Task scheduler: Start with suspended pool to avoid threading overhead on push

The idea is to avoid any threading overhead when we start pushing tasks in a
loop. Similarly to how we do it from the new dependency graph. Gives couple of
percent of speedup here, but also improves scalability.
This commit is contained in:
Sergey Sharybin 2018-01-22 12:57:08 +01:00
parent af83535dd0
commit 8553449cf2
1 changed files with 2 additions and 2 deletions

View File

@ -1101,7 +1101,7 @@ static void task_parallel_range_ex(
}
task_scheduler = BLI_task_scheduler_get();
task_pool = BLI_task_pool_create(task_scheduler, &state);
task_pool = BLI_task_pool_create_suspended(task_scheduler, &state);
num_threads = BLI_task_scheduler_num_threads(task_scheduler);
/* The idea here is to prevent creating task for each of the loop iterations
@ -1328,7 +1328,7 @@ void BLI_task_parallel_listbase(
}
task_scheduler = BLI_task_scheduler_get();
task_pool = BLI_task_pool_create(task_scheduler, &state);
task_pool = BLI_task_pool_create_suspended(task_scheduler, &state);
num_threads = BLI_task_scheduler_num_threads(task_scheduler);
/* The idea here is to prevent creating task for each of the loop iterations