Samuel Thibault
|
092f322b1c
Add CUDA concurrent kernel execution support through the STARPU_NWORKER_PER_CUDA environment variable.
|
12 年之前 |
Samuel Thibault
|
2963604386
port r12554 from 1.1: Disable timing on events for transfers, we don't use it
|
12 年之前 |
Samuel Thibault
|
28308c10be
Fix array size
|
12 年之前 |
Samuel Thibault
|
f400e9b510
Update variables to unset
|
12 年之前 |
Samuel Thibault
|
f249eeb264
Optimize getting the current task
|
12 年之前 |
Samuel Thibault
|
d996ed4da4
drop unused variable
|
12 年之前 |
Samuel Thibault
|
38fbca49e9
Fix _starpu_datawizard_progress prototype: it does not actually return anything
|
12 年之前 |
Samuel Thibault
|
8bfbe1ee39
Drop mentioning Gordon
|
12 年之前 |
Samuel Thibault
|
8bcae573b8
Drop duplicate wait
|
12 年之前 |
Samuel Thibault
|
e8a62c8fce
move spurious variable initialization
|
12 年之前 |
Samuel Thibault
|
c60d91cb3f
Enable asynchronous OpenCL wherever possible
|
12 年之前 |
Samuel Thibault
|
840af37490
push pending requests at the back of the list of pending requests, not the front
|
12 年之前 |
Samuel Thibault
|
058e113825
port r12533 from 1.1: Do not let non-CUDA workers use non-0 streams, CUDA seems not very threadsafe with that. Make the coherency engine avoid selecting non-CUDA workers to issue transfers, to avoid letting them use the 0 stream.
|
12 年之前 |
Samuel Thibault
|
9e7c9936ce
Use STARPU_MAIN_RAM where appropriate
|
12 年之前 |
Samuel Thibault
|
9aa1c5ba6c
Use STARPU_MAIN_RAM where appropriate
|
12 年之前 |
Samuel Thibault
|
2bafbc890a
fix build
|
12 年之前 |
Samuel Thibault
|
92d443d920
Make documentation clearer
|
12 年之前 |
Samuel Thibault
|
ea5d8f596b
Make a copy of the interface to the memchunk only when the latter gets detached from the data, and thus the interface code will not work on it. Drop the copy when the memchunk gets reattached. This allows interfaces to modify pointers in the interface in unpack, notably
|
12 年之前 |
Samuel Thibault
|
fb03301fdf
port r12523 from 1.1: Fix detaching a memchunk from a handle: the memchunk is not supposed to still have a pointer to the replicate
|
12 年之前 |
Samuel Thibault
|
1aad553cc2
Fix crash with perfmodels having type common
|
12 年之前 |
Andra Hugo
|
aed609a645
rollback r12518: don't allow unregister in callback or task codelets
|
12 年之前 |
Andra Hugo
|
84e77b5ad7
add prolog callbacks to mpi
|
12 年之前 |
Andra Hugo
|
a78e7bdc6a
allow unregisters in callbacks and codelets when we have only CPUs
|
12 年之前 |
Andra Hugo
|
2938dd4805
patch Terry: add a new prologue callback (one executed at pop time)
|
12 年之前 |
Andra Hugo
|
e991168dec
rollback book_workers
|
12 年之前 |
Samuel Thibault
|
5c4a168e72
Do not destroy containers before using them
|
12 年之前 |
Samuel Thibault
|
5a9ffce947
backport r12506: fix disabling out-of-order for data transfers
|
12 年之前 |
Samuel Thibault
|
bf8aa6338d
Let the OpenCL driver progress while the GPU is computing
|
12 年之前 |
Samuel Thibault
|
7d2bf0c630
Enable asynchronous CUDA kernels wherever possible
|
12 年之前 |
Samuel Thibault
|
7ce4ced85d
Let the CUDA driver progress while the GPU is computing
|
12 年之前 |