setpib - Qualcomm Atheros PIB File Editor
setpib [options] file base
[type data] [type data] [...]
Replace one or more values in a Qualcomm Atheros powerline
parameter (or PIB) file and re-compute the checksum. Values are
specified as a sequence of type/data pairs. Users must know the offset,
length and type of the values they change or problems will certainly result.
This program makes absolutely no effort to validate offsets or values.
This program can be used to change any PIB value, including those
that should not be changed. Be warned that some PIB values are derived from,
or interact with, other PIB values. Failure to observe such dependencies may
diminish performance or render a device inoperable.
Atheros does not recommend the use of this program to edit PIB
parameters because small mistakes can cause big problems. In addition,
Atheros is under no obligation to provide customers with the offset, length
or type of any PIB file parameters.
Atheros recommends using program modpib to change device
identity parameters and the Windows Device Manager to change other
parameters whenever possible.
This program handles older Thunderbolt/Lightning parameter files
and newer Panther/Lynx parameter files but it has a larger footprint that
may not be acceptable for embedded systems. In cases where only the older or
the newer parameter files are used, consider using program setpib1
for an older Thunderbolt/Lightning only version or setpib2 for a
newer Panther/Lynx only version.
This program is part of the Qualcomm Atheros Powerline Toolkit.
See the AMP man page for an overview and installation
instructions.
- -q
- Enable quiet mode which has not effect.
- -v
- Print a hexdump of the edited region. Print hexdump headers if this option
appears twice. The region window may be changed with option -w.
- -w window
- The display window in bytes. When option -v is present, PIB
information surrounding the edited region is displayed to provide some
sort of context. The edited region appears in the middle of this window.
The default window is 32 bytes. This means that the 32 bytes
before and the 32 bytes after the edited region are included in the
display.
- -x
- Repair the checksum. By default, this program does not modify files that
fail the size or checksum test on input. In addition, it does not
re-compute the checksum on output if no data has changed. This option
skips the size and checksum test on input and re-computes the checksum on
output even if no data has changed. This option only works on legacy PIB
files. It does not work on PIB files having NVM file format.
- -?, --help
- Print program help summary on stdout. This option takes precedence over
other options on the command line.
- -!, --version
- Print program version information on stdout. This option takes precedence
over other options on the command line. Usethis option when sending screen
dumps to Atheros Technical Support so that they know exactly which version
of the Linux Toolkit you are using.
- file
- The PIB filename. The file must exist and will be re-written. No
assumptions are made based on filename and no filename conventions are
enforced. This argument is required and must precede the base
argument.
- base
- The byte offset where changes will start. The offset is expressed in
hexadecimal with (or without) the leading "0x". Changes
are applied in order, left to right, and offset automatically increments
by the correct amount after each change is applied. This permits
consectuive PIB locations to be changed with one command. This argument is
required and must follow the file argument. It must also precede
all format/data pairs. Normally, you must use multiple setpib
commands to change non-consecutive PIB locations but careful use of format
skip can achive the same results using one command.
- type
- The implied length and encoding of the data. Valid types are byte,
word, long, data and hfid. A valid format must
precede each data item. A byte is 8 bits, a word is 16 bits, a long is 32
bits and a huge is 64 bits. The format implies the maximum data value
except for format data which is limited by the number of octets
entered on the command line.
- data
- The information written to the current PIB offset. Integer data may be
expressed in binary, decimal or hexadecimal but must not exceed the
type minimum or maximum. String data may contain any ASCII
characters that can be entered via keyboard but must be enclosed in quotes
when spaces are included. Binary values start with 0b or 0B
and hexadecimal values start with 0x or 0X.
- byte
integer
- A signed or unsigned integer stored as 8 bits or 1 byte. The value is
expressed in decimal by default. A binary value may be expressed with a
0b prefix. A hexadecimal value may be expressed using a 0x
prefix. Negative type/data values must appear at the end of the command
line, after option "--", to avoid parsing errors. The
offset increments by 1 prior to the next insertion.
- word
integer
- A signed or unsigned integer stored as 16 bits or 2 bytes. The value is
expressed in decimal by default. A binary value may be expressed with a
0b prefix. A hexadecimal value may be expressed using a 0x
prefix. Negative type/data values must appear at the end of the command
line, after option "--", to avoid parsing errors. The
offset increments by 2 prior to the next insertion.
