ITS launch fail: SALVAGER.317

[Richard McDonald]

I have a PiDP-10 installation without a front panel running on a Pi5/8GB with trixie.

For a while it worked, but for some time now launching ITS fails immediately with SAVLVAGER.317 on the console.  

I have done git pull which resulted in nothing being pulled down.  And I have completely reinstalled a new /opt/pidp10 directory with the same results.

Anyone have any suggestions?

[Eric Swenson]

You may have roached your disk and have been thrown in the salvager.  Try to run the following:

CHKR$G

Where the $, above, is the escape key.  This will check your disks for consistency, offering to repair any problems.

— Eric

--
You received this message because you are subscribed to the Google Groups "PiDP-10" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pidp-10+u...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/pidp-10/a8426ca9-1ac8-46d3-93b7-cab141a522b6n%40googlegroups.com.

[Steven Hirsch]

It sounds like the boot disk volume is corrupted. Shut down:

pdpcontrol stop

Go to

/opt/pidp10/systems/its

Move these someplace safe:

dskdmp.rim
rp03.0
rp03.1
rp03.2
rp03.3

Then unzip new disk images:

unzip its-system.zip

Now try again. I've had this happen a couple of times. If you have
anything critical in the original images, perhaps one of the gurus can
help get it mounted as a secondary volume so you can retrieve it.


Steve

[AltairArchaeology]

So I had a similar failure about a week ago, but mine was caused by an FTP fail from transferring a file from Windows to ITS. This was what I got:

I was able to solve it by deleting all files in the user folder (username of REV) which allowed Salvager to delete the folder on the next startup. With the folder deleted, it removed the "extra garbage in UFD" and ITS booted right up. I wasn't able to delete the files using DKSDMP commands, so I ended up having ChatGPT write a small Python script to do it (see the attached if you are interested in the specifics; I had to provide it the .img files for analysis.) I ran it, it deleted all files in the directory, and Salvager took care of the rest. 

Since then, I make a backup of the image files any time I make major changes in ITS or try to do any projects. If it fails to boot, I just replace the images with the backup ones and continue from there. 

[Eric Swenson]

I’d offer the same advice to you! Run the CHKR program to check and fix your hard disk. When you are in SALVAGER, you type CHKR$G ($ is escape) and it will your disks and offer to fix stuff:

CHKR$G

    

    You will be asked for the active unit numbers.  For example, if you have

    units 0 through 5 you would enter 012345.  Or, for all eight drives, you

    would enter 01234567.

    You will be asked for the unit that will provide the MFD.  Answer 0.

    You will be asked if you want to read all the blocks from all files.

    This is fairly quick in most cases, so it is recommended that you

    answer Y.

    You will be asked which unit to get the user directories from.  Answer 0.

    Finally, you will be asked if the changes should be written out.

    Answer Y.

— Eric (KC6EJS)

On Jun 1, 2026, at 22:25, AltairArchaeology <altairar...@gmail.com> wrote:

So I had a similar failure about a week ago, but mine was caused by an FTP fail from transferring a file from Windows to ITS. This was what I got:

--

You received this message because you are subscribed to the Google Groups "PiDP-10" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pidp-10+u...@googlegroups.com.

To view this discussion visit https://groups.google.com/d/msgid/pidp-10/10a7ae92-9d5d-4f54-a85d-409bc7b766dfn%40googlegroups.com.

<fix_rev_empty_ufd.py>

<20260525_223912.jpg>

[Richard McDonald]

Hey Eric,

Thanks for the reply.  When I get the SALVAGER.317 message, I am unable to type CHKR$G into the console window (at least nothing is echoed back).

Is there a special sauce/incantation to use at the DKSDMP prompt?

Cheers,

Rich

WB0NRE

[Eric Swenson]

Can you type <esc>U to enter DSKDMP?

If you can get to DSKDMP, then you can type

SALV<enter>

And then

CHKR$G

— Eric

To view this discussion visit https://groups.google.com/d/msgid/pidp-10/668de520-5f73-4157-9d24-1127ab5724aen%40googlegroups.com.

