FX CAST Atari ST core

https://github.com/MiSTer-devel/Main_MiSTer/wiki

Moderators: Mug UK, Zorro 2, Greenious, spiny, Sorgelig, Moderator Team

ijor
Hardware Guru
Hardware Guru
Posts: 3792
Joined: Sat May 29, 2004 7:52 pm
Contact:

Re: FX CAST Atari ST core

Postby ijor » Wed Mar 20, 2019 7:59 pm

ericgus wrote:Just curious if work on this core has halted?


Hi. Not halted, but paused. Sorry, I was too busy with real life I'm afraid. Hopefully core development would restart shortly.
Fx Cast: Atari St cycle accurate fpga core

vanfanel
Obsessive compulsive Atari behavior
Obsessive compulsive Atari behavior
Posts: 118
Joined: Tue Oct 09, 2018 10:19 pm
Location: Salamanca, España

Re: FX CAST Atari ST core

Postby vanfanel » Wed Mar 20, 2019 8:43 pm

ijor wrote:
ericgus wrote:Just curious if work on this core has halted?


Hi. Not halted, but paused. Sorry, I was too busy with real life I'm afraid. Hopefully core development would restart shortly.


May I politely suggest to opensource, Ijor? This core would benefit A LOT from opensourcing! (Scalers/Filters, for example...)

ericgus
Captain Atari
Captain Atari
Posts: 452
Joined: Fri Apr 08, 2016 4:53 am
Location: Boston MA - USA

Re: FX CAST Atari ST core

Postby ericgus » Thu Mar 21, 2019 4:53 am

ijor wrote:
ericgus wrote:Just curious if work on this core has halted?


Hi. Not halted, but paused. Sorry, I was too busy with real life I'm afraid. Hopefully core development would restart shortly.


Sure no worries, understandable life does tend to do that.. thanks for responding!

DYates69
Atarian
Atarian
Posts: 5
Joined: Wed Apr 25, 2018 3:25 pm

Re: FX CAST Atari ST core

Postby DYates69 » Thu Mar 21, 2019 11:37 am

What's the likelihood of a Falcon 030 core within the Cyclone V?

killersquirel
Atarian
Atarian
Posts: 6
Joined: Thu Dec 27, 2018 8:53 pm

Re: FX CAST Atari ST core

Postby killersquirel » Sun Mar 24, 2019 9:50 pm

I am loving this core. Hopefully disk write access will come soon. Is that on the cards to get implemented in the future?

ijor
Hardware Guru
Hardware Guru
Posts: 3792
Joined: Sat May 29, 2004 7:52 pm
Contact:

Re: FX CAST Atari ST core

Postby ijor » Mon Mar 25, 2019 1:10 pm

DYates69 wrote:What's the likelihood of a Falcon 030 core within the Cyclone V?


Cyclone V is a whole family of devices, so it depends on the exact size. If you mean the Cyclone V device present at the DE10-Nano/MiSTer, then yes, I believe it should fit. But note that it is not currently on my plans.

killersquirel wrote:Hopefully disk write access will come soon. Is that on the cards to get implemented in the future?


Yes, of course.
Fx Cast: Atari St cycle accurate fpga core

Milongero
Retro freak
Retro freak
Posts: 15
Joined: Sun Nov 25, 2018 2:24 pm
Location: Germany

Re: FX CAST Atari ST core

Postby Milongero » Thu Mar 28, 2019 8:11 am

Hi I found two games for the high resolution which unfortunately don't work properly.

Oxyd 1 and Oxyd 2

The image of the bricks is wrong and the ball always falls down automatically.

Esprit, which is actually Oxyd 0.5, works great.

Greetings Lutz

User avatar
Paradroyd
Captain Atari
Captain Atari
Posts: 287
Joined: Tue Sep 10, 2013 10:50 pm
Contact:

Re: FX CAST Atari ST core

Postby Paradroyd » Fri Mar 29, 2019 9:20 pm

I've found what seems to be a really interesting bug in this core, and I'd like to know if anyone else has seen it or can reproduce it.

