Tim Strike wrote in a message to All:

 TS> I'm using the following assembly code to try and determine
 TS> the drive type.  The reason I'm using this code is because
 TS> it's a DOS program running in a OS/2 session (but
 TS> potentially running on just vanilla DOS or Win95, or
 TS> whatever). 

 TS> mov bl,b     ; b contains the drive
 TS> add bl,1     ; adjust drive (for proper #)
 TS> mov ax,$4408 ; set I/O function
 TS> int $21      ; call INT 21
 TS> jc {at}nodrive  ; not a valid drive (oops)
 TS> cmp ax,0     ; check
 TS> je {at}floppy   ; if zero, it's a floppy!
 TS> mov al,2     ; set as type HD
 TS> jmp {at}done
 TS> {at}floppy:
 TS> mov al,1     ; floppy found!
 TS> jmp {at}done
 TS> {at}nodrive:
 TS> mov al,0
 TS> {at}done:

--------D-214408-----------------------------
INT 21 - DOS 3.0+ - IOCTL - CHECK IF BLOCK DEVICE REMOVABLE
        AX = 4408h
        BL = drive number (00h = default, 01h = A:, etc)
Return: CF clear if successful
            AX = media type (0000h removable, 0001h fixed)
        CF set on error
            AX = error code (01h,0Fh) (see #0811 at AH=59h)
SeeAlso: AX=4400h,AX=4409h,INT 2F/AX=122Bh

 TS> Now, under an OS/2 DOS box with HPFS, this returns the drive
 TS> type as "floppy" when it's really a HD.  This works fine
 TS> under DOS 6.22/Win95 without a problem. 

 TS> Does OS/2 return something for this function call?  Does it
 TS> return different values for this function call?  Is there
 TS> another interupt I can call instead? 

 TS> And no, I'm not presently an OS/2 user . . . just trying to 
 TS> track down a bug, that's all . . . otherwise I'd check myself.

It works as expected on my OS/2 system using the 8.241 (FP17) kernel.

My C: partition is HPFS:

-a100
1656:0100 mov bl,3
1656:0102 mov ax,4408
1656:0105 int 21
1656:0107 int 3
1656:0108
-g=100

AX=0001  BX=0003  CX=0000  DX=0000  SP=FFEE  BP=0000  SI=0000  DI=0000
DS=1656  ES=1656  SS=1656  CS=1656  IP=0107   NV UP EI PL NZ NA PO NC
1656:0107 CC            INT     3


It also works as expected on my D: partition, which is FAT:

-a100
1656:0100 mov bl,4
1656:0102 mov ax,4408
1656:0105 int 21
1656:0107 int 3
1656:0108
-g=100

AX=0001  BX=0004  CX=0000  DX=0000  SP=FFEE  BP=0000  SI=0000  DI=0000
DS=1656  ES=1656  SS=1656  CS=1656  IP=0107   NV UP EI PL NZ NA PO NC
1656:0107 CC            INT     3


My A: floppy is correctly shown as a floppy:

-a100
1656:0100 mov bl,1
1656:0102 mov ax,4408
1656:0105 int 21
1656:0107 int 3
1656:0108
-g=100

AX=0000  BX=0001  CX=0000  DX=0000  SP=FFEE  BP=0000  SI=0000  DI=0000
DS=1656  ES=1656  SS=1656  CS=1656  IP=0107   NV UP EI PL NZ NA PO NC
1656:0107 CC            INT     3


Surprisingly, my CD-ROM on E: generates an error with the carry bit set:

-a100
1656:0100 mov bl,5
1656:0102 mov ax,4408
1656:0105 int 21
1656:0107 int 3
1656:0108
-g=100

AX=0001  BX=0005  CX=0000  DX=0000  SP=FFEE  BP=0000  SI=0000  DI=0000
DS=1656  ES=1656  SS=1656  CS=1656  IP=0107   NV UP EI PL NZ NA PO CY
1656:0107 CC            INT     3


Be aware that the DOS version comes back as 20.30 under OS/2:

-a100
1656:0100 mov ah,30
1656:0102 int 21
1656:0104 int 3
1656:0105
-g=000

AX=1E14  BX=0000  CX=0000  DX=0000  SP=FFEE  BP=0000  SI=0000  DI=0000
DS=1656  ES=1656  SS=1656  CS=1656  IP=0104   NV UP EI PL NZ NA PO NC
1656:0104 CC            INT     3


If you want to pursue this further, the OS2DOS echo is probably appropriate.
 
-- Mike


---