Skip to content

Commit 18404c0

Browse files
author
Timmy
committed
Merge pull request #144 from TimmyLiu/develop
fast dtrsm (online compilation) and splitting big dgemm calls
2 parents c47ef12 + a08507d commit 18404c0

34 files changed

+5855
-73
lines changed

src/client/client.cpp

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -178,7 +178,7 @@ int main(int argc, char *argv[])
178178
deviceType = CL_DEVICE_TYPE_ALL;
179179
}
180180

181-
if( profileCount > 1 )
181+
if( profileCount >= 1 )
182182
{
183183
commandQueueFlags |= CL_QUEUE_PROFILING_ENABLE;
184184
}

src/library/CMakeLists.txt

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -641,6 +641,7 @@ include_directories(${OPENCL_INCLUDE_DIRS}
641641
${clBLAS_SOURCE_DIR}/library/blas/AutoGemm
642642
${clBLAS_SOURCE_DIR}/library/blas/AutoGemm/UserGemmKernelSources
643643
${clBLAS_SOURCE_DIR}/library/blas/specialCases/include
644+
${clBLAS_SOURCE_DIR}/library/blas/trtri
644645
)
645646

646647
option( BLAS_DUMP_CLBLAS_KERNELS "Force the library to dump OpenCL kernels to disk" OFF )

0 commit comments

Comments
 (0)