 Samuel Thibault
|
8ba2bf5419
fix build
|
vor 11 Jahren |
|
Samuel Thibault
|
1359bf77f3
use _WIN32 instead of just MINGW32
|
vor 11 Jahren |
|
Samuel Thibault
|
c8ef6dbd52
Windows doesn't respect posix' definition of timespec in time.h
|
vor 11 Jahren |
|
Samuel Thibault
|
6f04292eec
enable starpu_profiling_status_get optimization only with gcc and alike which support macros returning a value
|
vor 11 Jahren |
|
Samuel Thibault
|
fbcff3ec64
Use _starpu_sleep instead of usleep
|
vor 11 Jahren |
|
Samuel Thibault
|
7ba4274ea3
drop spurious include
|
vor 11 Jahren |
|
Samuel Thibault
|
c7291e5cb2
Use _starpu_callback_func_t instead of explicit function type, which bothers MSVC
|
vor 11 Jahren |
|
Samuel Thibault
|
1f1954c43b
Add OpenCL kernel submission pipelining, to overlap costs
|
vor 11 Jahren |
|
Samuel Thibault
|
b7efebe04d
Fix task accounting when the scheduler erroneously puts a non-cuda task to a cuda device
|
vor 11 Jahren |
|
Samuel Thibault
|
3abd76c814
shutdown StarPU if it was successfully started but the devices we wanted are not there
|
vor 11 Jahren |
|
Samuel Thibault
|
86eab72d26
revert r13284 and r13286: probes happen before the TSD is allocated
|
vor 11 Jahren |
|
Samuel Thibault
|
df782f4617
delete TSD only after fxt is really off
|
vor 11 Jahren |
|
Samuel Thibault
|
bec10be1cd
Use TSD to cache the value of gettid for FXT traces
|
vor 11 Jahren |
|
Samuel Thibault
|
f313079113
Factorize code to prepare for OpenCL pipelining
|
vor 11 Jahren |
|
Samuel Thibault
|
62e0ff89af
Fix missing include for windows port of locking
|
vor 11 Jahren |
|
Samuel Thibault
|
23636fb4f4
Fix simgrid build
|
vor 11 Jahren |
|
Samuel Thibault
|
77e107e4af
Warn if the user uses concurrent execution on a device which doesn't support it
|
vor 11 Jahren |
|
Samuel Thibault
|
62194c5c86
Fix filling the pipeline
|
vor 11 Jahren |
|
Samuel Thibault
|
f99ef13c88
Add CUDA kernel submission pipelining, to overlap costs and allow concurrent
|
vor 11 Jahren |
|
Samuel Thibault
|
647ccc35a7
Fix estimated allocated memory
|
vor 11 Jahren |
|
Samuel Thibault
|
41cfdc75cd
free data outside critical section
|
vor 11 Jahren |
|
Samuel Thibault
|
3b766361ce
free data outside critical section
|
vor 11 Jahren |
|
Samuel Thibault
|
206b3adf49
port r13269 from 1.1: free data outside critical section
|
vor 11 Jahren |
|
Samuel Thibault
|
aa5c3e5c50
Let interfaces declare which transfers they allow with the can_copy methode.
|
vor 11 Jahren |
|
Samuel Thibault
|
1ec6838c01
Implement perfmodel locking on windows
|
vor 11 Jahren |
|
Samuel Thibault
|
e5cf638f86
Pass codelet model when computing footprints for the bound computation
|
vor 11 Jahren |
|
Samuel Thibault
|
b294c2f942
Fix hardcoded limitation to 64 MPI nodes in traces
|
vor 11 Jahren |
|
Samuel Thibault
|
f8ff9c7d1c
Fix build without memcpy_peer
|
vor 11 Jahren |
|
Samuel Thibault
|
89ece6b14f
changelog
|
vor 11 Jahren |
|
Samuel Thibault
|
2c6e5a75b0
enable gpu-gpu transfers for matrices
|
vor 11 Jahren |