I hesitated to even post this as it's really weird, and I was waiting to see if anyone else posted anything like this, but as far as I know, no one has, so here goes...

The first thing I noticed is that if I use this core for an extended amount of time, and the actual length of time varies widely, eventually I will suddenly completely lose keyboard and joystick input into the core. When the input suddenly gets lost, whatever the state of the joystick or keyboard is at that instant, whether a key or joystick button/direction is up or down, it will stick that way. For example of you have the joystick pressed left when this happens, it will be stuck that way until the ST core is reset.

One interesting part of this is that though the ST core itself loses input, I can press F12 after other input has quit working, and still get into the menu and use it to reset the core or reset MiSTer itself. The mouse input also keeps working even after joystick and keyboard input are cut off.

After reset (of either the core or the MiSTer itself), input will work normally, until it happens again, which could happen after 3 minutes, or an hour, or any other seemingly arbitrary amount of time, at which point you can reset again, and it will work again for some amount of time. Repeat...

I thought that this only affected the latest version of the ST core, but I have verified that it happens at least several versions back. It even happens on the version prior to the MidiLink integration. I'm not sure if this is tied to recent input framework changes or not, but this particular thing seems to affect only the ST core / FXCast.

One other input oddity that I've noticed with this core which may or may not be related, and this one is easily reproducible on my MiSTer.

When I start the core, if I move the mouse in any direction, then pick up the joystick and rotate the stick though it's various contact points (N, S, E, W), as I do, the mouse pointer will creep in the last direction that I had just moved the mouse as I step the stick through it's different contact points, whatever that last mouse direction was. If I then move the mouse a short distance in a different direction, then stop moving it, and rotate the joystick again, the mouse pointer will creep in that new direction. It's like joystick events are leaking into the mouse input somehow.

If you can't reproduce this and need me to post a video of it happening, let me know, because as I said, this is easy to reproduce.

Hopefully there is one simple-to-find underlying cause.

I see absolutely nothing like this in any of the other cores, and I use several of them a lot.
- Paradroyd
@paradroyd on Twitter, @paradroyd@mastodon.sdf.org on Mastodon

BBond007
Captain Atari
Captain Atari
Posts: 366
Joined: Wed Feb 28, 2018 3:23 am

Re: FX CAST Atari ST core

Postby BBond007 » Sat Mar 30, 2019 2:30 am

Paradroyd wrote:I've found what seems to be a really interesting bug in this core, and I'd like to know if anyone else has seen it or can reproduce it.


I have had the same issue with the controller getting stuck (but still working in OSD) with every release - even the releases prior to it being named FX CAST.

I have not found a way to reproduce it as it just seems to happen at random intervals and sometimes taking hours before it occurs.

Some claim using a powered USB hub helps, but I have had it happen even when using a powered hub...

User avatar
Paradroyd
Captain Atari
Captain Atari
Posts: 287
Joined: Tue Sep 10, 2013 10:50 pm
Contact:

Re: FX CAST Atari ST core

Postby Paradroyd » Sat Mar 30, 2019 4:57 am

BBond007 wrote:
Paradroyd wrote:I've found what seems to be a really interesting bug in this core, and I'd like to know if anyone else has seen it or can reproduce it.


I have had the same issue with the controller getting stuck (but still working in OSD) with every release - even the releases prior to it being named FX CAST.

I have not found a way to reproduce it as it just seems to happen at random intervals and sometimes taking hours before it occurs.

Some claim using a powered USB hub helps, but I have had it happen even when using a powered hub...


Yeah..I don't believe that this is a hub problem. I NEVER see anything like this in other cores.

Also, can you reproduce the last part, where moving the joystick causes the mouse pointer to move?
- Paradroyd
@paradroyd on Twitter, @paradroyd@mastodon.sdf.org on Mastodon

Indcsion
Atari User
Atari User
Posts: 34
Joined: Fri Oct 12, 2018 11:48 am

Re: FX CAST Atari ST core

Postby Indcsion » Mon Apr 01, 2019 12:09 am

Any chance you could let someone upgrade it with the scaling options? Id like to test it more but it just doesn't look to good on my tv.

ijor
Hardware Guru
Hardware Guru
Posts: 3792
Joined: Sat May 29, 2004 7:52 pm
Contact:

Re: FX CAST Atari ST core

Postby ijor » Mon Apr 01, 2019 12:49 pm

Milongero wrote:Hi I found two games for the high resolution which unfortunately don't work properly.

