Notice: Applications that were compiled with older version of clib4, need to be recompiled to use the new version without issues
Highlights in this major release
- Memory allocator: overhauled. wmem now supports aligned allocations (e.g., memalign / posix_memalign).
- Pthreads: heavily optimized. Semaphores were replaced with mutexes where appropriate. Thread-local storage (TLS) operations are now much faster thanks to targeted inline assembly.
- I/O: fixes and speedups to stdio/unistd paths; performance now closely matches newlib’s.
- Startup: fixed issues when launching programs from Workbench tooltypes. Note: existing programs may need to be recompiled.
- ctype: fully rewritten; up to ~80% faster.
- Environment: getenv() / setenv() reimplemented for higher performance.
- Misc: numerous additional fixes and improvements.
What's Changed
- Version files changes by the github bot by @github-actions[bot] in #270
- Improved memalign by @afxgroup in #272
- Pthreads fixes by @afxgroup in #273
- Roundeven functions by @afxgroup in #274
- Profile fixes by @afxgroup in #278
- Small change on profile reverted by @afxgroup in #279
- Canonicalize file name by @afxgroup in #281
- Fixed a crash when debug was enabled by @afxgroup in #282
- Ctype by @afxgroup in #283
- Allocator alignment by @elfpipe in #284
- Allocator alignment (2) by @elfpipe in #285
- Remove #endif. by @elfpipe in #286
- Fixed pthread lib by @afxgroup in #287
- Update wmem/malloc/memalign code. by @elfpipe in #288
- Speedup I/O a little bit by @afxgroup in #289
- Fixed a problem on Input/Output/Error streams by @afxgroup in #290
- NECESSARY fix for the workbench startup message issue. by @elfpipe in #291
- This is another fix for the BLOCK allocator. by @elfpipe in #292
- Various speed up improvements by @afxgroup in #293
- Replaced Semaphores with Mutexes in pthread library by @afxgroup in #294
Full Changelog: v1.6.0...v2.0.0
Contributors
afxgroup and elfpipe