The Simple Code output format

Introduction

The file structure

The file header

The following table shows the order and sizes of the fields of the file header.

Name	Size in bytes
Magic number	4
Program flags	4
Number of Program Bytes	4
Version Information	2

Magic Number

The magic number for a simple-code file is 0x7F494152.

Program Flags

The program flags field supplies 32 bits. All are reserved for future use.

Number of program bytes

This is the number of program bytes in all data records in the file.

Version information

This is a plain 16-bit number, the current version number is 0.

The records

Note that there is no alignment since the size of the data record's program bytes not necessarily is a multiple of four.

Record tags

Currently (version 0) there are three record types. This might change in future versions. Note that tag 0 is invalid and not used by the format.

Name	Value
Invalid record	0
Data record	1
Entry record	2
End record	3

Data record

The data record contains the program bytes of a consecutive range of the program. The record contains the address of the first byte, record flags, the number of bytes and the actual bytes. The records are ordered, they are first ordered in increasing segment type order, then in increasing Record start address order.

The following table shows the order and size of the data record's fields.

Name	Size
Record tag	1
Segment type	1
Record flags	2
Record start address	4
Number of program bytes	4
Program bytes	Number of program bytes

Entry record

The entry record contains the program's entry point. The record only contains an address. There should be at most one entry record in a file. It is up to each individual loader how to handle the case where there is no entry record.

The following table shows the order and size of the entry record's fields.

Name	Size
Record tag	1
Entry address	4
Segment type	1

End record

The end record marks the end of the file. The file must contain exactly one such record and it must be the last record in the file. It also contains a checksum of the rest of the file.

The following table shows the order and size of the end record's fields.

Name	Size
Record tag	1
Checksum	4

Segment types

If there is more than one address space, this field is used to specify which address space the address in the file header or record uses. Note that segment type 0 is invalid and not used by the format.

Segment type	Value
INVALID	0
CODE	1
DATA	2
XDATA	3
IDATA	4
BIT	5

Record flags

There are currently no record flags. All these bits are set to 0 and are reserved for future use.

Record start address

This is the address of the first program byte in this record. The segment type field specifies which address space the record start address resides in.

Number of program bytes

This is the number of program bytes in the record.

Program bytes

These are the program bytes, the payload, of the data record. They form a consecutive range in memory, starting on the record's start address.

Entry address

This is the address of the program's entry point. The segment type specifies which address space the entry point resides in.

Checksum

All bytes in the file, except the four checksum bytes, are checksummed (the record tag for the checksum record is included in the checksum). To keep the flash loader simple the standard checksum is also simple; the sum of the file's individual program bytes and the checksum (as an unsigned 4-byte quantity, not the sum of the checksum's four bytes) must be zero. The checksum is ~(sum of the file's bytes) + 1. It is possible to allow more advanced check summing algorithms by using format modifiers, but there is currently no such modifier. A more advanced checksum algorithm will allow better error detection but will also require a more complex flash loader.