Copy of real disk cartridge image from my 1130

[Carl Claunch]

I ran a DCIP to copy one of my system cartridges, mounted in my internal drive on the 1130, to a blank cartridge mounted on my virtual 2310 disk drive. 

This can be used with the IBM 1130 Simulator

I put it up on Google Drive so that anyone can download a copy  Disk image to download

This should help resolve questions about the Fortran library and extended precision. 

Carl

[Bob Flanders]

Thank you, Carl.

--
You received this message because you are subscribed to the Google Groups "IBM1130" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ibm1130+u...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Al Kossow]

Carl, do you want to forward this on to the SIMH group?
If you don't want to support it, I can copy it over to bitsavers.org/bits/IBM/1130

On 7/27/16 8:24 AM, Carl Claunch wrote:
> I ran a DCIP to copy one of my system cartridges, mounted in my internal drive on the 1130, to a blank cartridge mounted
> on my virtual 2310 disk drive.
>
> This can be used with the IBM 1130 Simulator
>
> I put it up on Google Drive so that anyone can download a copy Disk image to download

> <https://drive.google.com/open?id=0B9_ekls8yE0fUWNwWWRkVmpfdUk>

>
> This should help resolve questions about the Fortran library and extended precision.
>
> Carl
>

> --
> You received this message because you are subscribed to the Google Groups "IBM1130" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to ibm1130+u...@googlegroups.com

> <mailto:ibm1130+u...@googlegroups.com>.

[Carl Claunch]

Hi Al

I can't guarantee a perpetual online copy, but happy to share it with anyone. Feel free to copy it over to Bitsavers. 

Carl

[Walter T. Mosscrop]

Hi Carl -

 

Does this image support the 2501 and 1403? My emulator does not support the 1442 and 1132.

 

Thanks,

 

Walter

To unsubscribe from this group and stop receiving emails from it, send an email to ibm1130+u...@googlegroups.com.

[Al Kossow]

On 7/27/16 4:09 PM, Carl Claunch wrote:
> Hi Al
>
> I can't guarantee a perpetual online copy, but happy to share it with anyone. Feel free to copy it over to Bitsavers.
>

will do

[Carl Claunch]

Hi Walter

It does not, but it would be a simple matter to run a System Reload to configure the IO devices.

If you are not sure how to do this, let me know, along with the memory config you want, and I will give you a modified version for 2501 and 1403 input output.

Carl

[Walter T. Mosscrop]

Hi Carl –

 

I would greatly appreciate it if you would do so for a 8K word system. It’s been quite a while since I generated my image and I can’t find the files I used.

[Richard Stofer]

Carl,

 

If you’re making custom images, I could test a 32k 2501 1403 variant.

[Carl Claunch]

Hi Walter

Here is a link to an 8K version of the disk with 2501 reader and 1403 printer as peripherals. https://drive.google.com/open?id=0B9_ekls8yE0fRlZ6cnJlVDFPR28

Carl

[Carl Claunch]

Hi Richard

A 32K version of the disk with 2501 reader and 1403 printer is at this link https://drive.google.com/open?id=0B9_ekls8yE0fenVBRUx4UVVMMXM 

Carl

[Walter T. Mosscrop]

Hi Carl –

 

For some reason the image won’t boot. I have a custom emulator; I have no idea at this time what the issue is but I suspect it’s with my emulator and not your image. It’s been a while since I’ve debugged something like this so I’m going to try to do it this weekend.

 

Thanks for the quick response. Time to dust off my assembler manual and see what’s going on…

[Richard Stofer]

Walter,

 

Bet on the problem being in your emulator.  Carl’s image works on my FPGA as does Brian’s DMS.  Notwithstanding these two little problems ALOG and STORECI.

 

ALOG (ELN) work on Carl’s image.  STORECI does not.

[Richard Stofer]

That is, STORECI does not work for me!  I don’t know if it works for others.

[Brian Knittel]

I noticed that the STORECI issue is noted in the ibm1130.pdf manual that I included with the emulator.

It was noted yesterday that out appears to involve printing. I'm sure it will be tracked down soon. It had to be working in most real world 1130 installations.

[Eddy Quicksall]

If I remember correctly the IBM1130.ORG cartridge has cylinder 0 (maybe just sector 0) as little endian while the rest is big endian. The IBM cartridges are all big endian. Could that be your problem?

