Merge branch 'develop' into rand_uniform

test/CMakeLists.txt

@@ -98,17 +98,11 @@ endfunction()
 
 function(add_test_executable TEST_NAME)
     add_executable(${TEST_NAME} EXCLUDE_FROM_ALL ${ARGN})
-    target_link_libraries(${TEST_NAME} ${CMAKE_THREAD_LIBS_INIT})
-
-    # Cmake does not add flags correctly for gcc
-    if(CMAKE_CXX_COMPILER_ID MATCHES "GNU")
-        set_target_properties(${TEST_NAME} PROPERTIES COMPILE_FLAGS -pthread LINK_FLAGS -pthread)
-    endif()
     set(TEST_COMMAND ${TEST_NAME})
     add_test_command(${TEST_NAME} ${TEST_COMMAND})
     add_dependencies(tests ${TEST_NAME})
     add_dependencies(check ${TEST_NAME})
-    target_link_libraries(${TEST_NAME} migraphx migraphx_onnx migraphx_ref)
+    target_link_libraries(${TEST_NAME} Threads::Threads migraphx migraphx_onnx migraphx_ref)
     target_include_directories(${TEST_NAME} PUBLIC include)
 endfunction(add_test_executable)
 

test/api/test_custom_op.cpp

@@ -99,7 +99,7 @@ TEST_CASE(run_sigmoid_custom_op)
     EXPECT(bool{result == migraphx::argument(s, expected_result.data())});
 }
 
-extern "C" void migraphx_test_private_disable_exception_catch(bool b);
+extern "C" MIGRAPHX_C_EXPORT void migraphx_test_private_disable_exception_catch(bool);
 
 TEST_CASE(run_sigmoid_with_incorrect_shape)
 {

test/gpu/mlir.cpp

@@ -84,7 +84,7 @@ migraphx::program create_program_from_mlir(const migraphx::module& mmlir)
     inputs.push_back(mm->add_parameter("output", mmlir.get_output_shapes().front()));
 
     migraphx::gpu::context ctx;
-    migraphx::gpu::insert_mlir(*mm, mm->end(), compile_mlir(ctx, mmlir, inputs), inputs);
+    migraphx::gpu::insert_mlir(*mm, mm->end(), compile_mlir(ctx, mmlir, inputs, {}), inputs);
     return p;
 }
 
