[Simh] Converting VMS BACKUP saveset file to SIMH tape file
Rob Jarratt
robert.jarratt at ntlworld.com
Tue Mar 30 14:14:58 EDT 2010
Yes, now posted:
https://cid-fc758a5a91b91301.skydrive.live.com/browse.aspx/Public/VirtualTap
e
Regards
Rob
> -----Original Message-----
> From: simh-bounces at trailing-edge.com [mailto:simh-bounces at trailing-
> edge.com] On Behalf Of Armistead, Jason
> Sent: 29 March 2010 13:35
> To: simh at trailing-edge.com
> Subject: Re: [Simh] Converting VMS BACKUP saveset file to SIMH tape
> file
>
> Hi Rob
>
> Is there source available for this program ? I'd like to understand
> how it ticks under the hood.
>
> Thanks,
> Jason
>
> -----Original Message-----
> From: simh-bounces at trailing-edge.com [mailto:simh-bounces at trailing-
> edge.com] On Behalf Of Rob Jarratt
> Sent: Sunday, 28 March 2010 3:00 PM
> To: 'Edward Berry'; simh at trailing-edge.com
> Subject: Re: [Simh] Converting VMS BACKUP saveset file to SIMH tape
> file
>
> The program I wrote will take a binary file and make a SIMH .TAP file
> out of
> it. I have used this to make BACKUP tapes from BACKUP files I have on
> the
> PC.
>
> I have posted the executable and its DLL here:
> https://cid-
> fc758a5a91b91301.skydrive.live.com/browse.aspx/Public/VirtualTap
> e. please note that the program requires the .NET Framework 2.0 to be
> installed.
>
> To be honest I have forgotten how to use it. I think you select the
> name of
> the .TAP file, select a label and then press the Write ANSI button to
> select
> the files to put on tape.
>
> Regards
>
> Rob
>
> > -----Original Message-----
> > From: simh-bounces at trailing-edge.com [mailto:simh-bounces at trailing-
> > edge.com] On Behalf Of Edward Berry
> > Sent: 28 March 2010 19:25
> > Cc: simh at trailing-edge.com
> > Subject: Re: [Simh] Converting VMS BACKUP saveset file to SIMH tape
> > file
> >
> > Rob Jarratt wrote:
> > > I have some C# code which may do this. If I can find the code would
> > you
> > > like to give it a go?
> > >
> > > Regards
> > >
> > > Rob
> >
> > I would also be interested in getting the code if it resurfaces!
> > Does it make a disk file to mount on simh, or a saveset to FTP into
> > the simulated filesystem and read as a disk-based saveset?
> >
> >
> > I've been dd'ing the savesets from exabyte tapes written by VMS
> > to linux (which I realize may not be the same thing as FTP'ing a
> disk-
> > based saveset)
> > This requires dividing the file into blocks (see simh_magtape.pdf
> > on trailing edge website). For small savesets I was able to use
> > Helbig's enblock.c:
> >
> > http://www.ba-stuttgart.de/~helbig/os/v6/enblock.c
> > with some description at
> > http://minnie.tuhs.org/pipermail/tuhs/2003-June/000612.html
> >
> > but for large files it was flakey - saveset came out a different size
> > each time I ran it.
> > Then I found that using fortran to read the saveset image unformatted
> > with the correct record length and write it out again, has the same
> > effect of blocking the records.
> >
> > Then the saveset had to be combined with its header and footer (also
> > dd'd
> > from the vaxtape), and if it was not the first saveset the header had
> > to be prepended with the tape label from the first set. I realize
> > this is quite different from what the original poster is doing,
> > but in case anyone needs to restore vax tapes to disk images for
> > simh I'll put the procedure and code below.
> >
> >
> >
> =======================================================================
> > ======
> > Restoring vax tapes for simh on a linux machine
> > Tape drive for reading the tapes: exabyte eliant
> >
> > script for reading tape files to disk (cd to an empty directory
> before
> > running)
> > -----------------------------------
> > ddvaxtape.csh:
> > #!/bin/csh -ef
> > #mt -f /dev/nst0 rewind
> > mt -f /dev/nst0 setblk 0
> > foreach nmbr (1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
> 22
> > \
> > 23 24 25 26 27 28 29 30)
> > dd if=/dev/nst0 bs=80 of=hdr${nmbr}.dat
> > if $nmbr == 1 then
> > head -c40 hdr${nmbr}.dat |awk '{print $1}'
> > head -c160 hdr${nmbr}.dat |tail -c80 |awk '{print $1,$2}'
> > else
> > head -c40 hdr${nmbr}.dat |awk '{print $1,$2}'
> > #cat hdr${nmbr}.dat |awk '{print $1,$2}'
> > endif
> > #dd if=/dev/nst0 bs=32768 of=saveset${nmbr}.dat
> > dd if=/dev/nst0 bs=65536 of=saveset${nmbr}.dat
> > head -c428 saveset${nmbr}.dat |tail -c135
> > #head -c444 saveset1.dat |tail -c151
> > echo
> > #dd if=/dev/nst0 bs=0 of=saveset${nmbr}.dat
> > dd if=/dev/nst0 bs=80 of=tail${nmbr}.dat
> > mt -f /dev/nst0 status
> > #if $status break
> > end
> > exit
> > If the tape has more than 30 savesets, after it finishes
> > create a subdirectory and run the program again there
> > without the rewind instruction.
