New in LRZIP 0.615 (Sep 5, 2013)
- Adjusting -O flag semantics, options help update.
- Making -O flag operative for lrztar, whitespace path fix, lrzuntar fix, other minor.
- Further tighten up ram restrictions with stdin/stdout to prevent running out of memory with all the buffers involved.
- Massive files fail with -U due to trying to allocate the whole lot in ram while doing checksums. Do it piecemeal to avoid the problem. Patch and debugging courtesy of Adam Tkáč.
- We have to run through the clear buffer function even for empty buffers or corrupt archives with empty match streams.
- MD5 code uses little endian so remove arbitrary SWAP macro and explicitly use htole32.
- Rewrite the magic if we receive lzma properties and have not yet written them yet during stdout operation.
- Set the control lzma properties only once.
- Add a control lock mutex for protecting certain control variables.
- Fix stdin fake mremap creating null bytes on osx, patch courtesy of John Boyle.
- Cache frequently used indirectly referenced variables in the sliding mmap code.
- Micro-optimise sliding_get_sb_range.
- A fix for a bug where large files containing the same non-zero bytes which requires a sliding window, courtesy of Serge Belyshev.
- Put vchar should take a 64 bit integer and is used from more than one call site so uninline it.
- Microoptimise in hash_search.
- Inline rzip functions used from only one caller.
- Check for successful calloc of hash table only after performing it.
- stdin_eof is just a bool
- hash_bits can only be up to 64 so use a char type.
- Check endianness of build with autotools to enable md5 support on more platforms.