@@ -163,7 +163,7 @@ TEST_CASE(conv_add_relu)
 {
     const std::string mlir_output = R"__migraphx__(
 module {
-  func.func @mlir_convolution(%arg0: tensor<1x2x2x2xf32>, %arg1: tensor<2x8x3x3xf32>, %arg2: tensor<1x8x4x4xf32>) -> tensor<1x2x2x2xf32> attributes {arch = "", kernel = "mixr"} {
+  func.func @mlir_convolution_add_relu(%arg0: tensor<1x2x2x2xf32>, %arg1: tensor<2x8x3x3xf32>, %arg2: tensor<1x8x4x4xf32>) -> tensor<1x2x2x2xf32> attributes {arch = "", kernel = "mixr"} {
     %0 = migraphx.convolution(%arg2, %arg1) {dilation = [1, 1], group = 1 : i64, padding = [0, 0, 0, 0], padding_mode = 0 : i64, stride = [1, 1]} : (tensor<1x8x4x4xf32>, tensor<2x8x3x3xf32>) -> tensor<1x2x2x2xf32>
     %1 = migraphx.add(%0, %arg0) : (tensor<1x2x2x2xf32>, tensor<1x2x2x2xf32>) -> tensor<1x2x2x2xf32>
     %2 = migraphx.relu(%1) : (tensor<1x2x2x2xf32>) -> tensor<1x2x2x2xf32>
@@ -191,7 +191,7 @@ TEST_CASE(quant_dot_add)
 {
     const std::string mlir_output = R"__migraphx__(
 module {
-  func.func @main(%arg0: tensor<1x5x4xi8>, %arg1: tensor<1x4x3xi8>, %arg2: tensor<1x5x3xi32>) -> tensor<1x5x3xi32> attributes {arch = "", kernel = "mixr"} {
+  func.func @mlir_quant_dot_add(%arg0: tensor<1x5x4xi8>, %arg1: tensor<1x4x3xi8>, %arg2: tensor<1x5x3xi32>) -> tensor<1x5x3xi32> attributes {arch = "", kernel = "mixr"} {
     %0 = migraphx.quant_dot(%arg0, %arg1) : (tensor<1x5x4xi8>, tensor<1x4x3xi8>) -> tensor<1x5x3xi32>
     %1 = migraphx.add(%0, %arg2) : (tensor<1x5x3xi32>, tensor<1x5x3xi32>) -> tensor<1x5x3xi32>
     return %1 : tensor<1x5x3xi32>
@@ -218,7 +218,7 @@ TEST_CASE(dot_add)
 {
     const std::string mlir_output = R"__migraphx__(
 module {
-  func.func @mlir_dot(%arg0: tensor<1x5x4xf32>, %arg1: tensor<1x4x3xf32>, %arg2: tensor<1x5x3xf32>) -> tensor<1x5x3xf32> attributes {arch = "", kernel = "mixr"} {
+  func.func @mlir_dot_add(%arg0: tensor<1x5x4xf32>, %arg1: tensor<1x4x3xf32>, %arg2: tensor<1x5x3xf32>) -> tensor<1x5x3xf32> attributes {arch = "", kernel = "mixr"} {
     %0 = migraphx.dot(%arg0, %arg1) : (tensor<1x5x4xf32>, tensor<1x4x3xf32>) -> tensor<1x5x3xf32>
     %1 = migraphx.add(%0, %arg2) : (tensor<1x5x3xf32>, tensor<1x5x3xf32>) -> tensor<1x5x3xf32>
     return %1 : tensor<1x5x3xf32>
@@ -244,7 +244,7 @@ TEST_CASE(conv_int8_dequantize_quantize)
 {
     const std::string mlir_output = R"__migraphx__(
 module {
-  func.func @main(%arg0: tensor<2x8x3x3xi8>, %arg1: tensor<1x8x4x4xi8>, %arg2: tensor<1x2x2x2xf32>, %arg3: tensor<1x2x2x2xi32>) -> tensor<1x2x2x2xi32> attributes {arch = "", kernel = "mixr"} {
+  func.func @mlir_quant_convolution_dequantizelinear_quantizelinear(%arg0: tensor<2x8x3x3xi8>, %arg1: tensor<1x8x4x4xi8>, %arg2: tensor<1x2x2x2xf32>, %arg3: tensor<1x2x2x2xi32>) -> tensor<1x2x2x2xi32> attributes {arch = "", kernel = "mixr"} {
       %0 = migraphx.quant_convolution(%arg1, %arg0) {dilation = [1, 1], group = 1 : i64, padding = [0, 0, 0, 0], padding_mode = 0 : i64, stride = [1, 1]} : (tensor<1x8x4x4xi8>, tensor<2x8x3x3xi8>) -> tensor<1x2x2x2xi32>
       %1 = migraphx.dequantizelinear(%0, %arg2, %arg3) : (tensor<1x2x2x2xi32>, tensor<1x2x2x2xf32>, tensor<1x2x2x2xi32>) -> tensor<1x2x2x2xf32>
       %2 = migraphx.quantizelinear(%1, %arg2, %arg3) : (tensor<1x2x2x2xf32>, tensor<1x2x2x2xf32>, tensor<1x2x2x2xi32>) -> tensor<1x2x2x2xi32>
@@ -277,7 +277,7 @@ TEST_CASE(dot_convert)
 {
     const std::string mlir_output = R"__migraphx__(
 module {
-  func.func @mlir_dot(%arg0: tensor<1x5x4xf32>, %arg1: tensor<1x4x3xf32>) -> tensor<1x5x3xf16> attributes {arch = "", kernel = "mixr"} {
+  func.func @mlir_dot_convert(%arg0: tensor<1x5x4xf32>, %arg1: tensor<1x4x3xf32>) -> tensor<1x5x3xf16> attributes {arch = "", kernel = "mixr"} {
     %0 = migraphx.dot(%arg0, %arg1) : (tensor<1x5x4xf32>, tensor<1x4x3xf32>) -> tensor<1x5x3xf32>
     %1 = migraphx.convert(%0) {target_type  =  1  :  i64} : (tensor<1x5x3xf32>) -> tensor<1x5x3xf16>
     return %1 : tensor<1x5x3xf16>
@@ -303,7 +303,7 @@ TEST_CASE(dot_where)
 {
     const std::string mlir_output = R"__migraphx__(
 module {
-  func.func @mlir_dot(%arg0: tensor<1x5x4xf32>, %arg1: tensor<1x4x3xf32>, %arg2: tensor<1x5x3xi8>, %arg3: tensor<1x5x3xf32>) -> tensor<1x5x3xf32> attributes {arch = "", kernel = "mixr"} {
+  func.func @mlir_dot_where(%arg0: tensor<1x5x4xf32>, %arg1: tensor<1x4x3xf32>, %arg2: tensor<1x5x3xi8>, %arg3: tensor<1x5x3xf32>) -> tensor<1x5x3xf32> attributes {arch = "", kernel = "mixr"} {
     %0 = migraphx.dot(%arg0, %arg1) : (tensor<1x5x4xf32>, tensor<1x4x3xf32>) -> tensor<1x5x3xf32>
     %1 = migraphx.where(%arg2, %0, %arg3) : (tensor<1x5x3xi8>, tensor<1x5x3xf32>, tensor<1x5x3xf32>) -> tensor<1x5x3xf32>
     return %1 : tensor<1x5x3xf32>

