6.4 KiB
% patchelfcrc(1) 0.0.2 % Mario Huettel % October 2022
NAME
patchelfcrc - Patch CRC checksums into ELF files
SYNOPSYS
patchelfcrc [-lrv?V] [-g GRANULARITY] [-p POLYNOMIAL] [-s STARTVALUE] [-x XORVAL] [-F FORMAT] [-O OUTPUTSECTION] [-S SEC] [--granularity=GRANULARITY] [--little-endian] [--dry-run] [--poly=POLYNOMIAL] [--reversed] [--start-value=STARTVALUE] [--verbose] [--xor-out=XORVAL] [--end-magic=MAGIC] [--crc-format=FORMAT] [--list-crcs] [--output-section=OUTPUTSECTION] [--start-magic=MAGIC] [--section=SECTION] [--help] [--usage] [--version] ELF
DESCRIPTION
patchelfcrc reads in sections of an ELF file and computes CRC checksums over the data. The calculated CRCs are placed in an output section of the ELFFILE. The output section must already exist inside the ELF file and must be big enough to hold all generated CRC checksums.
patchelfcrc supports multiple predefined CRCs as well as custom polynomials.
OPTIONS
- -h, --help
- Display help
- --usage
- Display usage information
- -p POLYNOMIAL, --polynomial=POLYNOMIAL
- Polynomial to use for CRC calculation. For a n bit wide CRC supply a number with n+1 bits and the MSB set. Alternatively, a predifined name can be supplied. See --list-crcs for details. At maximum a 32 bit wide CRC can be calculated. If nothing is given, crc-32-mpeg is used.
- -s STARTVALUE, --start-value=STARTVALUE
- Start value to preload CRC register with. This value is XORed with the XORVAL (see option -x).
- -x XORVAL, --xor-value=XORVAL
- XOR Value applied to initial start value (-s) and to the final CRC result.
- -r, --reversed
- Use bitreversed CRC. This is not implemented yet!
- -g GRANULARITY, --granularity=GRANULARITY
- CRC calculation granularity. This has no effect if big endian layout is used. For little endian layout, it specifies the sizes of the individual elements the CRC is computed over.
-
GRANULARITY = [word | halfword | byte]. Defaults to byte.
- -l, --little-endian
- The memory layout of the ELFFILE is in little-endian format.
- -F FORMAT, --crc-format=FORMAT
- Output format to place in output section. Options for FORMAT are bare or struct
- --start-magic=MAGIC, --endmagic=MAGIC
- MAGIC numbers (32 bit unsigned) that are expected to be found at the start and the end of the given output section. This serves as safety guard against accidental corruption of the output file. It is highly recommended to use these options.
- --help, -h, -?
- Print help.
- --dry-run
- Dry run. Do all calculations but do not write changes to file. ELF file will only be opened readonly. This mode implicitly activates the verbose output
- -v, --verbose
- Activate verbose output
- -V, --version
- Print version number
- --usage
- Print usage hints on command line options.
EXAMPLES
patchelfcrc --list-crcs
| Name | Polynomial | Reversed | Start Value | Output XOR |
|---|---|---|---|---|
| crc-8 | 0x107 | no | 0x0 | 0x0 |
| crc-8-darc | 0x139 | yes | 0x0 | 0x0 |
| crc-8-i-code | 0x11d | no | 0xfd | 0x0 |
| crc-8-itu | 0x107 | no | 0x55 | 0x55 |
| crc-8-maxim | 0x131 | yes | 0x0 | 0x0 |
| crc-8-rohc | 0x107 | yes | 0xff | 0x0 |
| crc-8-wcdma | 0x19b | yes | 0x0 | 0x0 |
| crc-16 | 0x18005 | yes | 0x0 | 0x0 |
| crc-16-buypass | 0x18005 | no | 0x0 | 0x0 |
| crc-16-dds-110 | 0x18005 | no | 0x800d | 0x0 |
| crc-16-dect | 0x10589 | no | 0x1 | 0x1 |
| crc-16-dnp | 0x13d65 | yes | 0xffff | 0xffff |
| crc-16-en-13757 | 0x13d65 | no | 0xffff | 0xffff |
| crc-16-genibus | 0x11021 | no | 0x0 | 0xffff |
| crc-16-maxim | 0x18005 | yes | 0xffff | 0xffff |
| crc-16-mcrf4xx | 0x11021 | yes | 0xffff | 0x0 |
| crc-16-riello | 0x11021 | yes | 0x554d | 0x0 |
| crc-16-t10-dif | 0x18bb7 | no | 0x0 | 0x0 |
| crc-16-teledisk | 0x1a097 | no | 0x0 | 0x0 |
| crc-16-usb | 0x18005 | yes | 0x0 | 0xffff |
| x-25 | 0x11021 | yes | 0x0 | 0xffff |
| xmodem | 0x11021 | no | 0x0 | 0x0 |
| modbus | 0x18005 | yes | 0xffff | 0x0 |
| kermit | 0x11021 | yes | 0x0 | 0x0 |
| crc-ccitt-false | 0x11021 | no | 0xffff | 0x0 |
| crc-aug-ccitt | 0x11021 | no | 0x1d0f | 0x0 |
| crc-24 | 0x1864cfb | no | 0xb704ce | 0x0 |
| crc-24-flexray-a | 0x15d6dcb | no | 0xfedcba | 0x0 |
| crc-24-flexray-b | 0x15d6dcb | no | 0xabcdef | 0x0 |
| crc-32 | 0x104c11db7 | yes | 0x0 | 0xffffffff |
| crc-32-bzip2 | 0x104c11db7 | no | 0x0 | 0xffffffff |
| crc-32c | 0x11edc6f41 | yes | 0x0 | 0xffffffff |
| crc-32d | 0x1a833982b | yes | 0x0 | 0xffffffff |
| crc-32-mpeg | 0x104c11db7 | no | 0xffffffff | 0x0 |
| posix | 0x104c11db7 | no | 0xffffffff | 0xffffffff |
| crc-32q | 0x1814141ab | no | 0x0 | 0x0 |
| jamcrc | 0x104c11db7 | yes | 0xffffffff | 0x0 |
| xfer | 0x1000000af | no | 0x0 | 0x0 |
- patchelfcrc -l -g word --start-magic=0x12345678 --end-magic=0x8754321 -p crc-32-mpeg -f bare -O .outputsection -S .text executable.elf
- Calculate the CRC over .text section and place the result in the .outputsection section. The output sections start and end are checked for the given magic numbers in order to assure correct memory layout. CRC-32-MPEG is used as CRC algorothm. The memory is interpreted as little endian and the CRC calculation granularity is a 32 bit word.
BUGS
Currently, reversed CRC algorithms are not implemented.