townforge/outcome
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

Commit Graph

  • a7d476e05b Next attempt! Niall Douglas (s [underscore] sourceforge {at} nedprod [dot] com) 2014-07-11 19:52:05 +0100
  • f88fd37e63 Reverted to pre-shared_ptr edition. Niall Douglas (s [underscore] sourceforge {at} nedprod [dot] com) 2014-07-10 20:53:06 +0100
  • 7aaaafc9ec Removed more stuff. Performance at 48%. This really sucks. Niall Douglas (s [underscore] sourceforge {at} nedprod [dot] com) 2014-07-10 20:47:09 +0100
  • e479fd1c22 Removed the embedded copy of the value in the iterator. Now at 39% performance. Niall Douglas (s [underscore] sourceforge {at} nedprod [dot] com) 2014-07-10 20:01:57 +0100
  • 681e7fe888 Well it works ... but at 33% of a standard unordered_map. This is becoming frustrating :( Niall Douglas (s [underscore] sourceforge {at} nedprod [dot] com) 2014-07-10 19:58:16 +0100
  • c1a18c1f6a Bumped to GCC 4.9 as libstdc++ 4.8 can't cope. Niall Douglas (s [underscore] sourceforge {at} nedprod [dot] com) 2014-07-10 19:33:21 +0100
  • 9c32acfc24 Next attempt, this time at a shared_ptr based map. Niall Douglas (s [underscore] sourceforge {at} nedprod [dot] com) 2014-07-10 19:28:07 +0100
  • 2b50067d06 Works, though slowly. Niall Douglas (s [underscore] sourceforge {at} nedprod [dot] com) 2014-07-08 18:24:43 +0100
  • 6548c1dd8f Turned GCC optimisation back on Niall Douglas (s [underscore] sourceforge {at} nedprod [dot] com) 2014-07-07 22:35:01 +0100
  • a328d6c8db Got it working properly on MSVC thankfully. Performance is about 20x slower than single threaded unordered_map. Niall Douglas (s [underscore] sourceforge {at} nedprod [dot] com) 2014-07-07 22:31:12 +0100
  • cd10fc9b84 Added more debug stuff. I am mystified why this is pushing all the nodes to the final buckets. Moving to Windows, it has a much better debugger. Niall Douglas (s [underscore] sourceforge {at} nedprod [dot] com) 2014-07-07 19:35:36 +0100
  • f2ae73311a wip, at least it runs without crashing, though it's expanding some buckets hugely more than it should. Niall Douglas (s [underscore] sourceforge {at} nedprod [dot] com) 2014-07-05 20:00:31 +0100
  • 9e67dab2aa Refactor complete, in the sense it compiles. Niall Douglas (s [underscore] sourceforge {at} nedprod [dot] com) 2014-07-03 19:50:51 +0100
  • 3599cf485a Wip of a complete rearchitecture of concurrent_unordered_map into a pure atomics implementation. Niall Douglas (s [underscore] sourceforge {at} nedprod [dot] com) 2014-07-02 20:10:07 +0100
  • c7c7e842f1 WIP on concurrent_unordered_map Niall Douglas (s [underscore] sourceforge {at} nedprod [dot] com) 2014-06-28 21:58:08 +0100
  • d53bd7b8d9 Added test for transacted reads but locked writes. Niall Douglas (s [underscore] sourceforge {at} nedprod [dot] com) 2014-06-27 18:47:49 +0100
  • 0fe2141136 Added malloc benchmark test. Niall Douglas (s [underscore] sourceforge {at} nedprod [dot] com) 2014-06-26 19:10:45 +0100
  • f200224edc Renamed whole project to boost.spinlock. Niall Douglas (s [underscore] sourceforge {at} nedprod [dot] com) 2014-06-25 19:06:13 +0100
  • 7ed7fb0484 Stopped hanging oneself around a cache line. Niall Douglas (s [underscore] sourceforge {at} nedprod [dot] com) 2014-06-23 20:23:26 +0100
  • 7fc8a33335 Added a benchmark. Niall Douglas (s [underscore] sourceforge {at} nedprod [dot] com) 2014-06-23 20:15:18 +0100
  • 5d9881f064 First import from Boost.AFIO Niall Douglas (s [underscore] sourceforge {at} nedprod [dot] com) 2014-06-23 19:26:59 +0100