srec_intel16 - Intel Hexadecimal 16‐bit file format
specification
This format is also known as the INHX16 format.
This document describes the hexadecimal object file format for
16‐bit microprocessors.
This format is very similar to the srec_intel(5) format,
except that the addresses are word addresses. The count field is a word
count.
The hexadecimal representation of binary is coded in ASCII
alphanumeric characters. For example, the 8‐bit binary value
0011‐1111 is 3F in hexadecimal. To code this in ASCII, one
8‐bit byte containing the ASCII code for the character '3'
(0011‐0011 or 0x33) and one 8‐bit byte containing the) ASCII
code for the character 'F' (0100‐0110 or 0x46) are required. For each
byte value, the high‐order hexadecimal digit is always the first
digit of the pair of hexadecimal digits. This representation (ASCII
hexadecimal) requires twice as many bytes as the binary representation.
A hexadecimal object file is blocked into records, each of which
contains the record type, length, memory load address and checksum in
addition to the data. There are currently six (6) different types of records
that are defined, not all combinations of these records are meaningful,
however. The record are:
- •
- Data Record
- •
- End of File Record
- •
- Extended Segment Address Record
- •
- Start Segment Address Record
- •
- Extended Linear Address Record
- •
- Start Linear Address Record
| Record Mark |
Record Length |
Load Offset |
Record Type |
Data |
Check sum |
- Record Mark.
- Each record begins with a Record Mark field containing 0x3A, the ASCII
code for the colon (“:”) character.
- Record Length
- Each record has a Record Length field which specifies the number of
16‐bit words of information or data which follows the Record Type
field of the record. This field is one byte, represented as two
hexadecimal characters. The maximum value of the Record Length field is
hexadecimal 'FF' or 255.
- Load Offset
- Each record has a Load Offset field which specifies the 16‐bit
starting load offset of the data words, therefore this field is only used
for Data Records (if the words are loaded as bytes, the address needs to
be doubled). In other records where this field is not used, it should be
coded as four ASCII zero characters (“0000” or 0x30303030).
This field one 16‐bit word, represented as four hexadecimal
characters.
- Record Type
Each record has a Record Type field which specifies the
record type of this record. The Record Type field is used to interpret the
remaining information within the record. This field is one byte, represented
as two hexadecimal characters. The encoding for all the current record types
are:
- 0
- Data Record
- 1
- End of File Record
- 5
- Execution Start Address Record
- Data
- Each record has a variable length Data field, it consists of zero or more
16‐bit words encoded as set of 4 hexadecimal digits, most
significant digit first. The interpretation of this field depends on the
Record Type field.
- Checksum
- Each record ends with a Checksum field that contains the ASCII hexadecimal
representation of the two's complement of the 8‐bit bytes that
result from converting each pair of ASCII hexadecimal digits to one byte
of binary, from and including the Record Length field to and including the
last byte of the Data field. Therefore, the sum of all the ASCII pairs in
a record after converting to binary, from the Record Length field to and
including the Checksum field, is zero.
(8‐, 16‐ or 32‐bit formats)
| Record Mark (“:”) |
Record Length |
Load Offset |
Record Type |
Data |
Check sum |
The Data Record provides a set of hexadecimal digits that
represent the ASCII code for data bytes that make up a portion of a memory
image.
The contents of the individual fields within the record are:
- Record Mark
- This field contains 0x3A, the hexadecimal encoding of the ASCII colon
(“:”) character.
- Record Length
- The field contains two ASCII hexadecimal digits that specify the number of
16‐bit data words in the record. The maximum value is 255
decimal.
- Load Offset
- This field contains four ASCII hexadecimal digits representing the word
address at which the first word of the data is to be placed. (For an
exquivalent bytes address, double it.)
- Record Type
- This field contains 0x3030, the hexadecimal encoding of the ASCII
character “00”, which specifies the record type to be a Data
Record.
- Data
- This field contains sets of four ASCII hexadecimal digits, one set for
each 16‐bit data word, most significant digit first.
- Checksum
- This field contains the check sum on the Record Length, Load Offset,
Record Type, and Data fields.
| Record Mark (“:”) |
Record Length (4) |
Load Offset (0) |
Record Type (5) |
EIP (4 bytes) |
Check sum |
The Execution Start Address Record is used to specify the
execution start address for the object file. This is where the loader is to
jump to begin execution once the hex load is complete.
The Execution Start Address Record can appear anywhere in a
hexadecimal object file. If such a record is not present in a hexadecimal
object file, a loader is free to assign a default execution start
address.
The contents of the individual fields within the record are:
- Record mark
- This field contains 0x3A, the hexadecimal encoding of the ASCII colon
(“:”) character.
- Record length
- The field contains 0x3032, the hexadecimal encoding of the ASCII
characters “02”, which is the length, in bytes, of the EIP
register content within this record.
- Load Offset
- This field contains 0x30303030, the hexadecimal encoding of the ASCII
characters “0000”, since this field is not used for this
record.
- Record Type
- This field contains 0x3035, the hexadecimal encoding of the ASCII
character “05”, which specifies the record type to be a
Start Address Record.
- EIP
- This field contains eight ASCII hexadecimal digits that specify the
address. The field is encoded big‐endian (most significant digit
first).
- Checksum
- This field contains the check sum on the Record length, Load Offset,
Record Type, and EIP fields.
This shall be the last record in the file.
| Record Mark (“:”) |
Record Length (0) |
Load Offset (0) |
Record Type (1) |
Check sum (0xFF) |
The End of File Record specifies the end of the hexadecimal object
file.
The contents of the individual fields within the record are:
- Record mark
- This field contains 0x3A, the hexadecimal encoding of the ASCII colon
(“:”) character.
- Record
Length
- The field contains 0x3030, the hexadecimal encoding of the ASCII
characters “00”. Since this record does not contain any Data
bytes, the length is zero.
- Load Offset
- This field contains 0x30303030, the hexadecimal encoding of the ASCII
characters “0000”, since this field is not used for this
record.
- Record Type
- This field contains 0x3031, the hexadecimal encoding of the ASCII
character “01”, which specifies the record type to be an End
of File Record.
- Checksum
- This field contains the check sum an the Record Length, Load Offset, and
Record Type fields. Since all the fields are static, the check sum can
also be calculated statically, and the value is 0x4646, the hexadecimal
encoding of the ASCII characters “FF”.
In general, binary data will expand in sized by approximately 2.3
times when represented with this format.
Here is an example INHX16 file. It contains the data
“Hello, World” to be loaded at address 0.
:0700000065486C6C2C6F5720726F646CFF0AA8
:00000001FF
srec_cat version 1.64
Copyright (C) 1998, 1999, 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007,
2008, 2009, 2010, 2011, 2012, 2013, 2014 Peter Miller
The srec_cat program comes with ABSOLUTELY NO WARRANTY; for
details use the 'srec_cat -VERSion License' command. This is free
software and you are welcome to redistribute it under certain conditions;
for details use the 'srec_cat -VERSion License' command.
| Scott Finneran |
E‐Mail: |
scottfinneran@yahoo.com.au |
| Peter Miller |
E‐Mail: |
pmiller@opensource.org.au |