Oxyd 1 and Oxyd 2
The image of the bricks is wrong and the ball always falls down automatically.
Esprit, which is actually Oxyd 0.5, works great.


Thanks for the report. Please let me know the version of the games that you are using so I could check them.
Fx Cast: Atari St cycle accurate fpga core

ijor
Hardware Guru
Hardware Guru
Posts: 3792
Joined: Sat May 29, 2004 7:52 pm
Contact:

Re: FX CAST Atari ST core

Postby ijor » Mon Apr 01, 2019 12:57 pm

Paradroyd wrote:The first thing I noticed is that if I use this core for an extended amount of time, and the actual length of time varies widely, eventually I will suddenly completely lose keyboard and joystick input into the core. When the input suddenly gets lost, whatever the state of the joystick or keyboard is at that instant, whether a key or joystick button/direction is up or down, it will stick that way. For example of you have the joystick pressed left when this happens, it will be stuck that way until the ST core is reset.


Yes, as I stated in the first release, and as commented by Bbond007, some testers reported this behavior. I could never reproduce it, but it does seem to be related to USB power and heat. Those that reported the problem and then changed to a self powered USB hub, either never had the problem again or it happened much rarely.

Note that connecting the MiSTer console to a PC has a similar effect as using a powered USB hub; the PC provides some power to the board in that case. Can you try that and see if it makes any change?

Again, it seems to be related to USB power and heat, I'm not really sure. A couple of users reported the same (or similar) behavior with the Amiga core.
Fx Cast: Atari St cycle accurate fpga core

someonesfortune
Atarian
Atarian
Posts: 1
Joined: Wed Oct 25, 2017 5:36 am

Re: FX CAST Atari ST core

Postby someonesfortune » Fri Apr 05, 2019 5:04 am

Hi Ijor, I want to start by saying "Thank you!" for creating an ST core for the MiSTer. I really do miss my old TT and all the other ST equipment I've had over the years. A couple years ago I tried to buy me an ST to play around and relive some of the old memories. I couldn't find one I was interested in and when I did it went for a crazy price. Guess if I really wanted it I could pay but one thing that always hovered over buying an old computer was the space I would need and supported monitor, etc, etc, etc... I ended up buying a MiST and really enjoyed it to recreate a lot of the experiences I had way back when. Now it is even better with your core as a lot of software works really well. The old spectrum bee and clown pictures never looked so good.

You may already know this and may already have plans to add something like this in a future release, but I thought to share it in case it may help.

As I have been playing around trying out things I thought I would mention something that might help improve the experience. When booting an ST or STE computer that had TOS 1.62 and below there is a long delay before you would see the desktop. On a real ST you would notice the floppy drive LED flicker as it tried to read from the floppy. This was because TOS would look at the floppy drives to see if a TOS file or other OS boot files were present to load them. If it didn't find a floppy it would try a few times before timing out and continuing to boot. A solution to make your ST boot much faster (even if you had a hard drive attached to your ST) was to leave a formatted blank disk in the floppy drive. This would reduce the boot up times enough to compare it to today's experience of booting from a spinning hard drive compared to an SSD boot times.

Have you thought about having an option to have a boot.rom that could be a blank floppy image? Having an option to always mount a floppy image upon core start or something like a boot.rom file would make it boot up much faster. There was something similar on the MiST core which helped speed up the boot time also.

