The format is specified by the argument after the file name. The format defaults to Motorola S-Record if not specified. The format specified are:

-Absolute_Object_Module_Format

This option says to use the Intel Absolute Object Module Format (AOMF) to read the file. (See srec_aomf(5) for a description of this file format.)

-Ascii-Hex

This option says to use the Ascii-Hex format to read the file. See srec_ascii_hex(5) for a description of this file format.

-Atmel_Generic

This option says to use the Atmel Generic format to read the file. See srec_atmel_genetic(5) for a description of this file format.

-Binary

This option says the file is a raw binary file, and should be read literally. (May also be written -Raw.)

-COsmac

This option says to use the RCA Cosmac Elf format to read the file. See srec_cosmac(5) for a description of this file format.

-Dec_Binary

This option says to use the DEC Binary (XXDP) format to read the file. See srec_dec_binary(5) for a description of this file format.

-Elektor_Monitor52

This option says to use the EMON52 format to read the file. See srec_emon52(5) for a description of this file format.

-FAIrchild

This option says to use the Fairchild Fairbug format to read the file. See srec_fairchild(5) for a description of this file format.

-Fast_Load

This option says to use the LSI Logic Fast Load format to read the file. See srec_fastload(5) for a description of this file format.

-Formatted_Binary

This option says to use the Formatted Binary format to read the file. See srec_formatted_binary(5) for a description of this file format.

-Four_Packed_Code

This option says to use the FPC format to read the file. See srec_fpc(5) for a description of this file format.

-Guess

This option may be uased to ask to guess the input format. This is slower than specifying an explicit format, as it may open and close the file a number of times.

-Intel

This option says to use the Intel hex format to read the file. See srec_intel(5) for a description of this file format.

-INtel_HeX_16

This option says to use the Intel hex 16 (INHX16) format to read the file. See srec_intel16(5) for a description of this file format.

-MOS_Technologies

This option says to use the Mos Technologies format to read the file. See srec_mos_tech(5) for a description of this file format.

-Motorola [ width ]

This option says to use the Motorola S-Record format to read the file. (May also be written -S-Record.) See srec_motorola(5) for a description of this file format.

The optional width argument describes the number of bytes which form each address multiple. For normal uses the default of one (1) byte is appropriate. Some systems with 16-bit or 32-bit targets mutilate the addresses in the file; this option will correct for that. Unlike most other parameters, this one cannot be guessed.

-Needham_Hexadecimal

This option says to use the Needham Electronics ASCII file format to read the file. See srec_needham(5) for a description of this file format.

-Ohio_Scientific

This option says to use the Ohio Scientific format. See srec_os65v(5) for a description of this file format.

-SIGnetics

This option says to use the Signetics format. See srec_spasm(5) for a description of this file format.

-SPAsm

This option says to use the SPASM assembler output format (commonly used by PIC programmers). See srec_spasm(5) for a description of this file format.

-SPAsm_LittleEndian

This option says to use the SPASM assembler output format (commonly used by PIC programmers). But with the data the other way around.

-STewie

This option says to use the Stewie binary format to read the file. See srec_stewie(5) for a description of this file format.

-Tektronix

This option says to use the Tektronix hex format to read the file. See srec_tektronix(5) for a description of this file format.

-Tektronix_Extended

This option says to use the Tektronix extended hex format to read the file. See srec_tektronix_extended(5) for a description of this file format.

-Texas_Instruments_Tagged

This option says to use the Texas Instruments Tagged format to read the file. See srec_ti_tagged(5) for a description of this file format.

-VMem

This option says to use the Verilog VMEM format to read the file. See srec_vmem(5) for a description of this file format.

-WILson

This option says to use the wilson format to read the file. See srec_wilson(5) for a description of this file format.

The -ignore-checksums option may be used to disable checksum validation of input files, for those formats which have checksums at all. Note that trhe checksum values are still read in and parse (so it is still an error if they are missing) but their values are not checked. Used after an input file name, the option affects that file alone; used anywhere else on the command line, it applies to all following files.

