mirror of
https://github.com/kspalaiologos/bzip3.git
synced 2026-10-23 14:48:53 +03:00
2b2c18c290
* Added: bz3_memory_needed API. * Added: bz3_decode_block_bound * Renamed: bz3_decode_bound -> bz3_decode_block_bound in documentation * Updated: libbz3.h to specify that it is a frame that is being compressed here. * Reworked how the fuzzer script works. Can now generate its own input for easier debugging. * Added: Multithreading to fuzz tutorial. * Added: 010editor template for fuzzer input. * Improved: Ignore the AFL Input/Outputs in .gitignore * Added a note saying `afl-whatsup` is not realtime. * Fixed: Variable name in embedded hex editor script. * Added: Note on where to find crashing test case :3 * Added: Integration for bz3_decode_block_bound in bz3_decode_block * Added: Guide for fuzzing with address sanitizer. * Fixed: Wrong branch in binary template of fuzz.c * Improved: Description in bz3_decode_block_bound * Revert: data_size should not compare with bz3_decode_block_bound * Removed: `bz3_decode_block_bound` API. * Added: BZ3_ERR_DATA_SIZE_TOO_SMALL when bz3_decode_block is called with insufficient buffer. * Implemented: Additional bounds checks in `bz3_decode_block` and `bz3_orig_size_sufficient_for_decode` API * Adjusted: Old fuzz script now named fuzz-decompress. Enforce includes from local files, not system. * Fixed: Bad quote usage in bz3_decode_block * Added: Standard fuzzer inputs. * Fixed: fuzz failures in decode_block and added a new fuzzer script for block decodes * Added: Sanity check that the user provided a valid compressed data size to bz3_decode_block * Changed: bz3_memory_needed -> bz3_min_memory_needed * Added: Round Trip Fuzzer * Fixed: Possible out of bounds read in `lzp_decode_block`. Mark exit branches as unlikely. * Update comment --------- Co-authored-by: Kamila Szewczyk <27734421+kspalaiologos@users.noreply.github.com>