Fix bugs with C++ TOPI flatten and relu (#869)

* Fix bugs with C++ TOPI flatten and relu

* Added regression tests. Fixed typo in CMakeLists.txt. Fixed topi cpp import removed.

CMakeLists.txt

@@ -193,7 +193,7 @@ if(USE_GRAPH_RUNTIME)
 endif(USE_GRAPH_RUNTIME)
 
 if(USE_LLVM)
-  find_spackage(LLVM CONFIG REQUIRED)
+  find_package(LLVM CONFIG REQUIRED)
   include_directories(${LLVM_INCLUDE_DIRS})
   add_definitions(${LLVM_DEFINITIONS})
   set(TVM_LLVM_VERSION ${LLVM_VERSION_MAJOR}${LLVM_VERSION_MINOR})

topi/include/topi/nn.h

@@ -47,7 +47,10 @@ inline tvm::Tensor relu(const tvm::Tensor& t,
                         std::string tag = kElementWise) {
   return tvm::compute(
       t->shape,
-      [&](const tvm::Array<tvm::Var>& i) { return tvm::max(t(i), threshold); },
+      [&](const tvm::Array<tvm::Var>& i) {
+        auto threshold_const = tvm::make_const(t->dtype, threshold);
+        return tvm::max(t(i), threshold_const);
+      },
       name,
       tag);
 }

topi/include/topi/nn/flatten.h

@@ -55,7 +55,7 @@ inline Tensor flatten(const Tensor& x,
       index.push_back(i);
       std::reverse(index.begin(), index.end());
       return x(index);
-    });
+    }, name, tag);
 }
 
 }  // namespace nn

topi/python/topi/__init__.py

@@ -24,3 +24,4 @@
 from . import testing
 from . import util
 from . import rocm
+from . import cpp

topi/tests/python_cpp/test_topi_basic.py

@@ -25,7 +25,12 @@ def test_apply(func, name):
     test_apply(topi.cpp.log, "log")
     test_apply(topi.cpp.sqrt, "sqrt")
 
+def test_flatten_tag():
+    A = tvm.placeholder((3, 4), name='A')
+    B = topi.cpp.nn.flatten(A)
+    assert B.op.tag == topi.tag.INJECTIVE
 
 if __name__ == "__main__":
     test_util()
     test_ewise()
+    test_flatten_tag()

topi/tests/python_cpp/test_topi_relu.py

@@ -5,9 +5,10 @@
 import topi
 from topi.util import get_const_tuple
 
-def verify_relu(m, n):
-    A = tvm.placeholder((m, n), name='A')
+def verify_relu(m, n, dtype):
+    A = tvm.placeholder((m, n), name='A', dtype=dtype)
     B = topi.cpp.nn.relu(A)
+    assert B.dtype == dtype
 
     a_np = np.random.uniform(size=get_const_tuple(A.shape)).astype(A.dtype)
     b_np = a_np * (a_np > 0)
@@ -51,7 +52,8 @@ def verify_leaky_relu(m, alpha):
 
 
 def test_relu():
-    verify_relu(10, 128)
+    for dtype in ['float32', 'float64', 'int32', 'int16', 'int8', 'int64']:
+        verify_relu(10, 128, dtype)
 
 def test_leaky_relu():
     verify_leaky_relu(100, 0.1)