Skip to content

Commit c22ee4d

Browse files
authored
Fix copy fast tuner (CNugteren#598)
* Fixed c_ld documentation. * Ran the generator script. * Fixed copy_fast tuner which was messed up by clang format reordering the includes. * Fixed the copy pad kernel which had the same issue as copy fast
1 parent b252e4d commit c22ee4d

File tree

2 files changed

+4
-2
lines changed

2 files changed

+4
-2
lines changed

src/tuning/kernels/copy_fast.hpp

Lines changed: 2 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -34,8 +34,9 @@ TunerSettings CopyGetTunerSettings(const int, const Arguments<T>& args) {
3434
settings.kernel_family = "copy";
3535
settings.kernel_name = "CopyMatrixFast";
3636
settings.sources =
37-
#include "../src/kernels/level3/copy_fast.opencl"
3837
#include "../src/kernels/level3/level3.opencl"
38+
// Comment to prevent reordering of includes
39+
#include "../src/kernels/level3/copy_fast.opencl"
3940
;
4041

4142
// Buffer sizes

src/tuning/kernels/copy_pad.hpp

Lines changed: 2 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -34,8 +34,9 @@ TunerSettings PadGetTunerSettings(const int, const Arguments<T>& args) {
3434
settings.kernel_family = "pad";
3535
settings.kernel_name = "CopyPadMatrix";
3636
settings.sources =
37-
#include "../src/kernels/level3/copy_pad.opencl"
3837
#include "../src/kernels/level3/level3.opencl"
38+
// Comment to prevent reordering of includes
39+
#include "../src/kernels/level3/copy_pad.opencl"
3940
;
4041

4142
// Buffer sizes

0 commit comments

Comments
 (0)