REV 3 - REV 5 - The beginning (ST536)

[exxos]

I did the video but I think I have solved the problem.. It actually would work fine on one of my later boards, but three of the earlier revisions all at the same problems of locking up somewhere around ROM copy on boot.

It looks like the problem was the CPU was not entering 50mhz fast enough on TTram access. It just must have been a fluke of tolerances that it worked before. I have shaved a cycle (about 10ns) off that switch , and its behaving fine at the moment :hide:

[PhilC]

:chairsmack: :chairsmack: :chairsmack:

[exxos]

I seem to have found a odd bug somewhere..It has been driving me nuts all day.. But I think I have found a method which seems repeatable now..

Basically if I cold boot and skip STram test (just pressing space) it gets to desktop fine.

If I do the same, but press spacebar on TTram test, it seems to load stuff from floppy , but resets just before it gets to desktop.

:shrug:

I've tried several boots with EMUTOS and it does not seem to behave the same way. Though it may not be a further test because it doesn't copy ROM to TTram like my ST536 TOS does. So that could be a factor. Or something is "broken" in TOS..

Not sure how to diagnose this one..

EDIT:

It might have something to do with my ROM copy to TTram ... If I don't run ROM from TTram, it seems to behave.. hmmm :shrug:

[exxos]

Let TOS do TTRAM test and it resets every time before getting to desktop. It shows the mouse for a split second then resets. If I skip TTRAM test, it's fine..

It will run GB6 from TTRAM. ROM shadow in TTRAM.

Its just done a full pass of YAARTTT without issue.

I've tried slowing TTRAM access down, disabled the CPU cache. Various code fiddles.. Not had any luck.

If I disable the ROM shadow then it works...

As the ROM copy still happens.. I'll write a STOS program tomorrow to see if ROM in TTRAM is still valid after TOS TTRAM test..

But there shouldn't be any way TOS can corrupt the ROM shadow. I only let TOS see 60MB.. My ROM shadow is around the 62MB mark.

[Badwolf]

But there shouldn't be any way TOS can corrupt the ROM shadow. I only let TOS see 60MB.. My ROM shadow is around the 62MB mark.

How do you do that?

Alternative suggestion: only do the shadow ROM after TT-RAM detection (& testing if it's allowed to proceed) happens?

BW

[exxos]

How do you do that?

Alternative suggestion: only do the shadow ROM after TT-RAM detection (& testing if it's allowed to proceed) happens?

That's what it does do...

I just issue berr after 60mb.. TOS doesn't find it and doesn't try to test it, at least AFAIK.. ROM copy happens after the detection.

[exxos]

I left it running overnight...

20260211_100351.jpg

[Badwolf]

How do you do that?

Alternative suggestion: only do the shadow ROM after TT-RAM detection (& testing if it's allowed to proceed) happens?

That's what it does do...

I just issue berr after 60mb.. TOS doesn't find it and doesn't try to test it, at least AFAIK.. ROM copy happens after the detection.

After the detection & testing?

BW

[exxos]

After the detection & testing

Testing only goes to 60mb as that is what TOS detects.. If it went beyond that, it'd trash rom and crash right away.. At least AIUI.

But this problem only seems to have surfaced recently.. I don't remember having it before.

I brainstormed with AI, it thinks my bus error logic might be glitching and the fault only shows up during TOS TTram check. :shrug: only avenue to go down currently. I can gate bus error right after rom copy.. So be easy to test.

[PhilC]

@exxos disable the ram test routine, it's not an essential and YAART does a better job. Itll also speed up boot times