[test]mv fluid op [reader, prim_ops, nccl, reduce_ops, lite] to test (#53429)

* [test]mv fluid reader to test/

* [test]mv fluid op prim_ops to test/cpp/fluid/prim_ops

* [test]mv fluid op nccl to /test/cpp/fluid/nccl/

* [test]mv fluid op reduce_ops to test/cpp/fluid/reduce_ops

* [test]mv fluid op lite to test/cpp/fluid/lite

* [test]fix lite

* [test]fix prim op path

* [fluid]clean prim ops cmakelists

paddle/fluid/operators/lite/CMakeLists.txt

 op_library(lite_engine_op DEPS lite_engine lite_tensor_utils)
-cc_test_old(test_lite_engine_op SRCS lite_engine_op_test.cc DEPS lite_engine_op
-            analysis)

paddle/fluid/operators/nccl/CMakeLists.txt

@@ -22,17 +22,3 @@ if(WITH_GPU OR WITH_ROCM)
       ${OPERATOR_DEPS} nccl_common
       PARENT_SCOPE)
 endif()
-
-if(WITH_GPU AND NOT WIN32)
-  nv_test(
-    nccl_op_test
-    SRCS nccl_op_test.cu.cc
-    DEPS nccl_op gpu_info device_context)
-endif()
-
-if(WITH_ROCM AND NOT WIN32)
-  hip_test(
-    nccl_op_test
-    SRCS nccl_op_test.cu.cc
-    DEPS nccl_op gpu_info device_context)
-endif()

paddle/fluid/operators/prim_ops/CMakeLists.txt

@@ -4,42 +4,3 @@ if(WITH_UNITY_BUILD)
   include(unity_build_rule.cmake)
 endif()
 register_operators()
-
-set(PRIM_OP_SRCS
-    reshape_p_op.cc
-    broadcast_p_op.cc
-    reduce_sum_p_op.cc
-    transpose_p_op.cc
-    split_p_op.cc
-    concat_p_op.cc
-    slice_select_p_op.cc
-    slice_assign_p_op.cc
-    gather_p_op.cc
-    scatter_add_p_op.cc
-    add_p_op.cc
-    sub_p_op.cc
-    mul_p_op.cc
-    div_p_op.cc
-    sqrt_p_op.cc
-    tanh_p_op.cc
-    sin_p_op.cc
-    cos_p_op.cc
-    exp_p_op.cc
-    matmul_p_op.cc
-    fill_constant_p_op.cc
-    log_p_op.cc
-    select_p_op.cc
-    eq_p_op.cc
-    gt_p_op.cc
-    ge_p_op.cc
-    ne_p_op.cc
-    pow_p_op.cc
-    max_p_op.cc
-    erf_p_op.cc
-    bernoulli_p_op.cc
-    abs_p_op.cc
-    cast_p_op.cc
-    rsqrt_p_op.cc
-    uniform_random_p_op.cc)
-
-cc_test_old(prim_op_test SRCS prim_op_test.cc ${PRIM_OP_SRCS} DEPS op_registry)

paddle/fluid/operators/reader/CMakeLists.txt

@@ -35,6 +35,5 @@ reader_library(create_py_reader_op SRCS create_py_reader_op.cc DEPS py_reader)
 
 op_library(read_op DEPS py_reader buffered_reader)
 
-cc_test(reader_blocking_queue_test SRCS reader_blocking_queue_test.cc)
 # Export local libraries to parent
 # set(READER_LIBRARY ${LOCAL_READER_LIBS} PARENT_SCOPE)

paddle/fluid/operators/reduce_ops/CMakeLists.txt

@@ -12,24 +12,3 @@ if(WITH_GPU)
 else()
   register_operators()
 endif()
-
-if(WITH_GPU)
-  if(${CMAKE_CUDA_COMPILER_VERSION} LESS 11.0)
-    nv_test(
-      check_reduce_rank_test
-      SRCS check_reduce_rank_test.cu
-      DEPS tensor cub)
-  else()
-    nv_test(
-      check_reduce_rank_test
-      SRCS check_reduce_rank_test.cu
-      DEPS tensor)
-  endif()
-endif()
-
-if(WITH_ROCM)
-  hip_test(
-    check_reduce_rank_test
-    SRCS check_reduce_rank_test.cu
-    DEPS tensor)
-endif()