You may specify zero or more filters to be applied. Filters are applied in the order the user specifies.

-Big_Endian_Checksum_BitNot address [ nbytes [ width ]]

This filter may be used to insert the one's complement checksum of the data into the data, most significant byte first. The data is literaly summed; if there are duplicate bytes, this will produce an incorrect result, if there are holes, it will be as if they were filled with zeros. If the data already contains bytes at the checksum location, you need to use an exclude filter, or this will generate errors. You need to apply and crop or fill filters before this filter. The value will be written with the most significant byte first. The number of bytes of resulting checksum defaults to 4. The width (the width in bytes of the values being summed) defaults to 1.

-Big_Endian_Checksum_Negative address [ nbytes [ width ]]

This filter may be used to insert the two's complement (negative) checksum of the data into the data. Otherwise similar to the above.

-Big_Endian_Checksum_Positive address [ nbytes [ width ]]

This filter may be used to insert the simple checksum of the data into the data. Otherwise similar to the above.

-Little_Endian_Checksum_BitNot address [ nbytes [ width ]]

This filter may be used to insert the one's complement (bitnot) checksum of the data into the data, least significant byte first. Otherwise similar to the above.

-Little_Endian_Checksum_Negative address [ nbytes [ width ]]

This filter may be used to insert the two's complement (negative) checksum of the data into the data. Otherwise similar to the above.

-Little_Endian_Checksum_Negative address [ nbytes [ width ]]

This filter may be used to insert the simple checksum of the data into the data. Otherwise similar to the above.

-Byte_Swap [ width ]

This filter may be used to swap pairs of odd and even bytes. By specifying a width (in bytes) it is possible to reverse the order of 4 and 8 bytes, the default is 2 bytes. (Widths in excess of 8 are assumed to be number of bits.) It is not possuble to swap non-power-of-two addresses. To change the alignment, use the offset filter before and after.

-Big_Endian_CRC16 address

This filter may be used to insert an industry standard 16-bit CRC checksum of the data into the data. Two bytes, big-endian order, are inserted at the address given. Holes in the input data are ignored. Bytes are processed in ascending address order (not in the order they appear in the input).

-Little_Endian_CRC16 address

As above, except little-endian order.

-Big_Endian_CRC32 address

This filter may be used to insert an industry standard 32-bit CRC checksum of the data into the data. Four bytes, big-endian order, are inserted at the address given. Holes in the input data are ignored. Bytes are processed in ascending address order (not in the order they appear in the input).

-Little_Endian_CRC32 address

As above, except little-endian order.

-Crop address-range

This filter may be used to isolate a section of data, and discard the rest.

-Exclude address-range

This filter may be used to exclude a section of data, and keep the rest. The is the logical complement of the -Crop filter.

-Fill value address-range

This filter may be used to fill any gaps in the data with bytes equal to value. The fill will only occur in the address range given.

-UnFill value [ min-run-length ]

This filter may be used to create gaps in the data with bytes equal to value. You can think of it as reversing the effects of the -Fill filter. The gaps will only be created if the are at least min-run-length bytes in a row (defaults to 1).

-Random_Fill address-range

This filter may be used to fill any gaps in the data with random bytes. The fill will only occur in the address range given.

-AND value

This filter may be used to bit-wise AND a value to every data byte. This is useful if you need to clear bits. Only existing data is altered, no holes are filled.

-eXclusive-OR value

This filter may be used to bit-wise XOR a value to every data byte. This is useful if you need to invert bits. Only existing data is altered, no holes are filled.

-OR value

This filter may be used to bit-wise OR a value to every data byte. This is useful if you need to set bits. Only existing data is altered, no holes are filled.

-NOT

This filter may be used to bit-wise NOT the value of every data byte. This is useful if you need to invert the data. Only existing data is altered, no holes are filled.