 

I posted a cartridge from my system which is all big endian … but since my emulator must support non IBM DSM’s then I don’t have the sector address before the data. You may not be able to boot from mine because all of the drivers are for TSO but it would be worth a quick try.

 

Others may want to pipe in and give more detail.

 

Eddy

 

From: ibm...@googlegroups.com [mailto:ibm...@googlegroups.com] On Behalf Of Walter T. Mosscrop

Sent: Thursday, July 28, 2016 6:22 PM
To: ibm...@googlegroups.com

[br...@quarterbyte.com]

> If I remember correctly the IBM1130.ORG cartridge has cylinder 0 (maybe just sector 0) as little endian while the rest is big endian. 

That would be surprising. The DMS disk image is created by running IBM's own 1130 code on the emulated CPU, the image is not created by any standalone non-1130 code. In turn, all 1130 disk IO is done by XIO commands, and all disk XIO reads and writes are performed by simh functions fxread and fxwrite. None of the code treats any sector or cylinder any differently than any other. fxread and fxwrite are used so that the same disk image file can be used on big-endian and little-endian machines interchangeably, and these routines adjusts byte order before writing or after reading (if necessary) so that words are interpreted correctly was words on whatever architecture simh is running on.

Brian

[br...@quarterbyte.com]

The extended LN program appears to have been solved by fixing the extended floating constant problem in asm1130.

I reassembled and reloaded all of DMS after fixing the cross assembler and the ln test program  no longer crashes. I suspect that routines other than EALOG were also affected by the extended constant bug.

Formal redistrubution will come soon, for now, you guys can play with this:

zip of revised DMS R2V12 disk image

modified asm1130.c source code

asm1130.ex the compiled assembler, you have to change extension from .ex to .exe to satisfy Google.

Storeci not fixed by this, as expected

brian

[John R Pierce]

On 7/28/2016 5:57 PM, br...@quarterbyte.com wrote:

> If I remember correctly the IBM1130.ORG cartridge has cylinder 0 (maybe just sector 0) as little endian while the rest is big endian. 

That would be surprising. The DMS disk image is created by running IBM's own 1130 code on the emulated CPU, the image is not created by any standalone non-1130 code. In turn, all 1130 disk IO is done by XIO commands, and all disk XIO reads and writes are performed by simh functions fxread and fxwrite. None of the code treats any sector or cylinder any differently than any other. fxread and fxwrite are used so that the same disk image file can be used on big-endian and little-endian machines interchangeably, and these routines adjusts byte order before writing or after reading (if necessary) so that words are interpreted correctly was words on whatever architecture simh is running on.

anyways, the 1130 was a 16 bit word addressing machine, there wasn't really any 'endian' about it in its software, it was word based.   now, the disk drive hardware naturally reads/writes bit serial data, so I suppose it matters if the bits are written from LSB or MSB first, that I can't tell you, although I suspect it was MSB first.

-- 
john r pierce, recycling bits in santa cruz

[Roland Langfeld]

Dear Brian,

Indeed, now everything is running smoothly with extended precision !! Thankyou and all the other members in this forum for fixing this problem so fast and so professionell !!

Roland

[Eddy Quicksall]

Brian

 

I noticed the endian issue a couple of years ago when I was trying my Windows File System Driver for the DM2 file system on an 1130.org cartridge. I will take a look at that code again and refresh my memory.

 

BTW, I thought you said you had to do some big/little conversion to my cartridge file. I thought it was due to this.

 

Eddy

[Eddy Quicksall]

Yes, it was MSB first.

 

From: ibm...@googlegroups.com [mailto:ibm...@googlegroups.com] On Behalf Of John R Pierce
Sent: Thursday, July 28, 2016 9:37 PM
To: ibm...@googlegroups.com
Subject: Re: [IBM1130] Copy of real disk cartridge image from my 1130

 

On 7/28/2016 5:57 PM, br...@quarterbyte.com wrote:

--

[Eddy Quicksall]

This thread was to address another emulator that won’t boot and I just wanted to pass this along to him so he could consider it.

 

I remembered wrong. I remembered “sector 0” because that was the quickest way to see why Windows file system driver did not work with your disk image. It turns out that the whole image is little endian. To compensate for this my driver first reads sector 0 and if I see 58 06 instead of 06 58 then I set a flag so it will properly read LET/FLET and all files being accessed.

