DIAGNOSE instructions should be documented

[sdl-hercules-390]

DIAGNOSE instructions should be documented

[marXtevens]

Could you point in a general direction, or its equivalent to give me, or someone else a place to start looking?

[Fish-Git]

Hi Mark!

LOVE your handle! "MarXtevens". Heh! Pretty cool. :)

If you clone the repository (or just download the .zip) and look in file diagnose.c you'll see Hercules's code for processing the DIAG instruction (which of course, technically, doesn't really have a mnemonic).

Starting on line 108, we do a switch on the operand-2 value (effective_addr2) (I'm presuming you know 'C' code?). It's each one of these diagnose codes that I'd like to see documented.

You might have to chase down the helper sub-function some of the case: statements call into to actually process that particular diagnose (e.g. DIAG X'224' for example calls the ARCH_DEP(diag224_call) helper function to perform the actual processing for that particular diagnose code), which might get kind of complicated (or might not! I don't know!), so be prepared to possibly dig through a lot of code.

What I'd like to see is something that identifies each of the diagnose codes that Hercules supports, what its name is, what it's overall purpose is (if the purpose can't be determined from the name itself), and what the register requirements are, if any. I'm talking minimal information. You don't have to explain the nitty gritty details of the instruction (unless of course you feel they're important).

Don't bother formatting it fancy or anything. I'll do that myself when I put the information into a web page somewhere. Or you can take a crack at that if you want to. Whatever. I'm easy. I'm just looking to get the overall information right now. Formatting it into a readable format can come later.

Does that help?

Thanks!!

[marXtevens]

It's been a while. as recently I spend most of my time dealing with IT audits. :-(

However, I was an [SA|SysProg|Programmer|...] in a movie once, so I can muddle through Git, C, and anything else I get thrown into. I will start simple and expand as I can manage.

Long ago and far away, as a SysProg, I had trouble getting into various VM Listserv groups. It was shortly after the CHRISTMA EXEC hit a number of VM sites, that some guy (me) was asking for access with a username of XMAS. Using marXtevens has had less repercussions.

[marXtevens]

I have created a rough version tonight of the documentation. See if the files I have created are what you had in mind:

https://github.com/marXtevens/hyperion

html/hercdiag.html
html/index.html

[Fish-Git]

Where are they?! Your repository does not seem to have a hercdiag.html file, and its index.html file is identical to the existing one. I'm not seeing what you want me to see!

Did you maybe forget to commit your changes to your repository and/or forget to push your commit to GitHub?

[Fish-Git]

p.s. If you want (to make it easier for you), you can simply drag and drop the files into your reply. GitHub will automatically upload it to their server create a link to it.

[Fish-Git]

OR... You could create a GitHub "gist" too:

• About gists
• Creating gists

Either way: Push your changes. Drag and drop. GitHub gist.

I'm cool with all three.

[Fish-Git]

Or you could just email it to me: user name: fish, domain: infidels, TLD: com.
Or if you don't like that one: user name: fish, domain: softdevlabs, TLD: com.

[marXtevens]

Sorry for the delay, work and grandchildren visit got in the way. Sneaking time from them (grandkids) to try to get this updated on my repository. I "think" it is uploaded and commited on my respository. This is not RCS ;-)

[Fish-Git]

Found them!

But I'm not seeing any real documentation. Just fake "lorem ipsem" junk.

Reminder:

What I'd like to see is something that identifies each of the diagnose codes that Hercules supports, what its name is, what it's overall purpose is (if the purpose can't be determined from the name itself), and what the register requirements are, if any. I'm talking minimal information.

I'm just looking to get the overall information right now. Formatting it into a readable format can come later.

[marXtevens]

It took about 45 minutes to generate the HTML template directly from the C code. Took me longer to figure out Git. Used shell scripts rather than manual cut and paste, so I could ensure I had most, if not all of the calls. At the moment, I believe I am missing only the Diag 308. The fake documentation will be replaced, as I read and understand each routine. I'm not done, by a long shot, but wanted you to know I had started, and what I had after a "few" minutes of work. Does this help?

[Fish-Git]

Oh. Okay. Sorry. That's fine then. Carry on.

[Fish-Git]

@marXtevens

Ping ...

(How's it going, Mark? I hope my previous comments didn't scare you off the project?)

[marXtevens]

... Ping

Takes a lot more than that to scare me away!   ;-)

I am done with the skeleton and have started reading the code to create the descriptions. I keep getting interrupted by real work and family events.   :-D

[Fish-Git]

Ah. Okay. We're cool then. I understand about interruptions. Take your time. No rush. :)

[Fish-Git]

@marXtevens

Ping?

I'm not trying to rush you or anything! I know this is a "As I can find the time to work on it" type of project. I just like to check in on things every once in a while to see how things are going.

[marXtevens]

I have not given up. I have been busy with new home projects, and work. The good news, is I was able to understand most of the code for one of the calls, and had started documentation of it, in my "fork." I will make an effort to complete that one this weekend.

[Fish-Git]

@marXtevens

Ping?

How's it going? :)

(once again I am not trying to rush you and I don't mean to petster you either, but it's been over 4 months since my last ping so I thought I'd check with you to see how things were going)

[Fish-Git]

Closing due to non-response/inactivity/apparent disinterest.