 Samuel Thibault
|
e7a67678f3
Port r17807 from Loris' work: added handle to _STARTPU_START_DRIVER_COPY macro
|
9 years ago |
|
Samuel Thibault
|
073f30fd18
fix coverity warnings
|
9 years ago |
 Nathalie Furmento
|
1fb4a4761c
src/datawizard/copy_driver.c: release opencl event when copy is completed
|
9 years ago |
|
Samuel Thibault
|
abb3c9b178
fix warnings
|
9 years ago |
|
Samuel Thibault
|
3c81d54382
Reduce a big overhead in starpu_pthread_mutex_lock/unlock by only checking for the sched mutex when locking a sched mutex
|
9 years ago |
|
Samuel Thibault
|
54f8793eb2
fix support for more that 4G tasks
|
10 years ago |
 Samuel Pitoiset
|
18e3e8138a
core: do not display the bus stats twice
|
10 years ago |
|
Samuel Thibault
|
f8c7102131
Add missing break
|
10 years ago |
|
Samuel Thibault
|
ecc274e4c0
Distinguish between fetches, prefetches and idle fetches in the paje trace
|
10 years ago |
|
Samuel Thibault
|
419f4e392c
TODOs
|
10 years ago |
|
Samuel Thibault
|
90bf39960e
Vastly improve simgrid simulation time by using wake queues
|
10 years ago |
|
Samuel Thibault
|
d06eb3bbd4
Make reclaiming prevent more fetches on the node being cleaned up, to avoid permanent battle between them
|
10 years ago |
|
Samuel Thibault
|
8a0144e402
fix crash when an interface without any_to_any is used with disks
|
10 years ago |
|
Samuel Thibault
|
fc5aa874a0
port r15822 from 1.2: fix small issues
|
10 years ago |
|
Samuel Thibault
|
1683d35a0d
Fix msg.h inclusion according to new location
|
10 years ago |
|
Samuel Thibault
|
c8efbb83e8
Inline list heads instead of keeping allocating them
|
10 years ago |
|
Nathalie Furmento
|
bb61f3808d
update cnrs copyright
|
10 years ago |
|
Samuel Thibault
|
f7ab9f4b09
Drop assertion when running with CUDA peer support in simgrid mode
|
11 years ago |
|
Samuel Thibault
|
51e895087b
Fix and test disk support with pack/unpack-only interfaces
|
11 years ago |
|
Samuel Thibault
|
aba9b7562b
Make disk test actually test all backends. Partly revert r13785 which was using a union as a structure.
|
11 years ago |
|
Samuel Thibault
|
ae1892fa7f
Improve the disk documentation, and clean the interface a bit
|
11 years ago |
|
Samuel Thibault
|
4b1a16edde
Fix University name after the fusion..
|
11 years ago |
|
Samuel Thibault
|
fa816b7fb8
Fix device selection when issuing a gpu-gpu transfer from the source
|
11 years ago |
|
Samuel Thibault
|
6a0c92daab
Avoid void* arithmetic, icc refuses it on windows...
|
11 years ago |
|
Samuel Thibault
|
7240b27585
port r12759 from 1.1: Make sure we use the local stream, and not the stream of another device. Define separate streams for GPU-GPU transfers, depending on which device emits it
|
11 years ago |
|
Samuel Thibault
|
f255d2c95a
port r12703 from 1.1: Allow data to be uninitialized but the application provide an init_cl to initialize it
|
11 years ago |
|
Samuel Thibault
|
f2b1d0f595
port r12584 from 1.1: Re-enable asynchronous transfers from CPUs, now that the issue has been found to be in application code
|
11 years ago |
|
Samuel Thibault
|
2963604386
port r12554 from 1.1: Disable timing on events for transfers, we don't use it
|
11 years ago |
|
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.
|
11 years ago |
|
Samuel Thibault
|
b4976f3d00
Use different streams for gpu-gpu transfers
|
12 years ago |