 

I just interpret XIO’s too and when the read is started, I read using the C function read() so what comes into memory is just one byte after another.

 

Here is the first part of sector 0.

[Eddy Quicksall]

We called it big endian … even from the get go in the early 60’s. That came up several times in discussions since most computers by then were little endian.

 

For example, in memory (and on disk) the value 0658 was in memory and disk that way … 06 58. That is important for some Fortran programs which used EQUIVALENCE and to get to the separate bytes. It was funky and I hated it (also the backward running arrays). I think this is an example (the int’s here are defaulting to TWO WORD INTEGERS):

 

      INTEGER CARD(80),LMAJ(100),LDET(100),HMAJ(100),HDET(100)

      INTEGER FROM,RNG(100,2,2),RPT,COS(360),CPT

      EQUIVALENCE (RNG(1,1,1),LMAJ),(RNG(1,1,2),LDET),

     *            (RNG(1,2,1),HMAJ),(RNG(1,2,2),HDET)

. . .

      CALL MASI(CARD, 3, 6,LMAJ(RPT))

      CALL MASI(CARD, 7,10,LDET(RPT))

      CALL MASI(CARD,12,15,HMAJ(RPT))

      CALL MASI(CARD,16,19,HDET(RPT))

 

From: ibm...@googlegroups.com [mailto:ibm...@googlegroups.com] On Behalf Of John R Pierce

Sent: Thursday, July 28, 2016 9:37 PM
To: ibm...@googlegroups.com

Subject: Re: [IBM1130] Copy of real disk cartridge image from my 1130

 

On 7/28/2016 5:57 PM, br...@quarterbyte.com wrote:

--

[Eddy Quicksall]

I thought about this. Was your program written with Fortran?

 

Mine came out big endian because a tape backup was made using the 1130 then that was read back using DOS. A DOS program was then used to transfer the cartridge over to an emulated DSM (at the time we were using 300MB CDC or Calcomp DSMs on the 1130). That was way back around ’95 so it is hard for this old guy to get that out of long term storage in my brain).

 

Eddy

 

From: ibm...@googlegroups.com [mailto:ibm...@googlegroups.com] On Behalf Of br...@quarterbyte.com

Sent: Thursday, July 28, 2016 8:58 PM
To: ibm...@googlegroups.com

[Eddy Quicksall]

We called it big endian … even from the get go in the early 60’s. That came up several times in discussions since most computers by then were little endian.

 

For example, in memory (and on disk) the value 0658 was in memory and disk that way … 06 58. That is important for some Fortran programs which used EQUIVALENCE and to get to the separate bytes. It was funky and I hated it (also the backward running arrays). I think this is an example (the int’s here are defaulting to TWO WORD INTEGERS):

 

      INTEGER CARD(80),LMAJ(100),LDET(100),HMAJ(100),HDET(100)

      INTEGER FROM,RNG(100,2,2),RPT,COS(360),CPT

      EQUIVALENCE (RNG(1,1,1),LMAJ),(RNG(1,1,2),LDET),

     *            (RNG(1,2,1),HMAJ),(RNG(1,2,2),HDET)

. . .

      CALL MASI(CARD, 3, 6,LMAJ(RPT))

      CALL MASI(CARD, 7,10,LDET(RPT))

      CALL MASI(CARD,12,15,HMAJ(RPT))

      CALL MASI(CARD,16,19,HDET(RPT))

 

From: ibm...@googlegroups.com [mailto:ibm...@googlegroups.com] On Behalf Of John R Pierce

Sent: Thursday, July 28, 2016 9:37 PM
To: ibm...@googlegroups.com

Subject: Re: [IBM1130] Copy of real disk cartridge image from my 1130

 

On 7/28/2016 5:57 PM, br...@quarterbyte.com wrote:

--

[Walter T. Mosscrop]

Hi Carl –

 

It appears that my issue with the disk image you provided is that it is trying to access the 1132 printer.

 

The last XIO device code accessed before the boot halts is 00110, which is the 1132. The disk image that I normally use (based on the IBM1130.org image) doesn’t attempt to access the 1132.

 

Is it possible you provided an image built for the 1442 and 1132 instead of the 2501 and 1403? My emulator does not support the 1132 or 1442.

 

Thanks,

 

Walter