|
| 1 | +------------------------------------------------------------------- |
| 2 | +Sun Apr 13 19:46:34 UTC 2025 - Mia Herkt < [email protected]> |
| 3 | + |
| 4 | +- Update to version 0.12.1: |
| 5 | +Features |
| 6 | + * Added --memory-limit=auto to mkdwarfs to use a more reasonable |
| 7 | + (hopefully) default for the block queue. The old default of |
| 8 | + 1 GiB was quite arbitrary and definitely not suitable for |
| 9 | + low-end systems. The new auto default will determine the limit |
| 10 | + based on the number of workers (which in turn is based on the |
| 11 | + number of CPUs), the block size, and the amount of physical |
| 12 | + memory of the system. |
| 13 | + * Replaced vector_byte_buffer with malloc_byte_buffer, which is |
| 14 | + internally based around a simple buffer that doesn't incur the |
| 15 | + cost of initializing each element like std::vector. Especially |
| 16 | + for large blocks which are known to be overwritten immediately, |
| 17 | + this can save a few CPU cycles. |
| 18 | +- Changes in version 0.12.0: |
| 19 | + * New Licensing Conditions: Instead of being all GPL-3.0 like all |
| 20 | + the previous releases, this release changes the license of a |
| 21 | + large fraction of the DwarFS code to MIT. All tools and |
| 22 | + libraries that only read DwarFS images are now MIT-licensed. |
| 23 | + Everything else (e.g. mkdwarfs) is still GPL-3.0 for the time |
| 24 | + being. |
| 25 | + |
| 26 | +Bugfixes |
| 27 | + * Changes for compatibility with Boost.Process v2. |
| 28 | + |
| 29 | +Features |
| 30 | + * Re-licensed all libraries required for reading DwarFS images |
| 31 | + under the MIT license. The source of all tools that just read |
| 32 | + DwarFS images (i.e. everything except for mkdwarfs) are also |
| 33 | + under the MIT license now. Everything else is still GPL-3.0. |
| 34 | + gh#mhx/dwarfs#255 |
| 35 | + * New hotness categorizer in mkdwarfs that allows a list of "hot" |
| 36 | + files to be stored in distinct file system blocks. |
| 37 | + * New explicit ordering mode in mkdwarfs that allows files to be |
| 38 | + ordered accoring to the order in a given list file. |
| 39 | + * dwarfs now shows the version of the FUSE library used. |
| 40 | + * New dwarfs options preload_all and preload_category to populate |
| 41 | + the block cache immediately after mounting. |
| 42 | + * New dwarfs option analysis_file that can be used for profiling |
| 43 | + and as input to mkdwarfs new hotness categorizer and explicit |
| 44 | + ordering mode. |
| 45 | + * New dwarfs option block_allocator that allows the user to |
| 46 | + switch from a malloc-based block allocator to an mmap-based |
| 47 | + one. This can help with returning memory back to the system if |
| 48 | + the blocks are evicted from the cache. |
| 49 | + |
1 | 50 | -------------------------------------------------------------------
|
2 | 51 | Fri Apr 4 07:31:03 UTC 2025 - Jan Engelhardt < [email protected]>
|
3 | 52 |
|
|
0 commit comments