2 Jan
2012
2 Jan
'12
6:02 p.m.
On 12/31/2011 05:37 PM, Enrico Lazzerini wrote:
Hi all,
Some days ago I posted a request to identify the FCB parameters to read
correctly a 8" disk CP/M.
I know that i was wrong in the subject so I would like to summarize and i
thank you for the suggestions and criticisms.
The data obtained directly with the STAT command DSK: are the following:
9600: 128 Byte Record Capacity
1200: Kilobyte Drive Capacity
128: 32 Byte Directory Entries
128: Checked Directory Entries
128: Records / Extent
16: Records / Block
64: Sectors / Track
Thats logical sectors per track, physical sector size is likely 1024 bytes.
2: Reserved tracks
2 SIDES
After further investigation the parameters for the correct reading would
appear as follows:
BEGIN SCO2 (1024 bytes/sector) - 8 DSDD "
MFM DENSITY, HIGH
CYLINDERS 77 SIDES 2 SECTORS 8,1024
Side1 0 1,2,3,4,5,6,7,8
Side2 0 1,2,3,4,5,6,7,8
from 0 to 8 is NINE sectors. Save for a few oddbals
and some hard
sector drives
floppies do not have a sector zero(0). The standard is the first
sector after the index
home is sector 1 and is numbered 1. So for eight 1k sectors they would
be 1 through 8.
ORDER CYLINDERS
BSH 4 BLM 15 EXM 0 DSM 300 DRM 127 AL0 SFO 00H 11110000B AL1 2
END
The following additional information which I hope can clarify or help:
The inclusion of incorrect data was due to a misreading of my screen on the
console CP/M in the data down the email. The calculations were performed by
acquiring more information from the operating system CP/M manuals which
shows how to derive the parameters needed to compose and then subsequently
identified by extrapolating the FCB. The stat command dsk: when possible on
a working machine that is able to read the disk whose parameters are to be
identified can be useful, but if you can not read the disc seems to have
realized that the only way is to use anadisk and 22disk.
IN CP/M language the FCB is how you talk to the CP/M OS and has no direct
connection to the on disk organization of data.
FCB = File Control Block and is parameter passed to the BDOS (reference
by pointer).
and the BDOS grinds that in to a logical reference to the physical
structure using
tables and data blocks in the BIOS for that machine.
Anadisk reports:
Tracks 0 and 1 = single density 26x128byte with OS CP/M that points to a
resident BIOS involved in F800H entry points;
Tracks 2-76 = 8x1024byte double density with directory consists of 2 blocks
2048Byte for a total of 128 entries and the remaining space for data
SCO2 means disk2 SCOMAR machine (in Italy there was a PCB derived from the
Ferguson Bigboard I with a different 1797 FDC controller and BIOS and it has
been used to drive a knitting machine now discontinued and no longer
supported)
The diskette was created from a friend using software IMD of Dave Dunfield
and myself played on floppy 3.5 "1.44MB respecting the original features (or
almost I hope)
Logically it would seem you can substitute a HD 5.25 floppy for a 8" but
in practice
I've not seen it work yet.
With these parameters i was able to read and write to
the disk even it
remains some doubt as for example by adding the file to disk MBASIC.COM this
does not seem to properly turn on the machine that uses this disk format
The system needs the reserved tracks filled with the right stuff to boot
and run.
CP/M does nto boot from the directory as it needs CP/M running to even read
the directory.
Infact the main problem was that probabily the EXM value not make me able to
write files greater than 16KB and the picture visualized the 3 times the
same filename.
The way CP/M works is the EXM is the limit to the size of an individual
entry
and files larger use multiple directory entries to capture the whole file.
That means a 56K file needs 4 directory entries and shows up in the
directory 4 times
and part of the physical directory entry tells the OS there are multiple
and how they
are used in sequence.
So any kind of suggestion to understand betfer the FCB structure and
parameters of CP/M will be welcome.
Please read the Digital Research CP/M Alteration Guide.
The problem is if your trying to replicate the on disk format for a old
machine
you need the BIOS documentation for that machine or lacking that some
serious code disassembly if you have a disk that works. If neither exists
then your designing your own BIOS from the ground up and that may be
more than most can take on.
I've designed machines and the BIOS to match and it's a straight forward
task. But doing it for a machine by someone else without documentation
of what they did is far harder as that if forensic reconstruction.
Allison
I would like to thank:
- For suggestions
allison for the valuable explanations and clarifications which go beyond the
simple description inherent in any textbook
Cisin Fred for the tip to detect the possible size of the FCB parameter BLS
analyzing the disk
About received criticism may i ask for:
- Could anybody give me the name of sw to reduce the size of a photo?
- And what used to convert it into text easily? (I have not had a chance to
do a dump from the screen)
- Where can I buy the registered version of 22disk or is no longer available
or supported?
I apologize again for my poor English that puts me in trouble to fully
understand the criticism received.
Have an happy new year 2012 to all.
Enrico - Pisa - ITaly