[sebhc] H8 Emulator update
dave04a at dunfield.com
Sun May 30 17:22:05 CDT 2004
>Dave, this just keeps getting better - I'm having a lot more trouble
>breaking it! ;>)
Thats the idea!
>- here's the only way I could do it this time around - my normal ini
>file looks like this:
>However, if I start with an .ini file that _only_ loads the rom -
>and don't mount a disk (f5) before hitting GO (4) then the entire
>emulator just vanishes! This is in a WinXP environment - I havn't tried
>it under Win2K or DOS.
This would be the old 80x86 CPU fart at running the stack misaligned off
the top of the 64k segment - I really don't want to fix it because it will
really slow the thing down (have to break all stack transfers (always a
word) into two separete byte operations, and then reassemble them afterward.
When you do this, you are doing EXACTLY the same thing that you did in the
beginning when you ran at 040,000 - you are executing the PAM8 address space.
The reason for this is that (as documented), the H17 system (IO ports and
ROM) are **NOT INSTALLED** in the simulation until you mount a disk - this
allows me to simulate my machine, and still let you have a disk.
So - pressing '4' goes at 030.000 which executes NOP's up until 040.000
where it does exactly what it did before.
I could give you an option to put in the INI file to install the H17
without mounting a disk (or you could just mount a disk - it can be
re-mounted at any time).
Another thing I could do would be to initialize the 8080 memory address
space to $76 (HLT) instead of $00 (NOP) - this would cause the cpu to
reset to PAM-8 if you execute unititialized memory - not really like a
true H8 (but hey - theoritically, an H8 *could* power-up with HLT's
>Here's a minor quibble - would you consider renaming the f5 function to
>"load" or "insert" a disk? Mount isn't really correct since after I
>mount a disk with f5, my next step is to return to tty mode and mount
>the disk again.
F5 isn't load or insert ... at least not for the tape files. Mount seems
most "correct" to me (It's also what I call the function in my other emulators
and I do like to maintain consistancy) - unles this is a real problem for
you, I would like to leave it as it is.
>Another "problem" is with the low level commands INIT and SYSGEN - if I
>try to run INIT, it exits gracefully with the error message "Wrong type
>of media ...". I can ctrl-D out of it and reboot, but I can't initialize
>a disk. Likewise with SYSGEN - if I attempt to gen a disk, I get an
>error saying the disk is not initialized by the correct version of INIT
>(I haven't tried to figure out which disk-images were made from disks
>created under which versions of HDOS, so this may be a legitimate
I'll have to look into them - I wouldn't be supprised if there's a software
Another thing - My driver builds the sector header on the fly, and I may not
gracefully handle someone trying to write to it - I'll look into it when I
have a chance - in the meantime, DOS copy works great to replicate a disk -
perhaps we could include a "blank" disk image for now.
Perhaps Steven will be able to provide me with some info on exactly what
these commands do and in what order - but no rush, I'm going to be tied
up with "real work" this week.
>Finally, just a few more suggestions, related to disk handling. Would
>you add a couple of "indicators" to the f5/D display so that we could
>see if drives are loaded R or W/R?
Already there - If you look at the status line when you press F5/D, you
will see something like: R1:name1 W2:name1 W3:name1
The R1 tells you that drive 1 is Read-only, the W2 and W3 tell you that
these drives are read-Write.
>obvious that we somehow have to find you an H17 setup so that it will be
>There is a 6-page "HDOS Cookbook" that is part of the software reference
>manual which I think would be helpful to museum visitors. I'll try to
>find the correct 1.x version and OCR it down to a minimal text file if
>you would like to include it as part of the "display".
That would be great! Most people will (like me) have only the HELP command
which is a bit "light" - even a few pages of reference would be really
dave04a (at) Dave Dunfield
dunfield (dot) Firmware development services & tools: www.dunfield.com
com Vintage computing equipment collector.
Delivered by the SEBHC Mailing List
More information about the Sebhc