In TOS 2.06 and above they replaced that routine in TOS and added the memory test screen which hides the floppy search. Thus allowing you to hit a key and continue booting. Those days of having to wait for 24mb of ram to test, oh good lord I was losing youth staring at the screen :lol:

User avatar
Paradroyd
Captain Atari
Captain Atari
Posts: 287
Joined: Tue Sep 10, 2013 10:50 pm
Contact:

Re: FX CAST Atari ST core

Postby Paradroyd » Sun Apr 07, 2019 3:21 am

ijor wrote:Yes, as I stated in the first release, and as commented by Bbond007, some testers reported this behavior. I could never reproduce it, but it does seem to be related to USB power and heat. Those that reported the problem and then changed to a self powered USB hub, either never had the problem again or it happened much rarely.

Note that connecting the MiSTer console to a PC has a similar effect as using a powered USB hub; the PC provides some power to the board in that case. Can you try that and see if it makes any change?

Again, it seems to be related to USB power and heat, I'm not really sure. A couple of users reported the same (or similar) behavior with the Amiga core.


I don't see why it would only affect FXCast, as I haven't seen this behavior with any other cores, and I spend many hours in the Amiga and C64 cores. Nevertheless it sounds like I should be using a powered hub anyway, so I have one on order and will do more testing after I get that installed.

There is still the fact that if I move the mouse in any direction, then move the joystick button or click the button on any gamepad or joystick, every time the joystick control is clicked, the mouse pointer will jump in the direction that the mouse was last moved. This happens consistently every time, and so is very easy to reproduce. This is not heat, as it happens immediately after turning Mister on cold, as well as later. It also does not ever happen in any other cores.

At any rate, I'll wait until I get the powered hub in place before testing with FXCast any further.
- Paradroyd
@paradroyd on Twitter, @paradroyd@mastodon.sdf.org on Mastodon

ijor
Hardware Guru
Hardware Guru
Posts: 3792
Joined: Sat May 29, 2004 7:52 pm
Contact:

Re: FX CAST Atari ST core

Postby ijor » Sun Apr 07, 2019 12:18 pm

Paradroyd wrote:I don't see why it would only affect FXCast, as I haven't seen this behavior with any other cores, and I spend many hours in the Amiga and C64 cores. Nevertheless it sounds like I should be using a powered hub anyway, so I have one on order and will do more testing after I get that installed.


At least one tester reported the same behavior on the Amiga core. But again, I can't say I am sure it is indeed related to Usb powering. If you don't have a powered hub you might try connecting the MiSTer console to a PC.

There is still the fact that if I move the mouse in any direction, then move the joystick button or click the button on any gamepad or joystick, every time the joystick control is clicked, the mouse pointer will jump in the direction that the mouse was last moved.


Didn't check that. It might be a bug in my code. Will check it and let you know.
Fx Cast: Atari St cycle accurate fpga core

User avatar
Paradroyd
Captain Atari
Captain Atari
Posts: 287
Joined: Tue Sep 10, 2013 10:50 pm
Contact:

Re: FX CAST Atari ST core

Postby Paradroyd » Mon Apr 08, 2019 1:34 am

Thanks!

I should have a separately powered hub in place in a day or two too, then I'll do some more testing.
- Paradroyd
@paradroyd on Twitter, @paradroyd@mastodon.sdf.org on Mastodon

laxer3a
Atarian
Atarian
Posts: 5
Joined: Sat Apr 02, 2016 7:58 pm

Re: FX CAST Atari ST core

Postby laxer3a » Mon Apr 08, 2019 9:58 pm

Hi !

Iam really glad to see again some activity from Ijor. I've been waiting for 6 month to finally be able to have a look at the HDL. I bet I can wait a little more... ;-)

But please consider that the longer you wait, the longer other feature and support could have been made by others. I sincerely hope you will find time outside of your work/life at least to push things at a level that is 'worth enough for you' to release.

(Also, the current state of FX Cast is probably locking anybody else to attempt doing progress on the Atari ST machine : why start/spend time on a HDL project that could takes month or years if you release your stuff in a few weeks)

Anyway, I am keeping hope to see an accurate open source Atari ST within this year ;-)

User avatar
Paradroyd
Captain Atari
Captain Atari
Posts: 287
Joined: Tue Sep 10, 2013 10:50 pm
Contact:

Re: FX CAST Atari ST core

Postby Paradroyd » Tue Apr 09, 2019 1:05 am

ijor wrote:
At least one tester reported the same behavior on the Amiga core. But again, I can't say I am sure it is indeed related to Usb powering. If you don't have a powered hub you might try connecting the MiSTer console to a PC.

