Jump to content
InsaneMultitasker

Horizon RAMdisk ROS and CFG Development

Recommended Posts

Tim,

 

Could you do a version of CFG832D with registers 8-15 cleared at startup and registers 0-7 set and the status byte cleared before exit? I'm still having some intermittent problems.

Share this post


Link to post
Share on other sites

Tim,

 

Could you do a version of CFG832D with registers 8-15 cleared at startup and registers 0-7 set and the status byte cleared before exit? I'm still having some intermittent problems.

Sure thing. Probably won't be until tomorrow evening as I'm tied up with work. I did look at the AMSTEST4 exit code for other clues, and will validate a few other things when I make the changes.

 

If you are still seeing problems, could you try one of the pre-SAMS detection versions - see if you can get it to fail in the same manner.

Share this post


Link to post
Share on other sites

Tim,

 

Could you do a version of CFG832D with registers 8-15 cleared at startup and registers 0-7 set and the status byte cleared before exit? I'm still having some intermittent problems.

 

Hey that was my idea!

 

(just another 80 column VDP innocent user here)

 

:)

 

Gerchoo (bless you Pebbles!)

Share this post


Link to post
Share on other sites

 

Hey that was my idea!

do you have some specific settings that you've used in the past for this similar problem?

 

I was contemplating whether the user interrupt hook has something hiding in it, or something else odd that XB27 and the V9938 would dislike. The only changes I've made this month to CFG are adding more devices and implementing the AMS test. The former should have no impact; the latter.. who knows ;)

Share this post


Link to post
Share on other sites

I gave CFG832C and CFG832D a good workput both last night and this morning. I have NOT been able to consistantly reproduce the exit failure. CFG832C did not fail at all in 25 tests and CFG832D failed ONCE in 25 tests and that was after exiting the configuration section and then pressing FCTN-= to exit from the detection section. The failure took the form of a "fractured" title screen that gave way to MENU after about a second. It was not like the initial exit failures that I encountered that required the console to be turned off. The only thing that I can surmise is that a "gremlin" was passing through the neighborhood and decided to stop at my machine for a few minutes. We'll have to wait until Gazoo gets his system re-assembled to see if he has the same or similar problems with it since he also has a Dijit AVPC card.

 

I'm not going to worry about the exit quirk since it occurs so infrequently now and does not affect performance. I still think it might be a good precautionary measure to add Gazoo's suggestion about the registers and status byte. It should ONLY apply to those of us with 80-column cards who use CFG832x and wouldn't interfere with 40-column systems.

Share this post


Link to post
Share on other sites

do you have some specific settings that you've used in the past for this similar problem?

 

I was contemplating whether the user interrupt hook has something hiding in it, or something else odd that XB27 and the V9938 would dislike. The only changes I've made this month to CFG are adding more devices and implementing the AMS test. The former should have no impact; the latter.. who knows ;)

 

I've only had 1 or 2 problems with registers 7 through 15 not being cleared. If you write those registers first, the 9918 will accept them as registers 0 through 7, and then you write registers 0 through 7. This procedure allows compatibility for both VDP chips.

 

Gazoo

Share this post


Link to post
Share on other sites

 

I've only had 1 or 2 problems with registers 7 through 15 not being cleared. If you write those registers first, the 9918 will accept them as registers 0 through 7, and then you write registers 0 through 7. This procedure allows compatibility for both VDP chips.

 

Gazoo

I did add instructions to clear the GPL status byte and user interrupt words as a safety measure.

 

What I do not understand is if the registers are set wrong before CFG (or another program) is loaded, why don't the bad results show up sooner? And if the program is causing the problem, then we should either see the garbled screen right away or at least on every restart. That isn't happening... there is no consistency.

 

I'm not able to connect the dots very easily on this one, o' great Gazoo!

Share this post


Link to post
Share on other sites

If the Great Gazoo said it, it is undoubtedly true! I always just nod my head and comply.

 

Don't do that, the insane one is much more knowledgable. He's making sense where I just do things that work for me and are generally overkill.

Clearing the status byte is probably enough.

Share this post


Link to post
Share on other sites

Does it matter if you eliminate a bug with a fly swatter or a hammer? It is just as dead. I defer to either of you when it comes to A/L (and try not to get in the way)..

Share this post


Link to post
Share on other sites

Does it matter if you eliminate a bug with a fly swatter or a hammer? It is just as dead. I defer to either of you when it comes to A/L (and try not to get in the way)..

 

Just don't get in the way of Gazoo's sledge hammer. :twisted:

  • Like 2

Share this post


Link to post
Share on other sites

Does it matter if you eliminate a bug with a fly swatter or a hammer? It is just as dead. I defer to either of you when it comes to A/L (and try not to get in the way)..

We both employ many of the same tools: fly-swatters, hammers, bull-dozers, and bit of machine language alchemy combined with a dash of "how the $%% did I get here?" magic. All depends on the mood - it is a hobby after all. Poor Gazoo has seen some of my warped code, he knows of what I speak.

 

;)

  • Like 1

Share this post


Link to post
Share on other sites
  • SAMS/AMS detection is limited to CRU address 0x1E00. CFG will only attempt to detect SAMS if no other card is found at this CRU first.
  • The Geneve detection has been corrected. If CFG is loaded with the standard DSR, no card detection occurs. However, if a ROMpage is activated, the card detection will poll peripherals as expected. In both cases, the Geneve will be listed as a peripheral at CRU >9600.
  • V9938 VDP registers are cleared at startup. VDP registers 0-7 are now set during startup.
  • Status byte is cleared prior to exiting.

While these changes should not affect ROS loading/saving, use with caution. I have been unable to test on the real hardware today due to a power outage in the area that knocked my system offline. I need to restore my work platter from its backup tomorrow.

 

 

 

 

 

Z-CFG832E.dsk

post-25764-0-47888400-1438413758_thumb.png

post-25764-0-11761400-1438413773_thumb.png

  • Like 2

Share this post


Link to post
Share on other sites

It seems that the "brute force" approach works quite well. That's what I did back when I was writing a bit of A/L code but that was "in a time long ago and a Galaxy far away".

Share this post


Link to post
Share on other sites

It looks like the "devious mind" (and cunning A/L programming skills) of InsaneMultitaker have won! It is with regret that I must admit that I have failed in my attempt to get his latest version of CFG832 to crash or otherwise misbehave at exit. It is completely stable and, despite his warning, I found no problems saving or loading ROS wit it. As my Aussie mates would say "Good on ya', mate!".

  • Like 1

Share this post


Link to post
Share on other sites

Does anyone happen to have documentation regarding the Mechatronics, AVPC, EVPC video cards and the BwG Disk Controller cards with respect to how they manipulate 0x8370 and what VDP bytes they reserve? I can't find any documentation explaining the 'what' or 'why' for these cards.

 

I finally got back to restoring some of my files and it looks like one of the last questions I had unanswered was related to these cards' use of VDP memory. I've all but eliminated the Ramdisk OS reliance upon VDP but I am concerned boot tracking may be an issue for some older programs.

 

Before I release the next release I'm hoping to track down these last pieces of 'lost' info. No pun intended.

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...

  • Recently Browsing   0 members

    No registered users viewing this page.

×
×
  • Create New...