Hi to all,

Ive got a problem with finding out the TracksPerHead value by calling Int13h ah=08h. Until now, I found out the following:

Value in cx after call:
4F42h
-------------ch----- ----cl------
cly. bit: 76543210 98
sect. bit:---------------543210

Expected value in DriveInfo.TracksPerHead:
14Fh
Value in DriveInfo.TracksPerHead:
0
(if I leave away the inc at the end of the procedure it is Fh)

f_read_driveinfo:
push es
mov dl, [DriveInfo.DriveNumber]
mov ah, 8h
int 13h
xor eax, eax
mov [DriveInfo.Heads], eax
mov [DriveInfo.TracksPerHead], eax
mov [DriveInfo.SectPerTrack], eax

mov al, dh
mov WORD [DriveInfo.Heads], ax

push cx
mov BYTE [DriveInfo.TracksPerHead], ch
shr cl, 6
and cl, 3
mov BYTE [DriveInfo.TracksPerHead+1], cl
pop cx

xor eax, eax
mov al, cl
and al, 63
mov [DriveInfo.SectPerTrack], eax

inc [DriveInfo.Heads]
inc [DriveInfo.TracksPerHead]
pop es
ret

What is my mistake?

THX MK

By the way, has someone a good tool to copy sectors from a binary file to a Floppy (without filesystems...)

Does nobody know a solution or can anybody try it out? For that I'll post it as ASM

Have you taken into account that bios( Int 13h ah=08h - Drive Parameters) has got maximum size limitations? The infamous 504mb barrier...

regards
Mac2004

Yes, I thought about, but this cant be the prob, I think:

When I calculate the values manually, I get the following:

Heads: 2
Tracks / Head: 335
Sectors / Track: 2

That's why it seems so, that my BIOS uses a logical CHS and translates it to a physical one. The effect of this is, that u can address more sectors if u don't use the real CHS-values.

The real problem is, that the computations in this code-fragment doesnt work:

mov BYTE [DriveInfo.TracksPerHead], ch
shr cl, 6
and cl, 3
mov BYTE [DriveInfo.TracksPerHead+1], cl

cx includes the Tracks and Sectors values from int 13h call as shown above. ch contains the 8 lower bits of the tracks value and cl <7;6> are the two upper bits of the tracks value. I want to have the track's value alone, so I copy the byte in ch as the lower one to my destination, then I shift cl 6 times right to arrange bits 8 and 9 of the tracks value, set the other bit's to all zeros and copy it to the upper byte of my destination. That's the theory, but it doesnt work as I want. There must somewhere be a little stupid mistake...

MK

Partcopy by John Fine is neat tool. I use it every day to write my binary files. Just use search engines to locate it...

regards,
Mac2004