1. 17 Oct, 2019 5 commits
  2. 06 Oct, 2019 2 commits
  3. 28 Jul, 2019 1 commit
  4. 22 May, 2019 1 commit
  5. 13 Mar, 2019 3 commits
  6. 10 Mar, 2019 2 commits
  7. 07 Nov, 2018 2 commits
  8. 02 Nov, 2018 5 commits
  9. 18 Oct, 2018 1 commit
  10. 17 Oct, 2018 1 commit
  11. 10 Oct, 2018 2 commits
  12. 09 Oct, 2018 3 commits
  13. 08 Oct, 2018 2 commits
  14. 02 Oct, 2018 4 commits
  15. 25 Sep, 2018 1 commit
  16. 19 Sep, 2018 1 commit
  17. 12 Sep, 2018 4 commits
    • Tomas Härdin's avatar
      Merge branch 'optimize' · 6e045d8d
      Tomas Härdin authored
      This has brought the speed of perftest2.sh from 2.3 kHz to 8.8 kHz, nearly
      quadruple the speed!
      
      There's some stuff still left to do in the master:
      
      * queueX() should only need to be called once for each "stage"
        in each master. It accounts for 23% of cycles.
      * initRefValues() accounts for 16% of cycles, perhaps there's still
        something that can be done to it.
      * There's probably some overhead in message packing/unpacking to take care of.
      
      Overall, the master spends about as much time inside itself as inside ZMQ.
      Memory allocations seem to account for most of that not accounted for in
      queueX() or initRefValues().
      
      As for the server, there's not much else to do since 36% of time is spent in
      FMI and 55% is spent inside ZMQ. The remaining 9% can be considered
      acceptable overhead.
      6e045d8d
    • Tomas Härdin's avatar
      Final test script · 074d1b95
      Tomas Härdin authored
      074d1b95
    • Tomas Härdin's avatar
      Truncate real_vrs instead of copying · d7d2b217
      Tomas Härdin authored
      1% callgrind improvement
      d7d2b217
    • Tomas Härdin's avatar
      Mico-optimize queue* · 74b66ce0
      Tomas Härdin authored
      No change in walltime, but callgrind says 3% fewer cycles.
      74b66ce0