test/cpp/fluid/CMakeLists.txt

@@ -7,17 +7,18 @@ endif()
 # add_subdirectory(dlnne)
 # add_subdirectory(elementwise)
 # add_subdirectory(fused)
-# if(WITH_LITE)
-#   add_subdirectory(lite)
-# endif()
+if(WITH_LITE)
+  add_subdirectory(lite)
+endif()
 # add_subdirectory(math)
 # if(WITH_MKLDNN)
 #   add_subdirectory(mkldnn)
 # endif()
-# add_subdirectory(nccl)
+add_subdirectory(nccl)
 # if(WITH_PSCORE)
 #   add_subdirectory(pscore)
 # endif()
-# add_subdirectory(prim_ops)
-# add_subdirectory(reduce_ops)
+add_subdirectory(prim_ops)
+add_subdirectory(reader)
+add_subdirectory(reduce_ops)
 # add_subdirectory(tensorrt)

test/cpp/fluid/lite/CMakeLists.txt

+cc_test_old(test_lite_engine_op SRCS lite_engine_op_test.cc DEPS lite_engine_op
+            analysis)

test/cpp/fluid/nccl/CMakeLists.txt

+if(WITH_GPU AND NOT WIN32)
+  nv_test(
+    nccl_op_test
+    SRCS nccl_op_test.cu.cc
+    DEPS nccl_op gpu_info device_context)
+endif()
+
+if(WITH_ROCM AND NOT WIN32)
+  hip_test(
+    nccl_op_test
+    SRCS nccl_op_test.cu.cc
+    DEPS nccl_op gpu_info device_context)
+endif()

test/cpp/fluid/prim_ops/CMakeLists.txt

+set(PRIM_OP_PATH ${PADDLE_SOURCE_DIR}/paddle/fluid/operators/prim_ops/)
+
+set(PRIM_OP_SRCS
+    ${PRIM_OP_PATH}reshape_p_op.cc
+    ${PRIM_OP_PATH}broadcast_p_op.cc
+    ${PRIM_OP_PATH}reduce_sum_p_op.cc
+    ${PRIM_OP_PATH}transpose_p_op.cc
+    ${PRIM_OP_PATH}split_p_op.cc
+    ${PRIM_OP_PATH}concat_p_op.cc
+    ${PRIM_OP_PATH}slice_select_p_op.cc
+    ${PRIM_OP_PATH}slice_assign_p_op.cc
+    ${PRIM_OP_PATH}gather_p_op.cc
+    ${PRIM_OP_PATH}scatter_add_p_op.cc
+    ${PRIM_OP_PATH}add_p_op.cc
+    ${PRIM_OP_PATH}sub_p_op.cc
+    ${PRIM_OP_PATH}mul_p_op.cc
+    ${PRIM_OP_PATH}div_p_op.cc
+    ${PRIM_OP_PATH}sqrt_p_op.cc
+    ${PRIM_OP_PATH}tanh_p_op.cc
+    ${PRIM_OP_PATH}sin_p_op.cc
+    ${PRIM_OP_PATH}cos_p_op.cc
+    ${PRIM_OP_PATH}exp_p_op.cc
+    ${PRIM_OP_PATH}matmul_p_op.cc
+    ${PRIM_OP_PATH}fill_constant_p_op.cc
+    ${PRIM_OP_PATH}log_p_op.cc
+    ${PRIM_OP_PATH}select_p_op.cc
+    ${PRIM_OP_PATH}eq_p_op.cc
+    ${PRIM_OP_PATH}gt_p_op.cc
+    ${PRIM_OP_PATH}ge_p_op.cc
+    ${PRIM_OP_PATH}ne_p_op.cc
+    ${PRIM_OP_PATH}pow_p_op.cc
+    ${PRIM_OP_PATH}max_p_op.cc
+    ${PRIM_OP_PATH}erf_p_op.cc
+    ${PRIM_OP_PATH}bernoulli_p_op.cc
+    ${PRIM_OP_PATH}abs_p_op.cc
+    ${PRIM_OP_PATH}cast_p_op.cc
+    ${PRIM_OP_PATH}rsqrt_p_op.cc
+    ${PRIM_OP_PATH}uniform_random_p_op.cc)
+cc_test_old(prim_op_test SRCS prim_op_test.cc ${PRIM_OP_SRCS} DEPS op_registry)

test/cpp/fluid/reader/CMakeLists.txt

+cc_test(reader_blocking_queue_test SRCS reader_blocking_queue_test.cc)

test/cpp/fluid/reduce_ops/CMakeLists.txt

+if(WITH_GPU)
+  if(${CMAKE_CUDA_COMPILER_VERSION} LESS 11.0)
+    nv_test(
+      check_reduce_rank_test
+      SRCS check_reduce_rank_test.cu
+      DEPS tensor cub)
+  else()
+    nv_test(
+      check_reduce_rank_test
+      SRCS check_reduce_rank_test.cu
+      DEPS tensor)
+  endif()
+endif()
+
+if(WITH_ROCM)
+  hip_test(
+    check_reduce_rank_test
+    SRCS check_reduce_rank_test.cu
+    DEPS tensor)
+endif()