[Richard McDonald]

OK, I was not able to <esc>U and get anything; so I pdpcontrol stop, waited some, ptpcontrol start 1, waited for type340 window to open, then opened a PDP-10 View window.

I then entered SALV<enter>CHKR$G and got the prompts you mentioned earlier.  Turns out I only have units 0 thru 3.

When I answered 0 for "GET USR DIRS FROM UNIT:" things seem to hang.  I noticed some activity in the SIMH window which showed:  <see simh.txt>

I;m not sure how much of that was recent, as I just noticed it out the corner of my eye.

Cheers,

Rich

WB0NRE

[Eric Swenson]

What did you type to the prompt “GET USR DIRS FROM UNIT”?  You should have typed “0” (no <enter>). That should have gotten the user directories from disk 0.  If the user directories on unit 0 were damaged, you could have used 1 or 2, since these are copied onto each unit, I think. — Eric

To view this discussion visit https://groups.google.com/d/msgid/pidp-10/f0244dee-ff2b-4bc8-8b30-013252621fden%40googlegroups.com.
<simh.txt>

[AltairArchaeology]

I went at it a bit of a different way to try it out. Didn't need to do it, but it's good practice. :) 

- Boot as normal (pdpcontrol 1, stop, read in, etc. I have the physical PiDP-10 hardware so I just use the switch for pdpcontrol select.)

- Once DSKDMP appears on the TTY prompt type in ($ = esc):

     - L$DDT <RETURN>
     - T$SALV BIN <RETURN>
     - CHKR$G <RETURN> 

- It'll then ask you some questions. This worked on my standard ITS images:

Active unit numbers? 0123 <RETURN> (You can type in more; it ignores any units that are inactive and proceeds. The default numbers are 0123 for the stock images)

Use MFD from unit: 0 <RETURN>
Read all blocks of all files? Y <RETURN>
Get user dirs from unit: 0 <RETURN>

Mine immediately went into deleting a .temp. folder that is normally done each time at the start of ITS' boot sequence. 

When it asks to write MFD changes: Y <RETURN>

Once complete (I throttle mine to 1400K; I recommend NOT doing that for this sequence as it takes quite a while), type in the following:

- $U <RETURN> 

This takes you from DDT back to DSKDMP. From there you can boot ITS as normal.

- ITS <RETURN>

- $G

[AltairArchaeology]

Realized with my past one some of the commands don't require a <RETURN>. Don't have the unit in front of me, but with some of them that automatically continue (Get user dirs from unit) just enter the number. Only press <RETURN> if it stays there for a few seconds. 

[Eric Swenson]

Are you sure you need a <return> after those prompts? I haven’t run the salvager in a very long time, but I thought it acted on the first digit/character typed for each.

— Eric (KC6EJS)

[AltairArchaeology]

So I covered this in my previous message, but for clarity for future readers:

I went at it a bit of a different way to try it out. Didn't need to do it, but it's good practice. :) 

- Boot as normal (pdpcontrol 1, stop, read in, etc. I have the physical PiDP-10 hardware so I just use the switch for pdpcontrol select.)

- Once DSKDMP appears on the TTY prompt type in ($ = esc):

     - L$DDT <RETURN>
     - T$SALV BIN <RETURN>

     - CHKR$G <Will proceed automatically>

- It'll then ask you some questions. This worked on my standard ITS images:

Active unit numbers? 0123 <RETURN> (You can type in more; it ignores any units that are inactive and proceeds. The default numbers are 0123 for the stock images)

Use MFD from unit: 0 < Will proceed automatically>
Read all blocks of all files? Y <Will proceed automatically>
Get user dirs from unit: 0 <Will proceed automatically>

Mine immediately went into deleting a .temp. folder that is normally done each time at the start of ITS' boot sequence. 

When it asks to write MFD changes: Y <Will proceed automatically>

Once complete (I throttle mine to 1400K; I recommend NOT doing that for this sequence as it takes quite a while), type in the following:

- $U <Will proceed automatically>   

This takes you from DDT back to DSKDMP. From there you can boot ITS as normal.

