Voyager Emulators Self Test

[agarza]

Has anyone noticed that running the ON/* or ON/+ self test work fine (don't give errors) on all ROMS (HP10, HP12, HP16, etc) except on the HP15?

HP15 always gives Error 9. And have to reset the emulator to continue working properly.

I have tried it with several emulators, all seem to be based on Nonpareil.

I am coding a hardware emulator and this has me baffeld.

[J-F Garnier]

After an Error 9, you should have an error code in register X that should give you an indication of what is failing in the emulation. For instance in my own emulator (private tool) I got code 3 indicating that the RAM test failed.

Which emulators did you check, and what error code in X did you get?

J-F

[agarza]

I am running Nonpareil v0.78 on Ubuntu. And also tried a version the Mac port by Maciej Bartosiak.

After the Error 9 is displayed, pressing the clear button show 3 on the X register.

So I asume it is a RAM error? Do you know what ram address are beeing checked but the self test?

[rprosperi]

Just to be clear, when a self-test on an emulator returns an error result, it does NOT mean that the emulator is not running properly, it only means that the original firmware code that is checking the emulated h/w is not having expected results, most likely because the emulated h/w behaves slightly differently from the original, true h/w (for example, the RAM on the PC you use is probably much faster than the expected behavior). This does not mean the emulator isn't working properly, it only means that the emulated h/w does not behave as the diagnostic code expects. There is typically no reason to be concerned about self-test failures on an emulator.

For example even HP's own HP-15C LE device does not produce proper results when using the original self-test diagnostics, for the same reason: the emulated h/w is slightly different from what the self-test code expects. But no problem, the 15C LE still works as expected.

[agarza]

(08-08-2021 03:33 PM)rprosperi Wrote:  Just to be clear, when a self-test on an emulator returns an error result, it does NOT mean that the emulator is not running properly, it only means that the original firmware code that is checking the emulated h/w is not having expected results, most likely because the emulated h/w behaves slightly differently from the original, true h/w (for example, the RAM on the PC you use is probably much faster than the expected behavior). This does not mean the emulator isn't working properly, it only means that the emulated h/w does not behave as the diagnostic code expects. There is typically no reason to be concerned about self-test failures on an emulator.

For example even HP's own HP-15C LE device does not produce proper results when using the original self-test diagnostics, for the same reason: the emulated h/w is slightly different from what the self-test code expects. But no problem, the 15C LE still works as expected.

I am trying to 'fix' or patch the emulator so that the it can pass the test.
It is strange that ONLY the HP15 is not passing the self test.

Do you think it is failing due to speed, even thou it seems it is a RAM error?

[J-F Garnier]

(08-08-2021 03:33 PM)rprosperi Wrote:  For example even HP's own HP-15C LE device does not produce proper results when using the original self-test diagnostics, for the same reason: the emulated h/w is slightly different from what the self-test code expects. But no problem, the 15C LE still works as expected.

The HP-15C LE, both the physical unit and the HP emulator for Windows (hp15c.exe, 2012), are giving Error 9 with code=2 in register X, meaning a ROM checksum error. HP patched some locations of the HP-15C ROM, but didn't update the checksums.

J-F

[agarza]

(08-08-2021 07:48 PM)J-F Garnier Wrote:  

(08-08-2021 03:33 PM)rprosperi Wrote:  For example even HP's own HP-15C LE device does not produce proper results when using the original self-test diagnostics, for the same reason: the emulated h/w is slightly different from what the self-test code expects. But no problem, the 15C LE still works as expected.

The HP-15C LE, both the physical unit and the HP emulator for Windows (hp15c.exe, 2012), are giving Error 9 with code=2 in register X, meaning a ROM checksum error. HP patched some locations of the HP-15C ROM, but didn't update the checksums.

J-F

Since I’m getting code=3 I tried to see what memory addresses were being checked.
It checks ram addresses from 0x07 to 0x00 and 0xff to c0 (backwards) and then it shows Error 9.

Maybe it is accessing an out of range memory address?

[J-F Garnier]

(08-08-2021 08:13 PM)agarza Wrote:  Since I’m getting code=3 I tried to see what memory addresses were being checked.
It checks ram addresses from 0x07 to 0x00 and 0xff to c0 (backwards) and then it shows Error 9.

Maybe it is accessing an out of range memory address?

Some information about the Voyager register access idiosyncrasies can be found in this very old thread:
https://www.hpmuseum.org/cgi-sys/cgiwrap...read=73287
with the best Nut emulation experts (Hi all )

J-F

[agarza]

(08-10-2021 07:40 AM)J-F Garnier Wrote:  

(08-08-2021 08:13 PM)agarza Wrote:  Since I’m getting code=3 I tried to see what memory addresses were being checked.
It checks ram addresses from 0x07 to 0x00 and 0xff to c0 (backwards) and then it shows Error 9.

Maybe it is accessing an out of range memory address?

Some information about the Voyager register access idiosyncrasies can be found in this very old thread:
https://www.hpmuseum.org/cgi-sys/cgiwrap...read=73287
with the best Nut emulation experts (Hi all )

J-F

Thanks for the info. I will try to path the ROM to skip the Reg 8 test. But my microcode programming is very limited.

[J-F Garnier]

(08-08-2021 07:48 PM)J-F Garnier Wrote:  The HP-15C LE, both the physical unit and the HP emulator for Windows (hp15c.exe, 2012), are giving Error 9 with code=2 in register X, meaning a ROM checksum error.

As a side question, can the 2012 HP emulator for Windows of the HP-15C still be found?
It is no more on HP site, it's not on hpcalc.org.
I know it can be found on some weird sites, but I mean a safe, known place.
Surely I'm keeping a private copy but it would be nice to know that it is archived somewhere.

Reason is that HP15C.exe is blazing fast (for an emulator), and a great tool to test and run HP-15C code.
It has the same "bugs" as the HP-15C LE, but honestly it's really minor compared to its performance.

J-F

[Stevetuc]

(08-11-2021 07:22 AM)J-F Garnier Wrote:  

(08-08-2021 07:48 PM)J-F Garnier Wrote:  The HP-15C LE, both the physical unit and the HP emulator for Windows (hp15c.exe, 2012), are giving Error 9 with code=2 in register X, meaning a ROM checksum error.

As a side question, can the 2012 HP emulator for Windows of the HP-15C still be found?
It is no more on HP site, it's not on hpcalc.org.
I know it can be found on some weird sites, but I mean a safe, known place.
Surely I'm keeping a private copy but it would be nice to know that it is archived somewhere.

Reason is that HP15C.exe is blazing fast (for an emulator), and a great tool to test and run HP-15C code.
It has the same "bugs" as the HP-15C LE, but honestly it's really minor compared to its performance.

J-F

This wayback archive of the hp download page has the hp15c emulator together with several others for download:

https://web.archive.org/web/201307021354...lator.html

[Eric Rechlin]

(08-11-2021 07:22 AM)J-F Garnier Wrote:  As a side question, can the 2012 HP emulator for Windows of the HP-15C still be found?
It is no more on HP site, it's not on hpcalc.org.

Not true. It's on my HP Emulators torrent:

https://www.hpcalc.org/torrents/

[J-F Garnier]

(08-11-2021 01:19 PM)Eric Rechlin Wrote:  

(08-11-2021 07:22 AM)J-F Garnier Wrote:  As a side question, can the 2012 HP emulator for Windows of the HP-15C still be found?
It is no more on HP site, it's not on hpcalc.org.

Not true. It's on my HP Emulators torrent:
https://www.hpcalc.org/torrents/

Thanks, good to know even if I'm not familiar with torrents.

J-F