- 22 Sep, 2017 4 commits
-
-
James Willis authored
-
James Willis authored
-
James Willis authored
-
James Willis authored
-
- 13 Sep, 2017 2 commits
-
-
James Willis authored
-
James Willis authored
Estimate an upper bound on the number of particles to read into the cache and how many particles to loop over in DOPAIR2_VEC.
-
- 12 Sep, 2017 1 commit
-
-
Matthieu Schaller authored
Force compiler to report shadowed variables. Removed all shadowed variables. Added the function scheduler_activate_send() to simplify the unskipping of MPI sends.
-
- 11 Sep, 2017 1 commit
-
-
Matthieu Schaller authored
-
- 02 Sep, 2017 2 commits
-
-
James Willis authored
-
James Willis authored
-
- 01 Sep, 2017 3 commits
-
-
Matthieu Schaller authored
-
James Willis authored
-
James Willis authored
-
- 31 Aug, 2017 4 commits
-
-
James Willis authored
-
James Willis authored
-
James Willis authored
-
James Willis authored
Changed call to part_is_active to new function. Also removed unnecessary load of v_hj in runner_doself2_vec.
-
- 24 Aug, 2017 5 commits
-
-
James Willis authored
-
James Willis authored
-
James Willis authored
-
James Willis authored
Removed alternate version of doself2_vec that was slower than current version. Renamed some variables. Removed left-packing functions for force interactions as they are not used anymore.
-
James Willis authored
Created new version of doself2_vec that computes force interaction on the fly, 2.1x faster than serial.
-
- 22 Aug, 2017 2 commits
-
-
James Willis authored
-
James Willis authored
-
- 17 Aug, 2017 2 commits
-
-
James Willis authored
-
James Willis authored
-
- 16 Aug, 2017 2 commits
-
-
Matthieu Schaller authored
-
James Willis authored
-
- 15 Aug, 2017 1 commit
-
-
James Willis authored
-
- 08 Aug, 2017 3 commits
-
-
James Willis authored
-
James Willis authored
-
James Willis authored
-
- 06 Aug, 2017 4 commits
-
-
James Willis authored
-
James Willis authored
-
James Willis authored
-
James Willis authored
Corrected bug in last_pj calculation. Using the last active particle in the cell as the upper bound now works.
-
- 04 Aug, 2017 4 commits
-
-
James Willis authored
-
James Willis authored
Fixed seg fault bug. Make sure that max_index_i[ci->count - 1] and max_index_j[0] are initialised to correct values when nothing is in range.
-
James Willis authored
-
James Willis authored
-