- long
integer
- A signed or unsigned integer stored as 32 bits or 4 bytes. The value is
expressed in decimal by default. A binary value may be expressed with a
0b prefix. A hexadecimal value may be expressed using a 0x
prefix. Negative type/data values must appear at the end of the command
line, after option "--", to avoid parsing errors. The
offset increments by 4 prior to the next insertions.
- huge
integer
- A signed or unsigned integer stored as 64 bits or 8 bytes. The value is
expressed in decimal by default. A binary value may be expressed with a
0b prefix. A hexadecimal value may be expressed using a 0x
prefix. The offset increments by 8 prior to the next insertion. Negative
type/data values must appear at the end of the command line, after option
"--", to avoid parsing errors. This format is only
available on hosts that support 64 bit data values.
- data hex
- A sequence of hexadecimal octets. Octets may be separated by a colon for
clarity. The number of octets determines the number of bytes written. The
offset increments by the number of bytes written. This is similar to
"key" and "mac" below but accepts
variable length input.
- key hex
- A sequence of hexadecimal octets. Octets may be separated by a colon for
clarity. The number of octets must convert to 16 bytes or an error will
occur. The offset increments by 16 prior to the next insertion. This is
similar "data" above but accepts only fixed length input.
This option may be used to set the DAK, NMK or NVAK values.
- mac hex
- A sequence of hexadecimal octets. Octets may be separated by a colon for
clarity. The number of octets must convert to 6 bytes or an error will
occur. The offset increments by 6 prior to the next insertion. This is
similar "data" above but accepts only fixed length input.
This option may be used to set the DAK, NMK or NVAK values.
- text
string
- An ASCII character string. The string string length determines the number
of bytes stored. The string is stored with NUL terminator included. It is
not padded or truncated. The offset increments by the number of bytes
stored prior to the next insertion. This option may be used to enter a
variable length string.
- hfid
string
- An ASCII character string. The string is always stored as 64 bytes. Short
strings are padded on the right with NUL characters. Long strings are
truncated on the right and the last byte is forced to NUL. The offset
increments by 64 bytes prior to the next insertion. This option may be
used to enter user, network and manufacturer identification strings.
- zero
count
- An unsigned integer representing the number of consecutive bytes to fill
with 0x00. The offset increments by the number of bytes written.
This option may be used to erase regions of the PIB.
- fill
count
- An unsigned integer representing the number of consecutive bytes to fill
with 0xFF. The offset increments by the number of bytes written.
This option may be used to erase regions of the PIB.
- skip
count
- An unsigned integer indicating the number of bytes to skip over before
staring another change. Intervening data data locations are unchanged.
- accesspassword
string
- An ASCII character string. The string is always stored as 257 bytes. Short
strings are padded on the right with NUL characters. Long strings are
truncated on the right and the last byte is forced to NUL. The offset
increments by 257 prior to the next insertion.
- accessusername
string
- An ASCII character string. The string is always stored as 33 bytes. Short
strings are padded on the right with NUL characters. Long strings are
truncated on the right and the last byte is forced to NUL. The offset
increments by 33 prior to the next insertion.
- adminpassword
string
- An ASCII character string. The string is always stored as 33 bytes. Short
strings are padded on the right with NUL characters. Long strings are
truncated on the right and the last byte is forced to NUL. The offset
increments by 33 prior to the next insertion.
- adminusername
string
- An ASCII character string. The string is always stored as 33 bytes. Short
strings are padded on the right with NUL characters. Long strings are
truncated on the right and the last byte is forced to NUL. The offset
increments by 33 prior to the next insertion.
- password
string
- An ASCII character string. The string is always stored as 257 bytes. Short
strings are padded on the right with NUL characters. Long strings are
truncated on the right and the last byte is forced to NUL. The offset
increments by 257 prior to the next insertion.
- url
string
- An ASCII character string. The string is always stored as 257 bytes. Short
strings are padded on the right with NUL characters. Long strings are
truncated on the right and the last byte is forced to NUL. The offset
increments by 257 prior to the next insertion.
