Code

History

Commits on git-commitgraph-v0.1.2 to git-odb/src525+
  1. refactorSebastian Thiel committed 6 years ago4dd9fd4
  2. closer to actually producing statisticsSebastian Thiel committed 6 years ago5f087ec
  3. refactorSebastian Thiel committed 6 years ago7add82c
  4. Also average statistics on chunk levelSebastian Thiel committed 6 years ago3b927e5
  5. Provide more detailed information when decoding an entrySebastian Thiel committed 6 years ago80c5da9
  6. no need to say 'begin' before doing something, it's primarily for loggingSebastian Thiel committed 6 years ago13eba3a
  7. throughput for packSebastian Thiel committed 6 years ago81f5c33
  8. print performance stats at the end of hashing the indexSebastian Thiel committed 6 years ago9c94417
  9. assure hashing progress is dropped when doneSebastian Thiel committed 6 years agodb6e067
  10. First implementation of logging per threadSebastian Thiel committed 6 years ago477dd90
  11. Support for providing progress to threadsSebastian Thiel committed 6 years ago2815858
  12. properly count objectsSebastian Thiel committed 6 years agod398e7e
  13. first very basic progress implementationSebastian Thiel committed 6 years agob820717
  14. Pass progress everywhere, for now just to discard itSebastian Thiel committed 6 years agoda3ae1c
  15. Control which hashing crates to use from the top-level as well.Sebastian Thiel committed 6 years agodfe9b20
  16. Use git-features to toggle 'parallel' mode from the 'gitoxide' levelSebastian Thiel committed 6 years agod944fbf
  17. sketch out `Progress` trait; don't forget to write docs at some pointSebastian Thiel committed 6 years ago534b3c7
  18. refactorSebastian Thiel committed 6 years agobaeb4ef
  19. refactorSebastian Thiel committed 6 years agoe12bfd6
  20. Make `in_parallel` trait bound more loose: Clone instead of copySebastian Thiel committed 6 years ago3e91b05
  21. Using all cores actually does speed things upSebastian Thiel committed 6 years agoed944b9
  22. Also run index+pack validation in parallel; only parallelize bigger packsSebastian Thiel committed 6 years agodc15b26
  23. avoid running anything in parallel for small packsSebastian Thiel committed 6 years agoc2df183
  24. Don't send every single entry, instead send reasonably sized chunksSebastian Thiel committed 6 years ago56298a6
  25. refactor (down to 6 minutes for big pack verification)Sebastian Thiel committed 6 years ago4157b51
  26. first working version of actually parallel `in_parallel`Sebastian Thiel committed 6 years ago145ee39
  27. first implementation of 'parallel' without threads. How will scoped fare?Sebastian Thiel committed 6 years ago735744e
  28. A sketch of a minimal helper for parallel workSebastian Thiel committed 6 years ago377252a
  29. refactorSebastian Thiel committed 6 years agobe4795f
  30. refactorSebastian Thiel committed 6 years ago3e2efff
  31. bigger LRU caches are better, but with this one we can't go too largeSebastian Thiel committed 6 years ago5e1f7ae
  32. First implementation of an LRU cache - it gets hit, let's see how it fares!Sebastian Thiel committed 6 years ago5a21031
  33. also set the cache with bases and deltasSebastian Thiel committed 6 years ago915a3fb
  34. first sketch of cache implementation - get() is there, next is put()Sebastian Thiel committed 6 years agoce54756
  35. Allow delta base resolution to fail (similar to how lookups can fail)Sebastian Thiel committed 6 years agob721424