 Nathalie Furmento
|
ef395f6e4b
when possible, use sizeof instead of a hard-coded value for a string length
|
%!s(int64=7) %!d(string=hai) anos |
|
Nathalie Furmento
|
a8cc467700
tests,examples: fix cppcheck warnings
|
%!s(int64=8) %!d(string=hai) anos |
 Samuel Thibault
|
2bcc6c9758
document what tests test
|
%!s(int64=9) %!d(string=hai) anos |
|
Nathalie Furmento
|
e3b44209ed
update copyright
|
%!s(int64=10) %!d(string=hai) anos |
|
Samuel Thibault
|
d35c3ec66b
Make quickcheck faster
|
%!s(int64=10) %!d(string=hai) anos |
|
Samuel Thibault
|
6757bca1b5
Remove implementation array terminator in tests, examples, and documentation, since it is not actually used
|
%!s(int64=10) %!d(string=hai) anos |
|
Nathalie Furmento
|
cc9aecb792
merge trunk
|
%!s(int64=10) %!d(string=hai) anos |
|
Samuel Thibault
|
4b1a16edde
Fix University name after the fusion..
|
%!s(int64=10) %!d(string=hai) anos |
 Andra Hugo
|
5593439277
changing perf_model structure: arch now contains several devices, so we can have for eg one device STARPU_CPU with 2 cores one device STARPU_CUDA_WORKER with 1 core
|
%!s(int64=11) %!d(string=hai) anos |
|
Nathalie Furmento
|
101eceed7a
tests/perfmodels/regression_based.c: Enable asynchronous CUDA kernels
|
%!s(int64=11) %!d(string=hai) anos |
|
Samuel Thibault
|
c60d91cb3f
Enable asynchronous OpenCL wherever possible
|
%!s(int64=11) %!d(string=hai) anos |
|
Samuel Thibault
|
7d2bf0c630
Enable asynchronous CUDA kernels wherever possible
|
%!s(int64=11) %!d(string=hai) anos |
|
Nathalie Furmento
|
5fdde967b0
tests: partly revert #11510 and turn all cpu implementations codelets not static so that MIC code is able to successfully call dlsym on the function
|
%!s(int64=11) %!d(string=hai) anos |
|
Nathalie Furmento
|
0c6c658cb1
tests: turn function static when it makes sense to do so
|
%!s(int64=11) %!d(string=hai) anos |
|
Samuel Thibault
|
c05e9e394f
import work from Brice & Damien to adapt tests to MIC and SCC
|
%!s(int64=12) %!d(string=hai) anos |
|
Nathalie Furmento
|
d56555097b
include: starpu_scheduler.h is no longer automatically included by
|
%!s(int64=12) %!d(string=hai) anos |
|
Samuel Thibault
|
a8df5c71bc
Add starpu_codelet_profile, a tool that draws the profile of a codelet.
|
%!s(int64=12) %!d(string=hai) anos |
|
Nathalie Furmento
|
a4f3da6a64
tests: only include starpu.h as it includes all the other starpu_ .h files
|
%!s(int64=12) %!d(string=hai) anos |
|
Samuel Thibault
|
d2cd1868e2
provide good examples by always using cudaMemsetAsync, not cudaMemset
|
%!s(int64=12) %!d(string=hai) anos |
 Cyril Roelandt
|
d6fd124ae4
Check the return value of starpu_opencl_unload_opencl().
|
%!s(int64=13) %!d(string=hai) anos |
|
Cyril Roelandt
|
86e332292f
Make tests/perfmodels/regression_based and tests/perfmodels/non_linear_regression_based work with OpenCL.
|
%!s(int64=13) %!d(string=hai) anos |
|
Cyril Roelandt
|
578cd4b507
Fix the size passed to cudaMemset().
|
%!s(int64=13) %!d(string=hai) anos |
|
Cyril Roelandt
|
974a90787d
Fix calls to memset.
|
%!s(int64=13) %!d(string=hai) anos |
|
Samuel Thibault
|
021892beba
Make more checks
|
%!s(int64=13) %!d(string=hai) anos |
|
Cyril Roelandt
|
685f84fcdb
Include valgrind.h in helper.h, not directly in the tests.
|
%!s(int64=13) %!d(string=hai) anos |
|
Cyril Roelandt
|
3b9c9c7b4c
Skip computations in the tests when running on valgrind.
|
%!s(int64=13) %!d(string=hai) anos |
|
Nathalie Furmento
|
9aaa7caf99
tests: if starpu_init() returns -ENODEV, skip the test, do not fail it
|
%!s(int64=13) %!d(string=hai) anos |
|
Samuel Thibault
|
d083734f5b
Fix test with only opencl
|
%!s(int64=13) %!d(string=hai) anos |
|
Nathalie Furmento
|
cbfe978b11
tests/perfmodels/regression_based.c: use new API to define data handles and access modes (this is needed as application accesses task's data before submitting it)
|
%!s(int64=13) %!d(string=hai) anos |
|
Nathalie Furmento
|
7599748470
update code w.r.t coding style
|
%!s(int64=13) %!d(string=hai) anos |
