Hepax/Zenrom conflict

[jwhsu]

Combination of Hepax from Ángel’s files and Zenrom from hp41.org causes calc to freeze, requiring a hard reset. Load order on active module page is 1) Zenrom, 2) Hepax. Catalog-2 afterwards shows both modules installed with Hepax listing before Zenrom. HEPDIR showed “No Filesystem”. Turning the calc off then on results in a freeze that requires a hard reset. Tried reversing module order, same problem. This combination does work on a 41CL (with YNFX and PowerCL extreme installed as well).

[jwhsu]

here is the rominit.log sequence 1) Zenrom, 2) Hepax:

Code: Select all

0-0: SM_NUT-CXX.mod pg:0
 R[0] 0 0 1 0 NUT0-N
 R[1] 1 0 1 0 NUT1-F
 R[2] 2 0 1 0 NUT2-L_X
 R[3] 3 0 1 1 -EXT FCN 2X
 R[4] 5 0 1 1 -TIME 2X
 R[5] 5 0 2 1 -X/F_BKN2
0-1: ZENROM.MOD pg:9
0-2: HEPAX_4H.MOD pg:8
1-0: SM_NUT-CXX.mod pg:0
1-1: ZENROM.MOD pg:9
1-2: HEPAX_4H.MOD pg:8
 F-PG[R0]: 8
 R[0] 8 1 1 3 -HEPAX`4H
 R[1] 8 1 2 3 -HEPAX_BK2
 R[2] 8 1 3 3 -HEPAX 4H
 R[3] 8 1 4 3 -HEPAX_BK4
 RAM[4] 8 WP=0
 RAM[5] 9 WP=0
2-0: SM_NUT-CXX.mod pg:0
2-1: ZENROM.MOD pg:9
 F-PG[R0]: 9
 R[0] 9 0 1 0 Zenrom-3B
2-2: HEPAX_4H.MOD pg:8

This is the rominit.log for sequence 1) Hepax, 2) Zenrom

Code: Select all

0-0: SM_NUT-CXX.mod pg:0
 R[0] 0 0 1 0 NUT0-N
 R[1] 1 0 1 0 NUT1-F
 R[2] 2 0 1 0 NUT2-L_X
 R[3] 3 0 1 1 -EXT FCN 2X
 R[4] 5 0 1 1 -TIME 2X
 R[5] 5 0 2 1 -X/F_BKN2
0-1: HEPAX_4H.MOD pg:8
0-2: ZENROM.MOD pg:9
1-0: SM_NUT-CXX.mod pg:0
1-1: HEPAX_4H.MOD pg:8
 F-PG[R0]: 8
 R[0] 8 1 1 2 -HEPAX`4H
 R[1] 8 1 2 2 -HEPAX_BK2
 R[2] 8 1 3 2 -HEPAX 4H
 R[3] 8 1 4 2 -HEPAX_BK4
 RAM[4] 8 WP=0
 RAM[5] 9 WP=0
1-2: ZENROM.MOD pg:9
2-0: SM_NUT-CXX.mod pg:0
2-1: HEPAX_4H.MOD pg:8
2-2: ZENROM.MOD pg:9
 F-PG[R0]: 9
 R[0] 9 0 1 0 Zenrom-3B

It looks like the Zenrom module is plugged in before the RAM pages are initialized, leading to a port conflict. Not sure how to fix this.

[Ángel Martin]

I also see the same problem, trying to understand the root cause now...

On V41 the HEPAX will only behave properly if it is "plugged" in an even page (8, A, C, E). This is because V41 doesn't seem to support the BLKMOV instruction properly, which is called during the initialization. The FileSys routine is then executed automatically, and it needs the HEPAX ROM to be in an even page as just mentioned. That not being the case it locks up V41 as well, you can try it by plugging the ZENROM first and the HEPAX later with pages #6 and #7 used by other modules, say the HP-IL.

On the CL none of this matters, because the file sys initialization is disabled (the CALC_ON polling point is void), and the filesystem initialization is made manually using HEPINI.

But the 41X does support BLKMOV, so why does it fail on it? We know the HEPAX is properly loaded with other modules, including those with CALC_ON events and even I/O interrupts like the CCD, AMC_OSX, SandMath, etc. Therefore there must be something unique to the ZenROM that makes it incompatible with the load_Module process on the 41X. It doesn't use any RAM pages, but it has a I/O interrupt itself, so I wonder if that may be the issue.

I think this needs to be investigated further, at the moment i don't have an explanation...

_________________________________________________________________________________
DM-41X: 00041; DM-42: 00042

[rprosperi]

HEPAX/ZenROM problem found and fixed. An updated BETA build will posted in the next couple days.

Please don't waste bandwidth hitting refresh, these bug fixes are not so exciting

[akaTB]

HEPAX/ZenROM problem found and fixed. An updated BETA build will posted in the next couple days.

Please don't waste bandwidth hitting refresh, these bug fixes are not so exciting

What a service! (2)

[jwhsu]

HEPAX/ZenROM problem found and fixed. An updated BETA build will posted in the next couple days.

Please don't waste bandwidth hitting refresh, these bug fixes are not so exciting

Thank you. Cannot wait to try it out.