> >
> > To list the savesets in each file:
> > grep EOF tail*.dat
> >
> > tail10.dat:EOF1RAXISD1. BCD3 00010010000100 91353 91353
> > 001208DECVMSBACKUP
> > EOF2F3276832768 M 00
> > tail11.dat:EOF1RAXIS2.BCK BCD3 00010011000100 92024 92024
> > 001208DECVMSBACKUP
> > EOF2F3276832768 M 00
> > tail12.dat:EOF1BCIIA.MRC BCD3 00010012000100 92025 92025
> > 000311DECVMSBACKUP
> > EOF2F3276832768 M 00
> > ------------------------------------
> > This produces 3 files for each saveset: hdr##.dat, saveset##.dat, and
> > tail##.dat
> > The first header is longer because it starts with the tape label.
> > The saveset##.dat files can be listed and extracted on the linux
> > host using the vmsbu utility.
> >
> > To use with simh they need to be blocked and recombined with tape
> > label, head, and tail.
> > http://www.ba-stuttgart.de/~helbig/os/v6/enblock.c will block small
> > files
> > (or maybe files with small reclength), but for the savesets with
> > reclength 32768 I used a fortran snippet (real hack, the file names
> are
> > coded in!
> > Also I think you need to redimension (character*32768 line) to match
> > the
> > reclen param the tape was written with.):
> >
> > c****writing in unformatted mode automatically wraps each record in
> 4-
> > byte integers giving
> > record size.
> > c record size can be variable
> > c
> > INTEGER*4 IR
> > character*32768 line
> > OPEN (UNIT=2,FILE='saveset19.dat',STATUS='OLD',recl=32768,
> > . form='unformatted',access='direct')
> > OPEN (UNIT=3,FILE='saveset19.tpe',STATUS='unknown',
> > . form='unformatted',access='sequential')
> > DO 77 IR = 1,50000
> > 3 READ(2,rec=ir,err=777) LINE
> > write(6,*) ir
> > write(unit=3) (LINE)
> > 77 continue
> > 777 close (3)
> > end
> >
> > (This works when compiled with g77 fortran. I just tried gfortran and
> > the file
> > comes out larger by about one byte per record, so I suppose it won't
> > work)
> >
> >
> > now putting it all together ("enblock" here is Helbig's enblock, and
> > saveset19.tpe is the output of the fortran blocking):
> >
> > #get the tapelabel, from first header, blocked, first 88 bytes:
> > cat hdr1.dat | enblock 80 > temp.dat
> > dd if=temp.dat count=88 bs=1 > $syn/simh/VAX/allback.tpe
> > #append the appropriate header:
> > cat hdr19.dat |enblock 80 >> $syn/simh/VAX/allback.tpe
> > #append the saveset:
> > cat saveset19.tpe >> $syn/simh/VAX/allback.tpe
> > #put end-of-file label?:
> > dd if=/dev/zero bs=1 count=4 >> $syn/simh/VAX/allback.tpe
> > #append the tail
> > cat tail19.dat | enblock 80 >> $syn/simh/VAX/allback.tpe
> > #put extra 4 zero's for end-of-tape
> > dd if=/dev/zero bs=1 count=4 >> $syn/simh/VAX/allback.tpe
> >
> > #mount it on simh!
> > attach tq2 allback.tpe
> > c
> > BACKUP/List/rewind mua2:
> > _______________________________________________
> > Simh mailing list
> > Simh at trailing-edge.com
> > http://mailman.trailing-edge.com/mailman/listinfo/simh
>
> _______________________________________________
> Simh mailing list
> Simh at trailing-edge.com
> http://mailman.trailing-edge.com/mailman/listinfo/simh
> _______________________________________________
> Simh mailing list
> Simh at trailing-edge.com
> http://mailman.trailing-edge.com/mailman/listinfo/simh
More information about the Simh
mailing list