diff options
| author | Erik Schnetter <schnetter@cct.lsu.edu> | 2008-02-04 21:19:00 +0000 |
|---|---|---|
| committer | Erik Schnetter <schnetter@cct.lsu.edu> | 2008-02-04 21:19:00 +0000 |
| commit | 5adcabc019a609157dd6e14021f1aab061aeb6ec (patch) | |
| tree | 80023d4d3e0958589047b6ce3a7a8c6ca5dbf738 /Carpet/LoopControl/README | |
| parent | c5198fcfaa82b7c9526f12e9b2f3b9dc2f35cbed (diff) | |
LoopControl:
Add parameter printstatus. Output statistics only if desired.
Add parameter legacy_init as new initial parallelisation choice.
Add parameter ignore_initial_overhead, which ignores the first time
the loop is encountered when optimising the parallelisation.
Correct serious inconsistency when choosing the tiling.
Output average loop time as well.
darcs-hash:20080204211904-dae7b-ef0175704efb2c8e0e20ee6ee04a468803b3a5ec.gz
Diffstat (limited to 'Carpet/LoopControl/README')
| -rw-r--r-- | Carpet/LoopControl/README | 4 |
1 files changed, 4 insertions, 0 deletions
diff --git a/Carpet/LoopControl/README b/Carpet/LoopControl/README index 231398bb0..50e2eacb5 100644 --- a/Carpet/LoopControl/README +++ b/Carpet/LoopControl/README @@ -79,3 +79,7 @@ Combine several strategies: - Note that a full sampling is too expensive. - Use a hierarchical strategy for this? - Use random sampling for this? + +5. Offer the parallelisation mechanism, or the tiling mechanism + separately, without actually looping over the array or over the + tiles. |