 Samuel Thibault
|
5a5eb593d1
Always reset the local worker key to worker0 for datawizard to get the proper wait queue when using multiple workers in the same driver thread
|
8 年 前 |
|
Samuel Thibault
|
a41080d2a6
Fix setting the proper CUDA stream when using several streams from the same driver thread
|
8 年 前 |
|
Samuel Thibault
|
52d97c6ae9
Fix cublas initialization/shutdown when using one thread per stream with multistream
|
8 年 前 |
|
Samuel Thibault
|
27cfafb119
Move setting worker_set->nworkers to topology, where it makes more sense
|
8 年 前 |
|
Samuel Thibault
|
6dbd5add56
Fix indent
|
8 年 前 |
|
Samuel Thibault
|
d3d32d6451
Fix priorities for schedulers which don't have unbound priority values
|
8 年 前 |
|
Samuel Thibault
|
ace9f6b23e
Use better priorities in cholesky algorithm
|
8 年 前 |
|
Samuel Thibault
|
5278ab7d1b
typo
|
8 年 前 |
|
Samuel Thibault
|
14492fb09a
typo
|
8 年 前 |
 Nathalie Furmento
|
461ca6a408
mpi/src/starpu_mpi_cache_stats.c: remove internal array which is not needed
|
8 年 前 |
|
Nathalie Furmento
|
d8961fb495
mpi: move cache data in the starpu_data_handle_t
|
8 年 前 |
|
Samuel Thibault
|
6a9541027d
comments
|
8 年 前 |
|
Samuel Thibault
|
5e61b183a5
Fix reporting CUDA workers when using several workers per CUDA
|
8 年 前 |
|
Samuel Thibault
|
8a713e9521
Fix missing include, thanks Léo
|
8 年 前 |
|
Nathalie Furmento
|
4ee7c64768
mpi/tests/insert_task_node_choice.c: use long long to make sure test is valid on a 32-bit architecture
|
8 年 前 |
|
Samuel Thibault
|
cd847ea3d3
Fix build with upcoming simgrid release
|
8 年 前 |
|
Samuel Thibault
|
e71124580f
shut up warning
|
8 年 前 |
|
Samuel Thibault
|
c16ae6ed15
Fix warning with simgrid
|
8 年 前 |
|
Samuel Thibault
|
e219b37a8d
Fix warnings with older simgrid
|
8 年 前 |
|
Samuel Thibault
|
cc7c39c4f7
Fix build with upcoming simgrid 3.15
|
8 年 前 |
|
Samuel Thibault
|
b24f685b47
When trying to reuse a memchunk, avoid locking the handle when we easily know that a transfer is needed and thus we will not afford doing it
|
8 年 前 |
|
Samuel Thibault
|
6734d1fd95
Fix initial value for number of cuda devices
|
8 年 前 |
|
Samuel Thibault
|
3bb7af49d1
Fix build without allocation cache
|
8 年 前 |
|
Samuel Thibault
|
afcbc106ef
Fix build without allocation cache
|
8 年 前 |
|
Samuel Thibault
|
03c9afe025
Do not set ncudagpus each time _starpu_init_cuda is called, it is useless
|
8 年 前 |
|
Nathalie Furmento
|
e2a6f6ca5a
tools/starpu_machine_display.c: also display SCC and MPI master slaves nodes
|
8 年 前 |
|
Samuel Thibault
|
1879558d5f
Drop now-useless case
|
8 年 前 |
|
Samuel Thibault
|
2f6fe20825
Only reuse memchunks which are easy to throw away
|
8 年 前 |
|
Samuel Thibault
|
72453f9f98
Disable forcing available memory by default, since this makes StarPU spuriously free data when prefetching fills the memory. Clean buffers are enough to be able to allocate data easily anyway.
|
8 年 前 |
|
Samuel Thibault
|
33399e8f5a
Improve data reuse: before calling generic reclaiming, try to reuse potentially-in-use mc of the same interface. Avoids having generic reclaiming free data just to be able to reallocate exactly the same kind of data
|
8 年 前 |
