 Samuel Thibault
|
ee4677f7de
more efficient loops
|
%!s(int64=14) %!d(string=hai) anos |
 Nathalie Furmento
|
2c86c03a5d
src/core/perfmodel/perfmodel_bus.c: improve display
|
%!s(int64=14) %!d(string=hai) anos |
|
Nathalie Furmento
|
d9e385fd42
tools/starpu_machine_display.c: check return value of starpu_init
|
%!s(int64=14) %!d(string=hai) anos |
|
Nathalie Furmento
|
2f80d102bc
tests/datawizard/acquire_cb.c: return EXIT_FAILURE when computation fails
|
%!s(int64=14) %!d(string=hai) anos |
|
Nathalie Furmento
|
5ee833f68b
src: only define variable when needed
|
%!s(int64=14) %!d(string=hai) anos |
|
Samuel Thibault
|
bd93dcb88e
Simply get back to original ifdef/endif :)
|
%!s(int64=14) %!d(string=hai) anos |
|
Samuel Thibault
|
29b7f32f77
Provide good examples by always using streams for cublas
|
%!s(int64=14) %!d(string=hai) anos |
|
Nathalie Furmento
|
7074411fbf
src/util/starpu_cublas.c: no need to define the function if CUDA is not enabled
|
%!s(int64=14) %!d(string=hai) anos |
|
Nathalie Furmento
|
f413aaa1cd
src/util/starpu_cublas.c: add missing #ifdef
|
%!s(int64=14) %!d(string=hai) anos |
|
Samuel Thibault
|
b613d58104
Assume that we have at least cuda 3.1, so that we can make our BLAS examples always use streams
|
%!s(int64=14) %!d(string=hai) anos |
|
Samuel Thibault
|
51a7e8c979
Use cudaMemcpyAsync instead of cudaMemcpy
|
%!s(int64=14) %!d(string=hai) anos |
|
Samuel Thibault
|
78ac4a07ca
follow-up r7018: add missing declarations
|
%!s(int64=14) %!d(string=hai) anos |
|
Samuel Thibault
|
d2cd1868e2
provide good examples by always using cudaMemsetAsync, not cudaMemset
|
%!s(int64=14) %!d(string=hai) anos |
|
Samuel Thibault
|
57e59bf2d9
provide good examples by always using cudaMemcpyAsync, not cudaMemcpy
|
%!s(int64=14) %!d(string=hai) anos |
|
Samuel Thibault
|
f18180c477
synchronize only the transfer stream
|
%!s(int64=14) %!d(string=hai) anos |
|
Samuel Thibault
|
3e58d92492
Add 2d shadow filters
|
%!s(int64=14) %!d(string=hai) anos |
|
Samuel Thibault
|
cb492ad84b
Keep documentation coherent with C convention
|
%!s(int64=14) %!d(string=hai) anos |
|
Samuel Thibault
|
972caa3d96
Fix build without OpenGL headers
|
%!s(int64=14) %!d(string=hai) anos |
|
Samuel Thibault
|
70864cd9ea
there are some optimizations with eager, just not wise scheduling
|
%!s(int64=14) %!d(string=hai) anos |
|
Samuel Thibault
|
1d7dcdc5b6
fix compilation warning
|
%!s(int64=14) %!d(string=hai) anos |
|
Samuel Thibault
|
c3610e1434
drop unused factor
|
%!s(int64=14) %!d(string=hai) anos |
|
Samuel Thibault
|
9803389d55
fix number
|
%!s(int64=14) %!d(string=hai) anos |
|
Samuel Thibault
|
fd06e119aa
more comments
|
%!s(int64=14) %!d(string=hai) anos |
|
Samuel Thibault
|
cf5341ae88
more details on the shadow at work
|
%!s(int64=14) %!d(string=hai) anos |
|
Samuel Thibault
|
6bc0d82b7c
fix non-cuda/opencl build
|
%!s(int64=14) %!d(string=hai) anos |
|
Samuel Thibault
|
a88b26f1a7
update changelog
|
%!s(int64=14) %!d(string=hai) anos |
|
Samuel Thibault
|
ee1e7926a3
Put single-combined-worker in nice order for starpu_machine_display
|
%!s(int64=14) %!d(string=hai) anos |
|
Samuel Thibault
|
3b540f359f
document that starpu_machine_display shows combined workers
|
%!s(int64=14) %!d(string=hai) anos |
|
Samuel Thibault
|
65688ef4d8
print topology before bandwidth
|
%!s(int64=14) %!d(string=hai) anos |
|
Samuel Thibault
|
1aeeab7392
mergeinfo for backport
|
%!s(int64=14) %!d(string=hai) anos |