As of tonight, I have a separately powered hub in place. No input control loss so far, but sometimes it doesn't happen for days, as you noted, so I'll try to use it a lot this week. So far, so good


ijor wrote:
There is still the fact that if I move the mouse in any direction, then move the joystick button or click the button on any gamepad or joystick, every time the joystick control is clicked, the mouse pointer will jump in the direction that the mouse was last moved.


Didn't check that. It might be a bug in my code. Will check it and let you know.


For what it's worth, this still happens when I'm on the powered USB hub, though it's possible (likely, actually) that this weird problem is completely separate from the loss of control that occasionally happens. Anyway, just thought I should let you know.
- Paradroyd
@paradroyd on Twitter, @paradroyd@mastodon.sdf.org on Mastodon

djmartins
Atari maniac
Atari maniac
Posts: 97
Joined: Wed Nov 21, 2018 10:26 pm

Re: FX CAST Atari ST core

Postby djmartins » Wed Apr 10, 2019 3:08 am

I had some USB problems with two different powered hubs I used and one keyboard.
Since then I upgraded to the nicely Sorgelig designed 7 port one and a short mechanical keyboard
and have had zero problems since.

ijor
Hardware Guru
Hardware Guru
Posts: 3792
Joined: Sat May 29, 2004 7:52 pm
Contact:

Re: FX CAST Atari ST core

Postby ijor » Wed Apr 10, 2019 5:29 pm

laxer3a wrote:Iam really glad to see again some activity from Ijor. I've been waiting for 6 month to finally be able to have a look at the HDL. I bet I can wait a little more... ;-)


You are welcome to ask as much as you like, but I'll release the source when I'll be ready :)

(Also, the current state of FX Cast is probably locking anybody else to attempt doing progress on the Atari ST machine : why start/spend time on a HDL project that could takes month or years if you release your stuff in a few weeks)


I'm not so sure about that. There is no reason for avoiding multiple implementations. I think it is actually a good thing. I said that I plan (eventually) to port FX CAST to MiST. This didn't stop slingshot that is enhancing the old MiST version with my 68K code. And I am actually, and very gladly, helping him as much as I can.

Btw, if there is something specific you want to ask about the code, or about the hardware, please don't hesitate to ask. I'm not keeping any actual information for myself. I never did. Yeah, I know, this is not nearly the same as looking at the source.
Fx Cast: Atari St cycle accurate fpga core

User avatar
Paradroyd
Captain Atari
Captain Atari
Posts: 287
Joined: Tue Sep 10, 2013 10:50 pm
Contact:

Re: FX CAST Atari ST core

Postby Paradroyd » Sun May 05, 2019 7:04 am

ijor wrote:
Paradroyd wrote:The first thing I noticed is that if I use this core for an extended amount of time, and the actual length of time varies widely, eventually I will suddenly completely lose keyboard and joystick input into the core. When the input suddenly gets lost, whatever the state of the joystick or keyboard is at that instant, whether a key or joystick button/direction is up or down, it will stick that way. For example of you have the joystick pressed left when this happens, it will be stuck that way until the ST core is reset.


Yes, as I stated in the first release, and as commented by Bbond007, some testers reported this behavior. I could never reproduce it, but it does seem to be related to USB power and heat. Those that reported the problem and then changed to a self powered USB hub, either never had the problem again or it happened much rarely.

Note that connecting the MiSTer console to a PC has a similar effect as using a powered USB hub; the PC provides some power to the board in that case. Can you try that and see if it makes any change?

Again, it seems to be related to USB power and heat, I'm not really sure. A couple of users reported the same (or similar) behavior with the Amiga core.


For what it's worth, I had the joystick input freeze on me twice tonight in the space of a few minutes (It's the first time I've attempted to play a game in FXcast in a while). In this particular instance I was playing Bubble Bobble, but I don't think the problem had anything to do with that.

