=> R0 = 23 (Canonicalise special field and disc name)
R1 = pointer to special field, or 0 if not present
R2 = pointer to disc name, or 0 if not present
R3 = pointer buffer for canonical special field, or 0 to find length
R4 = pointer buffer for canonical disc name, or 0 to find length
R5 = length of R3 buffer
R6 = length of R4 buffer
<= R1 = 0, or pointer to canonical special field
R2 = 0, or pointer to canonical disc name
R3 = bytes overflow from special field buffer
R4 = bytes overflow from disc name buffer
R5,R6 preserved
This call is used to convert special field and disc name to canonical forms. It is called twice:
=> R1-R4 = 0
R5,R6 don't matter
<= R1,R2 any non-zero value
R3 = length of buffer for special field (excluding null terminator)
R4 = length of buffer for disc name (excluding null terminator)
FileSwitch then allocates buffers, and calls again with
=> R1 = pointer to special field
R2 = pointer to disc name
R3 = pointer buffer
R4 = pointer buffer
R5 = length of R3 buffer
R6 = length of R4 buffer
<= R1 = entry R3
R2 = entry R4
R3,R4 = 0 (neither buffer overflowed)
This entry point is only called if bit 23 of the filing system information word is set - in this case you cannot just return an error, you must return sensible values. If your filing system doesn't use disc names or special fields, you could use this example code.
Example canonicalisation code
This code is suitable if your filing system ignores special fields and disc names.
; special field
TEQ r1, #0
MOVEQ r1, #0
MOVEQ r3, #0
BEQ %FT200
TEQ r3, #0
MOVEQ r1, #1
MOVEQ r3, #0
MOVNE r1, r3
MOVNE r14, #0
STRNEB r14, [r3]
MOVNE r3, #0
200
; disc name
TEQ r2, #0
MOVEQ r2, #0
MOVEQ r4, #0
BEQ %FT300
TEQ r4, #0
MOVEQ r2, #1
MOVEQ r4, #0
MOVNE r1, r4
MOVNE r14, #0
STRNEB r14, [r4]
MOVNE r4, #0
300
If you do want to use the fields, you will have to do more work to ensure the buffers don't overflow etc.
[sh-index] Back to list of manuals
|
| |
