 Samuel Thibault
|
8e7b92c6e8
silence warning
|
hace 15 años |
|
Samuel Thibault
|
6b70d00b21
silent warning
|
hace 15 años |
|
Samuel Thibault
|
846df70675
Fix build without fxt
|
hace 15 años |
|
Samuel Thibault
|
e9e66b6c78
Mention DriverCopyAsync in the trace.
|
hace 15 años |
|
Samuel Thibault
|
3926b7c7be
Fix hack which was wrong with peers: simply record both source and destination, it actually makes the statistics code simpler
|
hace 15 años |
|
Samuel Thibault
|
209ebef2ac
Trace async transfers differently to watch for cuda spurious blocking
|
hace 15 años |
 Cyril Roelandt
|
2787e1fa46
Added an SSE codelet to the vector scaling example.
|
hace 15 años |
 Ludovic Courtès
|
886913ffe3
gcc: Support interleaved declarations & definitions of task implementations.
|
hace 15 años |
|
Ludovic Courtès
|
68b2f37506
gcc: Simplify `task' attribute handling.
|
hace 15 años |
|
Samuel Thibault
|
ed747c4790
typo
|
hace 15 años |
|
Samuel Thibault
|
77007ead6a
revert r4190, it's completely bogus, we need to add -L for the AC_HAVE_LIBRARY. AC_HAVE_LIBRARY actually does not add -lcudart to LDFLAGS because it has a non-empty action. Let's thus add CUDA_LDFLAGS before the cublas test
|
hace 15 años |
|
Samuel Thibault
|
89f3df6586
keep -lcudart when checking for libcublas. This is needed for linking when that library path is not in LD_LIBRARY_PATH
|
hace 15 años |
 Olivier Aumage
|
6c4d62d0b0
- code disabled for years
|
hace 15 años |
|
Olivier Aumage
|
4a747573b4
- add missing ifdefs
|
hace 15 años |
|
Samuel Thibault
|
3d15b9bbb1
drop debugging
|
hace 15 años |
 Nathalie Furmento
|
fd06148167
configure.ac: remove trailing whitespaces
|
hace 15 años |
|
Nathalie Furmento
|
6db2c372ca
merge branch gpumem_prefetch
|
hace 15 años |
|
Samuel Thibault
|
61eda0dcef
Fix asynchronicity of the wt mechanism by keeping a read reference on the data
|
hace 15 años |
|
Samuel Thibault
|
4a0ed0dc25
Fix asynchronous prefetch: we also need to notify data dependencies in that case. Allocate the wrapper structure dynamically to permit asynchronous termination. Permit prefetch in callbacks and codelets
|
hace 15 años |
|
Samuel Thibault
|
b9f24287b1
Fix replicate reference counting
|
hace 15 años |
|
Samuel Thibault
|
d7c53078b1
avoid crashing on no model name
|
hace 15 años |
|
Olivier Aumage
|
1c6a3ad8b0
- fix initialization
|
hace 15 años |
|
Samuel Thibault
|
6f253ca156
do not make _starpu_prefetch_data_on_node_with_mode wait for the request when async is true
|
hace 15 años |
|
Samuel Thibault
|
73123b92d6
Count the number of workers in memory nodes, to avoid scheduling transfers from memory nodes without a worker (e.g. 0 cpus). Fixes at least the wt mask when no CPU is enabled.
|
hace 15 años |
|
Samuel Thibault
|
5c4a4e9c84
Add gdb functions to print data requests
|
hace 15 años |
|
Samuel Thibault
|
73ac11ecf8
provide file and line of errors
|
hace 15 años |
|
Samuel Thibault
|
0e3edeee5c
provide file and line of errors
|
hace 15 años |
|
Samuel Thibault
|
96c23945a0
revert 4155, there is another exp_end actually...
|
hace 15 años |
|
Samuel Thibault
|
6847e9ac82
exp_end is not actually the expected end, only estimations with the new task
|
hace 15 años |
|
Olivier Aumage
|
66e35b6222
- add support for complex number cases in the LU example
|
hace 15 años |