-Big_Endian_Length address [ nbytes ]

This filter may be used to insert the length of the data (high water minus low water) into the data. This includes the length itself. If the data already contains bytes at the length location, you need to use an exclude filter, or this will generate errors. The value will be written with the most significant byte first. The number of bytes defaults to 4.

-Little_Endian_Length address [ nbytes ]

As above, however the value will be written with the least significant byte first.

-Big_Endian_MAXimum address [ nbytes ]

This filter may be used to insert the maximum address of the data (high water + 1) into the data. This includes the maximum itself. If the data already contains bytes at the given address, you need to use an exclude filter, or this will generate errors. The value will be written with the most significant byte first. The number of bytes defaults to 4.

-Little_Endian_MAXimum address [ nbytes ]

As above, however the value will be written with the least significant byte first.

-Big_Endian_MINimum address [ nbytes ]

This filter may be used to insert the minimum address of the data (low water) into the data. This includes the minimum itself. If the data already contains bytes at the given address, you need to use an exclude filter, or this will generate errors. The value will be written with the most significant byte first. The number of bytes defaults to 4.

-Little_Endian_MINimum address [ nbytes ]

As above, however the value will be written with the least significant byte first.

-OFfset nbytes

This filter may be used to offset the addresses by the given number of bytes. No data is lost, the addresses will wrap around in 32 bits, if necessary.

-SPlit multiple [ offset [ width ] ]

This filter may be used to split the input into a subset of the data, and compress the address range so as to leave no gaps. This useful for wide data buses and memory striping. The multiple is the bytes multiple to split over, the offset is the byte offset into this range (defaults to 0), the width is the number of bytes to extract (defaults to 1) within the multiple. In order to leave no gaps, the output addresses are (width / multiple) times the input addresses.

-Un_SPlit multiple [ offset [ width ] ]

This filter may be used to reverse the effects of the split filter. The arguments are identical. Note that the address range is expanded (multiple / width) times, leaving holes between the stripes.

There are three ways to specify an address range:

minimum maximum

If you specify two number on the command line (decimal, octal and hexadecimal are understood, using the C conventions) this is an explicit address range. The minimum is inclusive, the maximum is exclusive (one more than the last address). If the maximum is given as zero then the range extends to the end of the address space.

-Within input-specification

This says to use the specified input file as a mask. The range includes all the places the specified input has data, and holes where it has holes. The input specification need not be just a file name, it may be anything any other input specification can be.

-OVER input-specification

This says to use the specified input file as a mask. The range extends from the minimum to the maximum address used by the input, and ignores any holes. The input specification need not be just a file name, it may be anything any other input specification can be.

In addition, all of these methods may be used, and used more than once, and the results will be added together.

Most of the places above where a number is expected, you may supply one of the following:

-MINimum input-specification

This inserts the minimum address of the specified input file. The input specification need not be just a file name, it may be anything any other input specification can be.

-MAXimum input-specification

This inserts the maximum address of the specified input file, plus one. The input specification need not be just a file name, it may be anything any other input specification can be.

-Length input-specification

This inserts the length of the address range in the specified input file, ignoring any holes. The input specification need not be just a file name, it may be anything any other input specification can be.

For example, the -OVER file option can be thought of a short-hand for ( -min file -max file )i, except that it is much easier to type, and also more efficient.

In addition, calculated values may optionally be rounded in one of three ways:

value -Round_Down number

The value is rounded down to the the largest integer smaller than or equal to a whole multiple of the number.

value -Round_Nearest number

The value is rounded to the the nearest whole multiple of the number.

value -Round_Up number

The value is rounded up to the the smallest integer larger than or equal to a whole multiple of the number.

The simplest case is converting files from Intel hex format to Motorola S-Record format: intel-file -intel -o srec-file Converting the other was is just as simple: srec-file -o intel-file -intel In each case, the default format is Motorola S-Record format, so it does not need to be specified.

