 Nathalie Furmento
|
f13cc390c7
examples/Makefile.am: fix program name
|
11 years ago |
 Andra Hugo
|
ab7bd6810c
nested scheduling contexts + worker collection keeping track of blocked workers
|
11 years ago |
 Samuel Thibault
|
eb4095be81
port r12836 from scheduler_modularization: Fix idleness computation on CUDA workers
|
11 years ago |
|
Samuel Thibault
|
af2b4ca2a8
port r12834 from scheduler_modularization: Also show "pop" scheduling state on CUDA
|
11 years ago |
|
Samuel Thibault
|
5e7bdfcdee
port r12832 from scheduler_modularization: Add a way to easily add thread-specific string events
|
11 years ago |
|
Samuel Thibault
|
654104db17
Fix macro names
|
11 years ago |
|
Samuel Thibault
|
1d8e85f121
port r12828 from 1.1: Add changelog for STARPU_CALIBRATE_MINIMUM
|
11 years ago |
|
Samuel Thibault
|
38ddafc107
Introduce a variable to specify the minimum number of calibration measurements
|
11 years ago |
|
Samuel Thibault
|
c727d2141d
Also pass simgrid CFLAGS to c++
|
11 years ago |
|
Samuel Thibault
|
b26c190c51
port r11792 from scheduler_modularization: properly test for task termination
|
11 years ago |
|
Samuel Thibault
|
3903429984
Fix build with simgrid and without opencl
|
11 years ago |
|
Samuel Thibault
|
3ae1f26666
Fix build with simgrid and without cuda
|
11 years ago |
|
Nathalie Furmento
|
f2b00293d8
examples/sched_ctx/sched_ctx_without_sched_policy.c: fix case when there is only 1 CPU, it is shared between both contexts
|
11 years ago |
|
Nathalie Furmento
|
2624c94c24
examples/cpp/incrementer_cpp.cpp: set async flags for CUDA and OpenCL implementations
|
11 years ago |
|
Nathalie Furmento
|
22a2b51d78
examples/binary/binary.c: set async flags for OpenCL implementation
|
11 years ago |
|
Samuel Thibault
|
d592415088
port r12811 from 1.1: Add changelog for the "scheduling" state in traces
|
11 years ago |
|
Samuel Thibault
|
8f8b885f30
Use XBT_BARRIER_SERIAL_PROCESS instead of hardcoded -1
|
11 years ago |
|
Samuel Thibault
|
e6096316de
Complete "scheduling" state with pre-exec, post-exec, push_notify, etc.
|
11 years ago |
|
Samuel Thibault
|
c5a8f90515
benefit from xbt_barrier_t when available
|
11 years ago |
|
Samuel Thibault
|
03bf7a2c4f
Benefit from MSG_get_as_by_name whenever it is available
|
11 years ago |
|
Nathalie Furmento
|
c18275f016
src/datawizard/malloc.c: improve error message
|
11 years ago |
|
Nathalie Furmento
|
8feca5ddce
examples/sched_ctx/sched_ctx.c: fix code when no cuda worker is available
|
11 years ago |
|
Nathalie Furmento
|
06778e0bb6
sched_ctx: new function starpu_sched_ctx_display_workers() to display the worker names of a context
|
11 years ago |
|
Nathalie Furmento
|
6786e419d9
examples/sched_ctx/sched_ctx_without_sched_policy.c: skip application when no worker cpu is found
|
11 years ago |
|
Nathalie Furmento
|
b1a6cad655
examples/sched_ctx/sched_ctx.c: skip application when no worker cpu is found
|
11 years ago |
|
Nathalie Furmento
|
adec4eb7ec
examples/worker_collections/worker_tree_example.c: skip application when no worker is found
|
11 years ago |
|
Nathalie Furmento
|
101eceed7a
tests/perfmodels/regression_based.c: Enable asynchronous CUDA kernels
|
11 years ago |
|
Nathalie Furmento
|
54305fca78
examples/sched_ctx/sched_ctx.c: when there is no CUDA devices, the context is created with a CPU, hence task submission does not fail
|
11 years ago |
|
Nathalie Furmento
|
56a874e16b
mpi/src/starpu_mpi_cache.c: protect acces to hashtables with mutexes
|
11 years ago |
|
Nathalie Furmento
|
f5ce9ebd53
mpi/src: move all functionalities related to early application requests to starpu_mpi_early_request
|
11 years ago |
