@Jon: By running the game from the command line, I get some curious info:
First I get:
https://www.dropbox.com/s/15u3jzq1jkxrg ... 7.jpg?dl=0
Then SHOWREGS says:
https://www.dropbox.com/s/uu51l7wkmitwx ... 6.jpg?dl=0
ADFFS 2.61 beta
Re: ADFFS 2.61 beta
Interesting...there's a bug in the code at that address which will cause Pi2 onward to report an illegal instruction.
As far as I can tell, it's the only occurrence. Add the following to !Run and see if it then works:
JITMEMORYA 12234 E96D400F E92D400F
More to the point, why doesn't it crash my Pi3? It should. I might see if I can automate fixups of this instruction as I've come across a few games that use it, such as Zarch. Although it's a bug in the code, up to the Pi1 it will be executed with the intended outcome, Pi2 onward will report it as illegal (except in my case it would appear!)
As far as I can tell, it's the only occurrence. Add the following to !Run and see if it then works:
JITMEMORYA 12234 E96D400F E92D400F
More to the point, why doesn't it crash my Pi3? It should. I might see if I can automate fixups of this instruction as I've come across a few games that use it, such as Zarch. Although it's a bug in the code, up to the Pi1 it will be executed with the intended outcome, Pi2 onward will report it as illegal (except in my case it would appear!)
Re: ADFFS 2.61 beta
@Jon: After adding the line you told me to the !Run script, it works!
BTW, could a "magical line" be added like in "Alone in the Dark" to skip certain very uneasy codes being asked at the beggining? It's very, very tedious to enter the game for testing purposes. I know it may be incompletable, but maybe it is (I have yet to finish Alone, but I am very far into the game nad there have been no problems at all).
NOTE: Maybe the illegal instruction is not being triggered on your Pi 3 because it's being down-clocked? Pi3 tend do downclock on power issues very easily.
BTW, could a "magical line" be added like in "Alone in the Dark" to skip certain very uneasy codes being asked at the beggining? It's very, very tedious to enter the game for testing purposes. I know it may be incompletable, but maybe it is (I have yet to finish Alone, but I am very far into the game nad there have been no problems at all).
NOTE: Maybe the illegal instruction is not being triggered on your Pi 3 because it's being down-clocked? Pi3 tend do downclock on power issues very easily.
Re: ADFFS 2.61 beta
I've had a very quick look, see if the following removes the manual protection in Flashback without side-effects:Vanfanel wrote:It's very, very tedious to enter the game for testing purposes.
JITMEMORYA 13298 E92D4FFF E1A0F00E
According to the ARM ARM, the instruction is not illegal, but unpredictable, which explains why it's not always triggering an abort.Vanfanel wrote:Maybe the illegal instruction is not being triggered on your Pi 3 because it's being down-clocked? Pi3 tend do downclock on power issues very easily.
Re: ADFFS 2.61 beta
I'm not entirely sure what's going on here, if the boot script forces a MODE change immediately after the JIT starts, the game always works. That implies the issue is related to the MODE change process, but quite how it would cause OS_File 15 to return incorrect file info is beyond meJonAbbott wrote:I have spotted an issue which can be reproduced by running Air Supremacy, quitting (via either CTRL-SHIFT-F12 or the in-game Exit) and then trying to run it again. After doing this, OS_File 15 returns incorrect file info and the error handler in BASIC hangs. Reloading the ADFFS Module doesn't resolve the issue, a reboot is required.
Starting the JIT manually before running the game also resolves the problem, so there's something odd going on.
EDIT: PON in Winterland suffers the same problem, in that switching MODE after the JIT starts resolved the issue. It's possibly related to running games from HD, so may be due to memory not being reset.
Re: ADFFS 2.61 beta
@Jon: Flashback is now working without the totally annoying codes (if you tried to access the games several times, even with a copy of the manual, you know how bad it was!) and I don't see any strange synthoms. I will play as far as I can, but I believe there's no problem at all: I have got to the second "world" with no problems.
Re: ADFFS 2.61 beta
Also, I can confirm ZARCH is hanging ramdomly, so I suspect the "impredictable" instruction is at work.
Re: ADFFS 2.61 beta
The boot script patches it, so it's not that particular bug. If you've created a HD install, check the script is up to date.Vanfanel wrote:Also, I can confirm ZARCH is hanging ramdomly, so I suspect the "impredictable" instruction is at work.
I've left it running in demo mode for 30 mins, no crashes yet.
Re: ADFFS 2.61 beta
@Jon: Yes, updating the boot script made it work reliably, no more random crashes.
Re: ADFFS 2.61 beta
Always the first thing to check if it's an HD install. I've made the HD installs that aren't truly HD installable use ADFBootFloppy to avoid having to keep the script up to date.