 Samuel Thibault
|
6b863c6ee1
make sure that _starpu_get_worker_struct doesn't overflow. Inline starpu_worker_get_count to make it not too expensive
|
9 years ago |
 Nathalie Furmento
|
94d626f9e1
fix bugs detected by coverity
|
9 years ago |
|
Samuel Thibault
|
943b5d56bf
Make starpu_worker_get_id_check public
|
9 years ago |
|
Samuel Thibault
|
ff2c06f79a
Better use an inline and macro to get best performance and file/line
|
9 years ago |
|
Nathalie Furmento
|
b7d6e2d4ce
src: introduce new function _starpu_worker_get_id_check() which fails when called from outside a worker, otherwise returns starpu_worker_get_id()
|
9 years ago |
|
Samuel Thibault
|
c14e34df69
Reduce cost of tree worker iterator by caching the set of workers using the same bindid
|
9 years ago |
|
Samuel Thibault
|
160b44ac45
inline the trivial functions accessing starpu_config
|
9 years ago |
|
Samuel Thibault
|
90bf39960e
Vastly improve simgrid simulation time by using wake queues
|
10 years ago |
 Andra Hugo
|
39ccc67d88
add hypervisor strategy monitoring perf counter (not portable, the counter reading the flops dependent on the architecture)
|
10 years ago |
|
Samuel Thibault
|
56bb0f1bf3
Make a copy of the user configuration, to avoid crashes when the config structure is freed by the user
|
10 years ago |
|
Samuel Thibault
|
bf8e9c78cb
Cache values gotten from environment variables, to fix concurrency with putenv() calls, and also reduce cost
|
10 years ago |
|
Andra Hugo
|
43f256f1c4
deal with shared workers between contexts having no starpu scheduling policy (allow the strategy to schedule tasks on not blocked shared workers)
|
10 years ago |
|
Samuel Thibault
|
5d7facd505
Drop unused field
|
10 years ago |
|
Nathalie Furmento
|
bb61f3808d
update cnrs copyright
|
10 years ago |
|
Samuel Thibault
|
ebf7530d52
Collapse thread and worker for CPU workers, since we are always synchronous there
|
10 years ago |
|
Samuel Thibault
|
ea38c92f6f
Make CUDA driver initialization able to cope with several card
|
10 years ago |
|
Samuel Thibault
|
5017e4ef01
Do not block in drivers run by the application
|
11 years ago |
|
Samuel Thibault
|
4b1a16edde
Fix University name after the fusion..
|
11 years ago |
|
Samuel Thibault
|
c3e6095f8d
Move the pause implementation into its own function, instead of _starpu_machine_is_running which is called in various places where the pause can actually lead to deadlocks
|
11 years ago |
|
Samuel Thibault
|
529e9bf200
Improve static scheduling by adding support for specifying the task execution order.
|
11 years ago |
|
Samuel Thibault
|
f99ef13c88
Add CUDA kernel submission pipelining, to overlap costs and allow concurrent
|
11 years ago |
|
Andra Hugo
|
ab7bd6810c
nested scheduling contexts + worker collection keeping track of blocked workers
|
11 years ago |
|
Andra Hugo
|
3c8000976d
fix deadlock: we have to lock the worker's sched_mutex when iterating on its list of contexts but unlock it before the worker is going to sleep
|
11 years ago |
|
Andra Hugo
|
c8def5adac
sharing workers btw ctxs with no sched_policy
|
11 years ago |
|
Andra Hugo
|
3ca3790ef1
* sched_ctxs without sched_policy
|
11 years ago |
|
Andra Hugo
|
5a219c8a8f
* patch Terry: fix book workers (book workers that have already been booked and wake up eventually workers that we don't want anymore in the group)
|
11 years ago |
|
Andra Hugo
|
aed609a645
rollback r12518: don't allow unregister in callback or task codelets
|
11 years ago |
|
Andra Hugo
|
a78e7bdc6a
allow unregisters in callbacks and codelets when we have only CPUs
|
11 years ago |
|
Andra Hugo
|
1c2f50cec5
fix exec_parallel code when called by the app thread
|
11 years ago |
|
Andra Hugo
|
6fd68c0e07
execute parallel code (openmp for eg) inside contexts ------ temporary solution
|
11 years ago |