test/module_test.cpp

@@ -83,7 +83,7 @@ TEST_CASE(calc_implict_deps)
     auto* else_mod = p.create_module("If_5_else");
     auto l2        = else_mod->add_literal(migraphx::literal(ys, datay));
     auto a2 = else_mod->add_instruction(migraphx::make_op("if"), {cond}, {then_mod1, else_mod1});
-    auto a3 = mm->add_instruction(migraphx::make_op("get_tuple_elem", {{"index", 0}}), a2);
+    auto a3 = else_mod->add_instruction(migraphx::make_op("get_tuple_elem", {{"index", 0}}), a2);
     else_mod->add_return({a3, l2});
 
     auto ret = mm->add_instruction(migraphx::make_op("if"), {cond}, {then_mod, else_mod});
@@ -95,6 +95,15 @@ TEST_CASE(calc_implict_deps)
     EXPECT(migraphx::contains(implicit_deps.at(ret), x1));
     EXPECT(migraphx::contains(implicit_deps.at(ret), x2));
     EXPECT(migraphx::contains(implicit_deps.at(ret), y2));
+    EXPECT(migraphx::contains(implicit_deps.at(ret), lx));
+    EXPECT(migraphx::contains(implicit_deps.at(ret), ly));
+    // test for sorting
+    p.sort();
+    auto ret_inputs = ret->inputs();
+    ret_inputs.insert(ret_inputs.end(), implicit_deps.at(ret).begin(), implicit_deps.at(ret).end());
+    EXPECT(std::all_of(ret_inputs.begin(), ret_inputs.end(), [&](const auto i) {
+        return std::distance(mm->begin(), i) < std::distance(mm->begin(), ret);
+    }));
 }
 
 TEST_CASE(module_annotate)

test/onnx/.onnxrt-commit

-d3295f4329d744fe1f8419e1220e123807282b99
+21a71d52bd2074b770807b209939ec11e2c64fa7