I've been using a separately powered USB hub since shortly after last time I posted here several weeks ago, and the MiSTer has a fan and heat sink on it. I switched to the MiniMig and had no problems there.
- Paradroyd
@paradroyd on Twitter, @paradroyd@mastodon.sdf.org on Mastodon

ijor
Hardware Guru
Hardware Guru
Posts: 3792
Joined: Sat May 29, 2004 7:52 pm
Contact:

Re: FX CAST Atari ST core

Postby ijor » Wed May 08, 2019 1:07 am

Paradroyd wrote:For what it's worth, I had the joystick input freeze on me twice tonight in the space of a few minutes (It's the first time I've attempted to play a game in FXcast in a while). In this particular instance I was playing Bubble Bobble, but I don't think the problem had anything to do with that.

I've been using a separately powered USB hub since shortly after last time I posted here several weeks ago, and the MiSTer has a fan and heat sink on it. I switched to the MiniMig and had no problems there.


Hi,

If you can, more or less, reproduce it somehow, then that would be extremely helpful to find and fix the problem.

Can you please connect the MiSTer USB serial console and let me know if you can see the debug messages and if you still get the input freeze problem? Do you don't know how to connect the console?
Fx Cast: Atari St cycle accurate fpga core

User avatar
Paradroyd
Captain Atari
Captain Atari
Posts: 287
Joined: Tue Sep 10, 2013 10:50 pm
Contact:

Re: FX CAST Atari ST core

Postby Paradroyd » Wed May 08, 2019 1:30 am

ijor wrote:Hi,

If you can, more or less, reproduce it somehow, then that would be extremely helpful to find and fix the problem.

Can you please connect the MiSTer USB serial console and let me know if you can see the debug messages and if you still get the input freeze problem? Do you don't know how to connect the console?


I will try to do this sometime within the next few days.

Thanks!
- Paradroyd
@paradroyd on Twitter, @paradroyd@mastodon.sdf.org on Mastodon

BBond007
Captain Atari
Captain Atari
Posts: 366
Joined: Wed Feb 28, 2018 3:23 am

Re: FX CAST Atari ST core

Postby BBond007 » Wed May 08, 2019 3:31 am

ijor wrote:Can you please connect the MiSTer USB serial console and let me know if you can see the debug messages and if you still get the input freeze problem? Do you don't know how to connect the console?


Unfortunately, it is no longer convenient for me to use the physical USB connection :(

It is more convenient, assuming it does not effect the debug messages, to use SSH (as opposed to physical USB connection) to achieve the same results:

Code: Select all

# killall MiSTer;/media/fat/MiSTer


I did report (viewtopic.php?f=123&t=34496#p355872) that the only messages I saw were concerning the visibility status of the OSD and nothing about button press events - like I did see with the initial version of FX CAST. At that time I was using the physical USB serial console (not the SSH method) + 4 watt PSU + powered USB hub... (https://youtu.be/ANGUdFz7AQ0)

Perhaps that (the lack of button press events messages) is related to running the most current 'MiSTer' bin and is unrelated to the version of the FX CAST core?

I now have 2 different MiSTers which both exhibit this same problem when using FX CAST yet don't seem to have the issue with other cores. Using the SSH method today on my newer DE10, I (still) see no event messages or error messages once a controller freeze occurs. In this latest case Bubble Bobble has frozen on me within the first few minutes of gameplay:

Code: Select all

Selected file AtariST/GAMES/B/Bubble Bobble (1987)(Firebird)[cr Electricbug].st                                                                                                                                                              with 368640 bytes to send for index 1.0
................................................................................                                                                                                                                                             ..........

OSD is now invisible
OSD is now visible
OSD is now invisible
FileLoad(open) File:/media/fat/config/FXCAST_input_18d8_0002_v2.map, error: -1.
FileLoad(open) File:/media/fat/config/input_18d8_0002_v2.map, error: -1.
OSD is now visible
OSD is now invisible
OSD is now visible
OSD is now invisible


Thanks. Hope this helps...


Return to “MiSTer”

Who is online

Users browsing this forum: elfbutt, flain and 5 guests