- ITS <RETURN>

- $G

[Eric Swenson]

Thanks for the corrections. This matches my expectations and experience. 

While the person who first had the issue could well recover by reverting to a saved copy of the disk, as someone suggested, the “right” fix (and what would have been done “back in the day” on ITS, is to do the equivalent of the steps you outlined, to salvage the disk and not lose any data by reverting to a disk backup (which wasn’t actually done much back then) or a full backup of the disk to tape (which would have needed lots of tapes, including all the incremental backup dumps since the last full dump was done).

I highly recommend people learn how to run and administer ITS, including salvaging and running backups, rather than simply throwing away their disks and reverting to a host disk copy.

I say this, as a person who used ITS “back in the day”!

— Eric (KC6EJS)

On Jun 2, 2026, at 21:06, AltairArchaeology <altairar...@gmail.com> wrote:

So I covered this in my previous message, but for clarity for future readers:

--

You received this message because you are subscribed to the Google Groups "PiDP-10" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pidp-10+u...@googlegroups.com.

To view this discussion visit https://groups.google.com/d/msgid/pidp-10/49f5ce6a-c872-42b5-b415-5bb015e7cec7n%40googlegroups.com.

[Richard McDonald]

Many thanks to @AltairArchaeology and Eric for your suggestions and patience.  So far, nothing gets past "GET USR DIRS FROM UNIT:0" when the system sends a <CR><LF> and appears to hang.  I even left it overnight, thinking it was silently working hard.  I tried using "GET USR DIRS FROM UNIT:" with 0, 1, 2, and 3 separately with the same null result.

I have tried a whole new installation with fresh git downloads of /opt/pidp10; all with the same SAVLVAGER.317 result.  I am beginning to think this may be related to my trying to use the latest Raspi OS Trixie.  Which underlying OS version are you using to run SIMH?

Cheers,

Rich

(WB0NRE)

[Eric Swenson]

If salvager prints a CR/LF it is done with no errors. Do $U to get to DSKDMP and “its<enter>” to load ITS and the $g to start it.

— Eric (KC6EJS)

On Jun 3, 2026, at 09:36, Richard McDonald <ranc...@gmail.com> wrote:



To view this discussion visit https://groups.google.com/d/msgid/pidp-10/441d8c06-1664-47f6-84ed-4f00687acdbcn%40googlegroups.com.

[Richard McDonald]

After SALVAGER prints <cr><lf>, esc u does nothing.

In looking closely at the SIMH window when SALVAGER seems to hang I see a couple of SIMH prompts.  So I took a shot in the dark and typed continue in the SIMH window.  Lo and behold something happened.  

I'm thinking something is dropping simh into a single-step mode.  Since I don't have the blinkenlights perhaps and open circuit pin is fooling something.  Thoughts?

Sent from my T-Mobile 5G Device
Get Outlook for Android

---

From: Eric Swenson <er...@swenson.org>
Sent: Wednesday, June 3, 2026 9:47:39 AM
To: Richard McDonald <ranc...@gmail.com>
Cc: PiDP-10 <pid...@googlegroups.com>
Subject: Re: [pidp-10] ITS launch fail: SALVAGER.317

 

[Eric Swenson]

For that, you’ll need a PiDP-10 person (I’m just an ITS person!).

And when I reacted to SALVAGER printing CRLF, I actually meant after a CHKR$G command.  Once it finishes the prompts and starts doing it work, it “returns to DDT” by printing CRLF.  I’m guessing you were referring to the CRLF it prints out after displaying the initial “SALVAGER 364” (or whatever your version of salvager is).  In that case, you’re right, something is causing you to return to SIMH. I’m surprised there isn’t a SIMH prompt though.

— Eric

[Lars Brinkhoff]

I'm thinking something is dropping simh into a single-step mode.

It could have been a HALT instruction (i.e. JRST 4,).  If that is the case, the program probably encountered a fatal condition from which it can't proceed.  You should note the address of the HALT and figure out where in the program it is located.