test/op_shape_test.cpp

@@ -323,7 +323,7 @@ TEST_CASE(conv_dyn_batch)
 TEST_CASE(conv_dyn_img)
 {
     migraphx::shape input_dyn_shape  = {migraphx::shape::float_type,
-                                       {{1, 1}, {3, 3}, {5, 20}, {5, 20}}};
+                                        {{1, 1}, {3, 3}, {5, 20}, {5, 20}}};
     migraphx::shape weights_shape    = {migraphx::shape::float_type, {1, 3, 3, 3}};
     migraphx::shape output_dyn_shape = {migraphx::shape::float_type,
                                         {{1, 1}, {1, 1}, {3, 18}, {3, 18}}};
@@ -376,7 +376,7 @@ TEST_CASE(conv_autopad_dyn_batch)
 {
     // auto_pad dynamic batch
     migraphx::shape input_dyn_shape  = {migraphx::shape::float_type,
-                                       {{1, 10}, {3, 3}, {5, 5}, {5, 5}}};
+                                        {{1, 10}, {3, 3}, {5, 5}, {5, 5}}};
     migraphx::shape weights_shape    = {migraphx::shape::float_type, {1, 3, 3, 3}};
     migraphx::shape output_dyn_shape = {migraphx::shape::float_type,
                                         {{1, 10}, {1, 1}, {5, 5}, {5, 5}}};
@@ -393,7 +393,7 @@ TEST_CASE(conv_autopad_dyn_img)
 {
     // auto_pad dynamic img
     migraphx::shape input_dyn_shape  = {migraphx::shape::float_type,
-                                       {{1, 1}, {3, 3}, {5, 10}, {5, 10}}};
+                                        {{1, 1}, {3, 3}, {5, 10}, {5, 10}}};
     migraphx::shape weights_shape    = {migraphx::shape::float_type, {1, 3, 3, 3}};
     migraphx::shape output_dyn_shape = {migraphx::shape::float_type,
                                         {{1, 1}, {1, 1}, {5, 10}, {5, 10}}};
@@ -2619,6 +2619,36 @@ TEST_CASE(reshape_non_fixed_not_matching_error)
     throws_shape(migraphx::make_op("reshape", {{"dims", new_shape}}), input);
 }
 
+TEST_CASE(return_shape_tuple)
+{
+    using migraphx::shape;
+    auto op = migraphx::make_op("@return");
+    shape s0{shape::bool_type, {1, 1}};
+    shape s1{shape::float_type, {2, 3}};
+
+    std::vector<shape> s{s0, s1};
+    auto s_out = op.compute_shape(s);
+    EXPECT(s_out.type() == shape::tuple_type);
+    EXPECT(s0 == s_out.sub_shapes()[0]);
+    EXPECT(s1 == s_out.sub_shapes()[1]);
+}
+
+TEST_CASE(return_shape_half)
+{
+    using migraphx::shape;
+    auto op = migraphx::make_op("@return");
+    std::vector<shape> s{{shape::half_type}};
+    EXPECT(op.compute_shape(s) == shape{shape::half_type});
+}
+
+TEST_CASE(return_shape_empty)
+{
+    using migraphx::shape;
+    auto op = migraphx::make_op("@return");
+    std::vector<shape> s;
+    EXPECT(op.compute_shape(s) == shape{});
+}
+
 TEST_CASE(rnn)
 {
     {

test/py/CMakeLists.txt

@@ -27,7 +27,7 @@ include(PythonModules)
 function(add_py_test NAME SCRIPT)
     foreach(PYTHON_VERSION ${PYTHON_VERSIONS})
         set (ENV_COMMAND ${CMAKE_COMMAND} -E env 
-            "PYTHONPATH=$<TARGET_FILE_DIR:migraphx_py_${PYTHON_VERSION}>"
+            "PYTHONPATH=$<TARGET_FILE_DIR:migraphx_pybind_${PYTHON_VERSION}>"
             "PYTHONMALLOC=debug"
             "MALLOC_CHECK_=3"
         )
@@ -41,10 +41,8 @@ function(add_py_test NAME SCRIPT)
 
     endforeach()
 endfunction()
-foreach(PYTHON_VERSION ${PYTHON_VERSIONS})
-    add_dependencies(tests migraphx_py_${PYTHON_VERSION})
-    add_dependencies(check migraphx_py_${PYTHON_VERSION})
-endforeach()
+add_dependencies(tests migraphx_py)
+add_dependencies(check migraphx_py)
 
 add_py_test(ref test_cpu.py WORKING_DIRECTORY ${TEST_ONNX_DIR})
 add_py_test(save_load test_save_load.py WORKING_DIRECTORY ${TEST_ONNX_DIR})

