Skip to content

Commit 6a83d37

Browse files
authored
Merge pull request #51 from reyoung/fix_lstm_avx_error
Fix bug in NO_AVX when using lstm
2 parents 3304de7 + 4a880f0 commit 6a83d37

File tree

2 files changed

+7
-4
lines changed

2 files changed

+7
-4
lines changed

paddle/.gitignore

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -40,3 +40,4 @@ HPPL_ERROR_LOG
4040
unittest.list
4141
proto
4242
dist
43+
setup.py

paddle/cuda/CMakeLists.txt

Lines changed: 6 additions & 4 deletions
Original file line numberDiff line numberDiff line change
@@ -1,8 +1,11 @@
1+
set(AVX_SOURCES
2+
src/hl_math.cc
3+
src/hl_avx_functions.cc
4+
)
15
set(CUDA_SOURCES
26
src/hl_time.cc
3-
src/hl_math.cc
47
src/hl_cpu_functions.cc
5-
src/hl_avx_functions.cc)
8+
${AVX_SOURCES})
69

710
set(CUDA_CXX_WITH_GPU_SOURCES
811
src/hl_cuda_cublas.cc
@@ -12,7 +15,7 @@ set(CUDA_CXX_WITH_GPU_SOURCES
1215
set_source_files_properties(${CUDA_CXX_WITH_GPU_SOURCES}
1316
PROPERTIES COMPILE_FLAGS "-D__NVCC__")
1417

15-
set_source_files_properties(${CUDA_SOURCES}
18+
set_source_files_properties(${AVX_SOURCES}
1619
PROPERTIES COMPILE_FLAGS "-mavx")
1720

1821
set(CUDA_DSO_SOURCES
@@ -73,4 +76,3 @@ endif()
7376

7477
add_style_check_target(paddle_cuda ${CUDA_SOURCES})
7578
add_style_check_target(paddle_cuda ${CUDA_HEADERS})
76-
# add_style_check_target(hppl ${HPPL_CU_SOURCES}) # TODO(yuyang18): Format hppl style

0 commit comments

Comments
 (0)