You can not select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
112 lines
4.5 KiB
112 lines
4.5 KiB
3 years ago
|
%! Detailed register description file for PCICFG.EXE by Ralf Brown
|
||
|
|
%!
|
||
|
|
%! Filename VVVVDDDD.PCI = Vendor VVVVh, Device DDDDh
|
||
|
|
%! Last Edit 27jun98 by Ralf Brown
|
||
|
|
%!
|
||
|
|
|
||
|
|
Everything preceding a line beginning with the six characters "!begin" is
|
||
|
|
a comment and will be ignored (with the proviso that the total file size
|
||
|
|
not exceed 64K). Everything from the !begin line to a line starting with
|
||
|
|
the four characters "!end" forms part of the device description, in a
|
||
|
|
format similar to that used for printf().
|
||
|
|
|
||
|
|
To create your own device description file for PCICFG, copy this file to
|
||
|
|
a file in the same directory containing PCICFG.EXE called VVVVDDDD.PCI,
|
||
|
|
where VVVV are the four hexadecimal digits of the PCI vendor ID reported
|
||
|
|
by PCICFG, and DDDD are the four hexadecimal digits of the PCI device
|
||
|
|
ID. Then, place the appropriate formatting directives between the
|
||
|
|
"!begin" and "!end" lines below.
|
||
|
|
|
||
|
|
|
||
|
|
format spec:
|
||
|
|
everything is literal except backslashes and conversion ops
|
||
|
|
introduced by a percent sign; the percent-sign is followed by
|
||
|
|
PCI field specifier, an optional print width, and the conversion
|
||
|
|
character
|
||
|
|
|
||
|
|
conversion characters:
|
||
|
|
<newln> ignore the newline and paste together the two lines
|
||
|
|
% literal percent sign
|
||
|
|
! rest of line is a comment and is skipped
|
||
|
|
b print number in binary
|
||
|
|
o print number in octal
|
||
|
|
d print number in decimal
|
||
|
|
x print number in hexadecimal
|
||
|
|
e print boolean value as "disable" (false) or "enable" (true)
|
||
|
|
E print boolean value as "enable" (false) or "disable" (true)
|
||
|
|
f print boolean value as dash (false) or check mark (true)
|
||
|
|
n print boolean value as Y (false) or N (true), i.e. inverted
|
||
|
|
N print boolean as Yes (false) or No (true), i.e. inverted
|
||
|
|
y print boolean value as N (false) or Y (true)
|
||
|
|
Y print boolean value as No (false) or Yes (true)
|
||
|
|
(name) print enumerated value given by enum spec "name"
|
||
|
|
{xyz} print Nth character in "xyz" (prints last if N > # chars)
|
||
|
|
/lit/ print literal string if boolean is true, dashes otherwise
|
||
|
|
|a1;a2;a3;a4| select among 2-4 alternate values (index 0-3), prints
|
||
|
|
last alternative given if N > # alternatives
|
||
|
|
|
||
|
|
backslash conversions supported:
|
||
|
|
\t tab character
|
||
|
|
\\ literal backslash
|
||
|
|
\X (any X not listed above:) literal X
|
||
|
|
|
||
|
|
PCI field specifier:
|
||
|
|
'[' hex-address ']' entire byte at offset 'hex-address'
|
||
|
|
'[' addr ':' bitnum ']' bit 'bitnum' of byte at offset 'addr'
|
||
|
|
'[' addr ':' bit1 '-' bit2 ']'
|
||
|
|
bits 'bit1' to 'bit2' of byte at 'addr'
|
||
|
|
'[' bitfield-spec '|' bitfield-spec ... ']'
|
||
|
|
concatenate multiple bitfields (MS first)
|
||
|
|
any of the above may also add the following modifiers just prior to
|
||
|
|
the closing bracket or a separating vertical bar:
|
||
|
|
'<' shift shift the field left 'shift' bits
|
||
|
|
'*' mult multiply the field's value by 'mult'
|
||
|
|
'+' offset add 'offset' to the field value
|
||
|
|
'<' shift '+' offset
|
||
|
|
shift left and then add
|
||
|
|
'*' mult '+' offset multiply and then add
|
||
|
|
bit numbers up to 31 are fully supported, and address the appropriate
|
||
|
|
byte beginning at the specified address; all fields are treated as
|
||
|
|
32-bit values internally. Bit numbers above 31 are supported
|
||
|
|
provided that the indicated field does not occupy more than four
|
||
|
|
bytes (i.e. if the low bit is not a multiple of 8, the maximum field
|
||
|
|
size is less than 32 bits).
|
||
|
|
Note: addresses are in hex, all other values are decimal (unsigned except
|
||
|
|
for 'offset', which may include a leading minus sign after the plus
|
||
|
|
sign introducing it)
|
||
|
|
|
||
|
|
When printing, an explicit width greater than the item's actual width
|
||
|
|
causes the addition of padding on the left to ensure the desired
|
||
|
|
width; the padding uses blanks for all formats except binary (b),
|
||
|
|
octal (o), and hexadecimal (x) numbers, which use zeros.
|
||
|
|
|
||
|
|
|
||
|
|
!begin
|
||
|
|
|
||
|
|
%! insert device description here
|
||
|
|
|
||
|
|
!end
|
||
|
|
|
||
|
|
; everything between the above !end line and a following !enum line is also
|
||
|
|
; a comment
|
||
|
|
|
||
|
|
!enum enum-name
|
||
|
|
Value0
|
||
|
|
Value1
|
||
|
|
Value2
|
||
|
|
...
|
||
|
|
ValueN (also used for any values > N)
|
||
|
|
!end
|
||
|
|
|
||
|
|
; enums are matched by name; first, an exact match is attempted, then the
|
||
|
|
; first named enum starting with the string in the (enum) is selected
|
||
|
|
; name matches are case-sensitive!
|
||
|
|
;
|
||
|
|
; leading and trailing whitespace is ignored, allowing you to indent the
|
||
|
|
; value strings without forcing the space to appear in the output; if you
|
||
|
|
; really do need to retain leading or trailing whitespace, add a backslash
|
||
|
|
; to the beginning or end of the line
|
||
|
|
|
||
|
|
|
||
|
|
%! end of file
|