test/verify/test_add_nhwc.cpp

+/*
+ * The MIT License (MIT)
+ *
+ * Copyright (c) 2015-2022 Advanced Micro Devices, Inc. All rights reserved.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to deal
+ * in the Software without restriction, including without limitation the rights
+ * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+ * copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
+ * THE SOFTWARE.
+ */
+
+#include "verify_program.hpp"
+#include <migraphx/program.hpp>
+#include <migraphx/generate.hpp>
+#include <migraphx/make_op.hpp>
+
+struct test_add_nhwc : verify_program<test_add_nhwc>
+{
+    migraphx::program create_program() const
+    {
+        migraphx::program p;
+        auto* mm = p.get_main_module();
+        auto s   = migraphx::shape::from_permutation(
+            migraphx::shape::float_type, {4, 3, 8, 8}, {0, 2, 3, 1});
+        auto x   = mm->add_parameter("x", s);
+        auto y   = mm->add_parameter("y", s);
+        auto add = mm->add_instruction(migraphx::make_op("add"), x, y);
+        mm->add_return({add});
+        return p;
+    }
+};

test/verify/test_reduce_mean_nhwc.cpp

+/*
+ * The MIT License (MIT)
+ *
+ * Copyright (c) 2015-2022 Advanced Micro Devices, Inc. All rights reserved.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to deal
+ * in the Software without restriction, including without limitation the rights
+ * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+ * copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
+ * THE SOFTWARE.
+ */
+
+#include "verify_program.hpp"
+#include <migraphx/program.hpp>
+#include <migraphx/generate.hpp>
+#include <migraphx/make_op.hpp>
+#include <migraphx/instruction.hpp>
+
+struct test_reduce_mean_nhwc : verify_program<test_reduce_mean_nhwc>
+{
+    migraphx::program create_program() const
+    {
+        migraphx::program p;
+        auto* mm = p.get_main_module();
+        auto s   = migraphx::shape::from_permutation(
+            migraphx::shape::float_type, {4, 256, 2, 2}, {0, 2, 3, 1});
+        auto x      = mm->add_parameter("x", s);
+        auto reduce = mm->add_instruction(migraphx::make_op("reduce_mean", {{"axes", {1}}}), x);
+        auto abs    = mm->add_instruction(migraphx::make_op("abs"), reduce);
+        auto sqrt   = mm->add_instruction(migraphx::make_op("sqrt"), abs);
+        mm->add_return({sqrt});
+        return p;
+    };
+};

tools/api.py

@@ -36,6 +36,8 @@ error_type = ''
 success_type = ''
 try_wrap = ''
 
+export_c_macro = 'MIGRAPHX_C_EXPORT'
+
 c_header_preamble: List[str] = []
 c_api_body_preamble: List[str] = []
 cpp_header_preamble: List[str] = []
@@ -125,7 +127,7 @@ class Type:
 
 
 header_function = Template('''
-${error_type} ${name}(${params});
+${export_c_macro} ${error_type} ${name}(${params});
 ''')
 
 function_pointer_typedef = Template('''
@@ -177,7 +179,7 @@ class CFunction:
                                **kwargs)
 
     def generate_header(self) -> str:
-        return self.substitute(header_function)
+        return self.substitute(header_function, export_c_macro=export_c_macro)
 
     def generate_function_pointer(self, name: Optional[str] = None) -> str:
         return self.substitute(function_pointer_typedef,

tools/api/api.cpp

@@ -44,7 +44,7 @@ namespace migraphx {
 
 static thread_local bool disable_exception_catch = false; // NOLINT
 
-extern "C" void migraphx_test_private_disable_exception_catch(bool b)
+extern "C" MIGRAPHX_C_EXPORT void migraphx_test_private_disable_exception_catch(bool b)
 {
     disable_exception_catch = b;
 }

tools/api/migraphx.h

@@ -26,6 +26,9 @@
 
 #include <stdlib.h>
 #include <stdbool.h>
+
+#include <migraphx/api/export.h>
+
 // Add new types here
 // clang-format off
 #define MIGRAPHX_SHAPE_VISIT_TYPES(m) \