shaped Bastard
Posts: 2978
Joined: Mon Feb 20, 2012 4:42 pm
Contact:

Re: Testers needed

Postby AtariZoll » Sun Feb 07, 2016 9:03 am

Foxie wrote:Thanks for the help! I've been doing some debugging, I've got a new version that stands a better chance of working on the STE, but I'm still puzzled why it crashes on the Falcon:
...
I've found what was causing it to crash on Steem - apparently I'd accidentally left the CPU type in Devpac set to 68020, and then I used cmpi.w #...,variable(pc). I've removed the (pc) and it's working fine now.
What's puzzling me is this: I could have sworn program counter relative addressing was allowed for cmpi, even on the base 68000. Hatari also seems to think so! A couple of 68000 instruction references I've got to hand seem to suggest it's valid, so I'm puzzled. It can't be a bug in Steem because Devpac also disallows this mode on the 68000.

Just as I expected - assembling error. Another problem is that Hatari's CPU emulation had/has some flaws - there was thread about it here, and I even made some SW for detecting which instructions work, and which cause illegal on diverse configs, CPUs and in emulators. Considering this case - cmp with PC relative (+other PC relative addressings), Steem is accurate. And Hatari should be updated too in meantime - do you using latest version ?
Generally, PC relative is not valid for target and in case of single operand - like tst on 68000. I did not recommend Steem Debugger because want to play smart - that's based on thousands of hours spent with it.
Famous Schrodinger's cat hypothetical experiment says that cat is dead or alive until we open box and see condition of poor animal, which deserved better logic. Cat is always in some certain state - regardless from is observer able or not to see what the state is.

User avatar
Foxie
Captain Atari
Captain Atari
Posts: 354
Joined: Wed Feb 03, 2016 7:12 pm

Re: Testers needed

Postby Foxie » Wed Feb 10, 2016 8:34 pm

After some testing in Steem, I have what I hope to be the final version - at least as far as the ST goes. It's still a bit of a mystery why it didn't work on the Falcon.

STFM version:

https://www.dropbox.com/s/j6zkhey2m800d ... 5.zip?dl=0

I've also written code for STE DMA sound, which has a separate test program:

https://www.dropbox.com/s/lygjiszc3z70m ... e.zip?dl=0

For some reason in Hatari's Falcon emulation mode, the sound is terrible. It works fine in STE mode, and also on Steem. It would be interesting to know if it also runs into problems on a real Falcon (RGB only, not VGA). I'm not touching any of the Falcon-specific registers for sound so I don't know why it sounds so bad.

User avatar
AtariCrypt
Captain Atari
Captain Atari
Posts: 359
Joined: Fri Mar 14, 2014 5:04 pm
Location: Lancashire, England
Contact:

Re: Testers needed

Postby AtariCrypt » Wed Feb 10, 2016 10:12 pm

Works on my Mega STe (obviously downloaded the DMA version)
There was a little wobble from the raster bars at the very start, but moving the mouse later didn't make them move at all. Sound quality appears nice but I wasn't able to have it loud due to the little one being in bed! Pressing the right mouse button produces a load of code on the screen that scrolls upwards?!! Pressing left mouse quits...

Steve

User avatar
Foxie
Captain Atari
Captain Atari
Posts: 354
Joined: Wed Feb 03, 2016 7:12 pm

Re: Testers needed

Postby Foxie » Sun Feb 14, 2016 9:25 pm

Splendid, thanks for testing! A bit of wobble while it's still loading is kind of expected, as long as it's stable once it gets going.

The right mouse button prints my bitcoin private key; it's a well-hidden feature that I'm sure nobody will find ^.^

User avatar
AtariCrypt
Captain Atari
Captain Atari
Posts: 359
Joined: Fri Mar 14, 2014 5:04 pm
Location: Lancashire, England
Contact:

Re: Testers needed

Postby AtariCrypt » Sun Feb 14, 2016 9:27 pm

NP :cheers:


Social Media

     

Return to “680x0”

Who is online

Users browsing this forum: No registered users and 2 guests