I had a vacation so not that much coding (Legoland. YAAY)
But I have been working on a lot of internal stuff and the biggest issue for sure is handling of keyboard. a machine with no kickstart whatsoever loaded behaves very different then a machine with kickstart started. So working on getting it to work but programming new roms every try is tiresom, and UAE does not really handle keyboard as the real hardware.
I guess it WOULD be more easy if I did use IRQs etc, but as I want this software to run on very defective machines, this is not an option. (as now, the rom starts with paula removed, this will not be possible then as an example)
Anyway, when I got that issue solved I will release that version as next stable.
8 thoughts on “Soon next stable version should be ready”
Did some testing of the build dated 20/7.
Found some issues:
– On A500 (i.e. 68000) getting “illegal instruction” errors on test: audio – simple waveform, graphictests (after it says not implemented, on other non implemented screens as well)
00000000 00000001 00011956 00000000 00000002 00000000 00000003 0000031
00F8690B 00002628 FFFFFFFC 00000000 00011956 00002404 00002058
(the 0002058 value is counting down until bad crash 🙂
– (Chipmem) tests seem to not work, get flashy screen and after a while it says Bad block start at $000003FF and ends at $00000403 with a size of 0KB (screen still flashy)
Both can be reproduced by setting WinUAE to A500 config. Update: suddenly without changing WinUAE settings the memtest works again .. 🙁
– Keyboard handling on A1200 is better
– IRQ test on A1200 takes really long then says IRQ1 LEVEL: CANCELLED then crash
– on real hardware, the dots line on press LMB/RMB boot screen and colorcycling doesnt happen, goes to menu immediately
This is it for now.. Thanks for the good work! 🙂
If possible I would like some features that are in Keir’s systest disk but probably coding those in asm might be hard.
ok I have no clue what I did but ONE TIME all tests started working on 1200. Chipmem, IRQ etc.
But power cycle and again same issues .. flashy copper bars at every test/menu item..
Haven’t been able to reproduce what I did, it’s been running for 2 hours trying various stuff.. no go.
illegal instruction.. interesting.. well, then I atleast know that my errorhandling seems to work. is this on a working machine?
the hexnumbers are actually dumps of all registers. D0-D7, A0-D7, SR.
flashy screen comes when a user pressed a button, and not released it yet, waiting for the button to be released.
Anyway.. thanx for input. will test on my labmachines here.
there will be more tests and more advanced ones. I guess I need to sort out bugs first 🙂
I hope you have one of these or something similar.
they even show the usage of your diag rom in one of their vids.
the combination of your diagrom and their flash is no doubt very useful (though only for A500/2000 as they haven’t made a socket for gary for A600)
That one was really interesting…
It’s been awhile since I played with ROM programming (um…maybe 25 years?) and back then I somehow recall only working with PIC ROMs for GVP SCSI & 030 boards. We gang-programmed 8 FastROM v3/v4 at a time when we were short of pre-programmed ROMs from production sources, or a new update came from Ralph Babel. I am curious to locate a burner and get my hands on some EPROMs and try a few things for the Amigas and the C64/128/disk drives I have.
Can you add to the DiagROM ReadMe the EPROM/ROM type you know works directly in the 16-bit A500/2000 machines? I see you have mentioned the ROM type expected for the 32-bit machines (27C400-105). I vaguely remember there being different types (pinout different?) at one point when KS 2.04 came out for the A500/2000.
I use 27C400-105 even on 16bit machines. I have read that SOME revisions of the 500 needs some pins to be swapped. actually do not know what versions need that and what pins.
on rev8 A500 and A2000/600 27C400 works fine anyway.
I’ve got some parts ordered, and a burner, so I shall see. Rev 6A A500 here. Thanks.