A common activity is to crop your data to match your EPROM location. Your linker may add other junk that you are not interested in, e.g. at the RAM location. In this example, there is a 1MB EPROM at the 2MB boundary: infile -crop 0x200000 0x300000 -o outfile The lower bound is inclusive, the upper bound is exclusive.

Just possibly, you have a moronic EPROM programmer, and it barfs if the eprom doesn't start at zero. Rather than butcher the linker command file, just offset the addresses: infile -crop 0x200000 0x300000 -offset -0x200000 -o outfile This example also demonstrates how the input filters may be chained together.

The command takes its name from the UNIX cat(1) command, which is short for `catenate' or `to join'. Joining files together into a single file is simple, just name as many files on the command line as you need: infile1 infile2 -o outfile However, this assumes that the files don't overlap in any way (you will get an error if they do). If both files start from address zero, you may need to use the offset filter: infile1 infile2 -offset 0x80000 -o outfile Sometimes you want the two files to follow each other exactly, but you don't know the offset in advance: infile1 infile2 -offset -maximum infile1 -o outfile Notice that where the was a number (0x80000) before, there is now a calculation (-maximum infile1). This is possible most places a number may be used (also -minimum and -range).

It is possible to fill the blanks where our data does not lie. The simplest example of this fills the entire EPROM: infile -fill 0x00 0x200000 0x300000 -o outfile This example fills the holes, if any, with zeros. You must specify a range - with a 32-bit address space, filling everything generates huge load files.

If you only want to fill the gaps in your data, and don't want to fill the entire EPROM, try: infile -fill 0x00 -over infile -o outfile This example demonstrates the fact that wherever an address range may be specified, the -over and -within options may be used.

It is common to need to ``unfill'' an eprom image after you read it out of a chip. Usually, it will have had all the holes filled with 0xFF (areas of the EPROM you don't program show as 0xFF when you read them back).

To get rid of all the 0xFF bytes in the data, use this filter: infile -unfill 0xFF -o outfile This will get rid of all the 0xFF bytes, including the ones you actually wanted in there. There are two ways to deal with this. First, you can specify a minimum run length to the un-fill: infile -unfill 0xFF 5 -o outfile This says that runs of 1 to 4 bytes of 0xFF are OK, and that a hole should only be created for runs of 5 or more 0xFF bytes in a row. The second method is to re-fill over the intermediate gaps: outile -fill 0xFF -over outfile -o outfile2 Which method you choose depends on your needs, and the shape of the data in your EPROM. You may need to combine both techniques.

If you have a 16-bit data bus, but you are using two 8-bit EPROMs to hold your firmware, you can generate the even and odd images by using the -SPlit filter. Assuming your firmware is in the firmware.hex file, use the following:

 firmware.hex -split 2 0 -o firmware.even.hex
 firmware.hex -split 2 1 -o firmware.odd.hex

 firmware.hex \
	-offset -0x10000 -split 2 0 \
	-offset 0x10000 -o firmware.even.hex
 firmware.hex \
	-offset -0x10000 -split 2 1 \
	-offset 0x10000 -o firmware.odd.hex

A second use for the -SPlit filter is memory striping. In this example, the hardware requires that 512-byte blocks alternate between 4 EPROMs. Generating the 4 images would be done as follows:

 firmware.hex -split 0x800 0x000 0x200 -o firmware.0.hex
 firmware.hex -split 0x800 0x200 0x200 -o firmware.1.hex
 firmware.hex -split 0x800 0x400 0x200 -o firmware.2.hex
 firmware.hex -split 0x800 0x600 0x200 -o firmware.3.hex

The unsplit filter may be used to reverse the effects of the split filter. Note that the address range is expanded leaving holes between the stripes. By using all the stripes, the complete input is reassembled, without any holes. For example, to reverse our previous 16-bit data bus example, use the following command:

 -o firmware.hex \
	firmware.even.hex -unsplit 2 0 \
	firmware.odd.hex  -unsplit 2 1