 Nathalie Furmento
|
9054490002
happy 2021
|
%!s(int64=5) %!d(string=hai) anos |
|
Nathalie Furmento
|
6a16f8370f
update copyright holders
|
%!s(int64=6) %!d(string=hai) anos |
 Samuel Thibault
|
7da3610c7b
Factorize calls to MSG_process_sleep
|
%!s(int64=6) %!d(string=hai) anos |
|
Nathalie Furmento
|
f7abbfa6fc
update copyright
|
%!s(int64=7) %!d(string=hai) anos |
|
Samuel Thibault
|
b8d44ad342
small optimization
|
%!s(int64=8) %!d(string=hai) anos |
|
Nathalie Furmento
|
089c051bf5
fix copyrights
|
%!s(int64=8) %!d(string=hai) anos |
|
Nathalie Furmento
|
6901c4441b
src/: fix copyrights
|
%!s(int64=8) %!d(string=hai) anos |
 Corentin Salingue
|
f45e0a0e7c
merge from trunk
|
%!s(int64=9) %!d(string=hai) anos |
|
Corentin Salingue
|
6bcb21c3f1
Progress also data in NUMA nodes
|
%!s(int64=9) %!d(string=hai) anos |
|
Samuel Thibault
|
d6d6378102
Don't stay stuck for too long pushing requests to GPUs
|
%!s(int64=9) %!d(string=hai) anos |
|
Samuel Thibault
|
922ff28a1e
Make datawizard_progress use worker 0 of worker sets as reference for determining which memory nodes to make progress
|
%!s(int64=9) %!d(string=hai) anos |
|
Samuel Thibault
|
cb9e2e2495
Avoid static analysis warnings
|
%!s(int64=9) %!d(string=hai) anos |
|
Samuel Thibault
|
9a4011b6a0
Fix _starpu_datawizard_progress calls from the application context
|
%!s(int64=9) %!d(string=hai) anos |
|
Corentin Salingue
|
cc4eb6b654
fix indent
|
%!s(int64=9) %!d(string=hai) anos |
|
Corentin Salingue
|
e1f46dab74
really fix r19261 : add prefix _starpu_ to worker_drives_memory
|
%!s(int64=9) %!d(string=hai) anos |
|
Corentin Salingue
|
1361c43dbd
Inline new functions for managing memoy nodes
|
%!s(int64=9) %!d(string=hai) anos |
|
Corentin Salingue
|
3bbc1443c4
Each worker can manage the good set of memories
|
%!s(int64=9) %!d(string=hai) anos |
|
Samuel Thibault
|
0b5135e60e
re-apply r18403 again, we'll be using maxmin/precision simgrid parameter to coarsen the time
|
%!s(int64=9) %!d(string=hai) anos |
|
Samuel Thibault
|
514f24d352
revert r18403 for now, it seems to increase starpu-mpi-simgrid execution time a lot
|
%!s(int64=9) %!d(string=hai) anos |
|
Samuel Thibault
|
4fb4ce92aa
Reduce the granularity of simulation of CUDA and MPI probes, it is not that expensive.
|
%!s(int64=9) %!d(string=hai) anos |
|
Samuel Thibault
|
6bd3bb1b96
revert spurious change
|
%!s(int64=9) %!d(string=hai) anos |
|
Samuel Thibault
|
8bb9a66912
Fix data migration: we need to flush the cache too. Add starpu_mpi_data_migrate which does things properly
|
%!s(int64=9) %!d(string=hai) anos |
|
Samuel Thibault
|
1683d35a0d
Fix msg.h inclusion according to new location
|
%!s(int64=11) %!d(string=hai) anos |
|
Nathalie Furmento
|
bb61f3808d
update cnrs copyright
|
%!s(int64=11) %!d(string=hai) anos |
|
Samuel Thibault
|
4c6de74837
Add starpu_data_idle_prefetch_on_node
|
%!s(int64=11) %!d(string=hai) anos |
|
Samuel Thibault
|
f433b48230
Add STARPU_SIMGRID_CUDA_MALLOC_COST and STARPU_SIMGRID_QUEUE_MALLOC_COST environment variables to control whether simgrid simulates CUDA costs or not
|
%!s(int64=11) %!d(string=hai) anos |
|
Samuel Thibault
|
5a20717d13
periodically perform reclaiming if needed
|
%!s(int64=11) %!d(string=hai) anos |
|
Samuel Thibault
|
3937eb658e
Make _starpu_memchunk_tidy available to the application
|
%!s(int64=11) %!d(string=hai) anos |
|
Samuel Thibault
|
d737e2d729
Periodically reclaim memory instead of waiting the last moment.
|
%!s(int64=11) %!d(string=hai) anos |
|
Samuel Thibault
|
2a197a52b7
This was fixed
|
%!s(int64=11) %!d(string=hai) anos |