- username
string
- An ASCII character string. The string is always stored as 257 bytes. Short
strings are padded on the right with NUL characters. Long strings are
truncated on the right and the last byte is forced to NUL. The offset
increments by 257 prior to the next insertion.
The following example edits file abc.pib by writing decimal
value 1 at offset 01F7 followed by hexadecimal value
00B052BABE01. A partial dump is printed showing 16 bytes before and
16 bytes after the changed data because the default display windows is 16
bytes. Unless the change occurs right at the start or end of the file, it
will appear in the center of the display window.
# setpib -v abc.pib 01F7 byte 1 data 00:B0:52:BA:BE:01
000001D0 00 00 00 00 00 00 00 00 00 .........
000001E0 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
000001F0 00 00 00 00 00 00 00 01 00 B0 52 BA BE 01 00 00 ..........R.....
00000200 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
00000210 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ..............
The following example sets the manufacturer HFID string in PIB
file def.pib. HFID strings are 64 characters long and either
truncated or padded on the right with NUL characters. You may need to
enclose the string in quotes it it contains spaces.
# setpib -v def.pib 24 hfid "Galactic Software Pirates, Inc."
00000000 38 1F 00 00 30 1B EB 04 00 B0 52 00 8...0.....R.
00000010 00 66 50 D3 E4 93 3F 85 5B 70 40 78 4D F8 15 AA .fP...?.[p@xM...
00000020 8D B7 00 00 47 61 6C 61 63 74 69 63 20 53 6F 66 ....Galactic Sof
00000030 74 77 61 72 65 20 50 69 72 61 74 65 73 2C 20 49 tware Pirates, I
00000040 6E 63 2E 00 00 00 00 00 00 00 00 00 00 00 00 00 nc..............
00000050 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
00000060 00 00 00 00 50 D3 E4 93 3F 85 5B 70 40 78 4D F8 ....P...?.[p@xM.
00000070 15 AA 8D B7 49 6E 74 65 6C 6C 6F 6E 20 45 6E 61 ....Atheros Ena
00000080 62 6C 65 64 bled
The following example displays the contents of PIB file
abc.pib without changing any data. The region displayed is 24 bytes
starting at offset 0x24. The format skip specifies the region but
does not change it. By default, the 32 bytes before and 32 bytes after are
included in the display. This technique can be used to inspect a specific
portion of a PIB file.
# setpib -v abc.pib 24 skip 64
00000000 38 1F 00 00 30 1B EB 04 00 B0 52 00 8...0.....R.
00000010 00 66 50 D3 E4 93 3F 85 5B 70 40 78 4D F8 15 AA .fP...?.[p@xM...
00000020 8D B7 00 00 47 61 6C 61 63 74 69 63 20 53 6F 66 ....Galactic Sof
00000030 74 77 61 72 65 20 50 69 72 61 74 65 73 2C 20 49 tware Pirates, I
00000040 6E 63 2E 00 00 00 00 00 00 00 00 00 00 00 00 00 nc..............
00000050 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
00000060 00 00 00 00 50 D3 E4 93 3F 85 5B 70 40 78 4D F8 ....P...?.[p@xM.
00000070 15 AA 8D B7 49 6E 74 65 6C 6C 6F 6E 20 45 6E 61 ....Atheros Ena
00000080 62 6C 65 64 bled
The next example does the same thing for a negative value. This is
tricky because the minus sign will be interpreted as the start of another
option. We must use "--" to end normal option parsing. This
is a POSIX standard feature. See IEEE Std 1002.1-2001, Section 12.2,
Utility Syntax Guidelines for more an explanation.
# setpib -v abc.pib 1471 -- long -50
00001450 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ...............
00001460 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
00001470 00 CE FF FF FF 00 00 00 00 00 00 00 00 00 00 00 ................
00001480 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
00001490 00 00 00 00 00 .....
The following example computes a new checksum without changing any
data. The filename and an offset are required but no other values are
needed. This technique can be used to compute a new checksum after modifying
the file using software that does not update the checksum.
# setpib abc.pib 0 -x
PIB file structure and content is proprietary to Qualcomm Atheros,
Ocala FL USA. Consequently, public information is not available. Qualcomm
Atheros reserves the right to modify PIB file structure or content in future
firmware releases without any obligation to notify or compensate users of
this program.