Add std C++ components to rocm namespace and add unit tests by pfultz2 · Pull Request #4752 · ROCm/AMDMIGraphX

pfultz2 · 2026-04-08T01:15:55Z

Motivation

This adds std C++ components to rocm namespace using many of our existing implementations. In a followup PR, I will be removing the implementation from migraphx.

This will help reduce name collisions in the migraphx namespace with std named components.

Technical Details

Changelog Category

Add a CHANGELOG.md entry for any option other than Not Applicable

- Added: New functionality.
- Changed: Changes to existing functionality.
- Removed: Functionality or support that has been removed. (Compared to a previous release)
- Optimized: Component performance that has been optimized or improved.
- Resolved Issues: Known issues from a previous version that have been resolved.
- Not Applicable: This PR is not to be included in the changelog.

github-actions

Remaining comments which cannot be posted as a review comment to avoid GitHub Rate Limit

format.py

[format.py] _{reported by reviewdog 🐶}

AMDMIGraphX/test/gpu/kernels/rocm/iterator/iterator_traits.cpp

Lines 114 to 116 in 71e9d44

    
           { 
        
               EXPECT(rocm::is_same<rocm::iterator_traits<volatile int*>::reference, volatile int&>{}); 
        
           }

[format.py] _{reported by reviewdog 🐶}

AMDMIGraphX/test/gpu/kernels/rocm/iterator/iterator_traits.cpp

Lines 121 to 123 in 71e9d44

    
           { 
        
               EXPECT(rocm::is_same<rocm::iterator_traits<const volatile int*>::value_type, int>{}); 
        
           }

[format.py] _{reported by reviewdog 🐶}

AMDMIGraphX/test/gpu/kernels/rocm/iterator/iterator_traits.cpp

Lines 128 to 130 in 71e9d44

    
           { 
        
               EXPECT(rocm::is_same<rocm::iterator_traits<float*>::value_type, float>{}); 
        
           }

[format.py] _{reported by reviewdog 🐶}

AMDMIGraphX/test/gpu/kernels/rocm/iterator/iterator_traits.cpp

Lines 133 to 135 in 71e9d44

    
           { 
        
               EXPECT(rocm::is_same<rocm::iterator_traits<double*>::difference_type, rocm::ptrdiff_t>{}); 
        
           }

[format.py] _{reported by reviewdog 🐶}

AMDMIGraphX/test/gpu/kernels/rocm/iterator/iterator_traits.cpp

Lines 138 to 140 in 71e9d44

    
           { 
        
               EXPECT(rocm::is_same<rocm::iterator_traits<char*>::reference, char&>{}); 
        
           }

[format.py] _{reported by reviewdog 🐶}

AMDMIGraphX/test/gpu/kernels/rocm/iterator/iterator_traits.cpp

Lines 154 to 156 in 71e9d44

    
           { 
        
               EXPECT(rocm::is_same<rocm::iterator_traits<mock_iterator>::difference_type, int>{}); 
        
           }

[format.py] _{reported by reviewdog 🐶}

AMDMIGraphX/test/gpu/kernels/rocm/iterator/iterator_traits.cpp

Lines 159 to 161 in 71e9d44

    
           { 
        
               EXPECT(rocm::is_same<rocm::iterator_traits<mock_iterator>::value_type, double>{}); 
        
           }

[format.py] _{reported by reviewdog 🐶}

AMDMIGraphX/test/gpu/kernels/rocm/iterator/iterator_traits.cpp

Lines 164 to 166 in 71e9d44

    
           { 
        
               EXPECT(rocm::is_same<rocm::iterator_traits<mock_iterator>::pointer, double*>{}); 
        
           }

[format.py] _{reported by reviewdog 🐶}

AMDMIGraphX/test/gpu/kernels/rocm/iterator/iterator_traits.cpp

Lines 169 to 171 in 71e9d44

    
           { 
        
               EXPECT(rocm::is_same<rocm::iterator_traits<mock_iterator>::reference, double&>{}); 
        
           }

[format.py] _{reported by reviewdog 🐶}

AMDMIGraphX/test/gpu/kernels/rocm/utility/forward.cpp

Lines 33 to 35 in 71e9d44

    
           { 
        
               return true; 
        
           }

[format.py] _{reported by reviewdog 🐶}

AMDMIGraphX/test/gpu/kernels/rocm/utility/forward.cpp

Lines 39 to 41 in 71e9d44

    
           { 
        
               return false; 
        
           }

[format.py] _{reported by reviewdog 🐶}

AMDMIGraphX/test/gpu/kernels/rocm/utility/forward.cpp

Lines 140 to 142 in 71e9d44

    
           { 
        
               return rocm::forward<T>(x); 
        
           }

[format.py] _{reported by reviewdog 🐶}

AMDMIGraphX/test/gpu/kernels/rocm/utility/integer_sequence.cpp

Lines 32 to 34 in 71e9d44

    
           { 
        
               EXPECT(rocm::is_same<rocm::integer_sequence<int, 0, 1, 2>::value_type, int>{}); 
        
           }

[format.py] _{reported by reviewdog 🐶}

AMDMIGraphX/test/gpu/kernels/rocm/utility/integer_sequence.cpp

Lines 37 to 39 in 71e9d44

    
           { 
        
               EXPECT(rocm::is_same<rocm::integer_sequence<long, 0, 1>::value_type, long>{}); 
        
           }

[format.py] _{reported by reviewdog 🐶}

AMDMIGraphX/test/gpu/kernels/rocm/utility/integer_sequence.cpp

Lines 42 to 44 in 71e9d44

    
           { 
        
               EXPECT(rocm::is_same<rocm::integer_sequence<char, 'a', 'b'>::value_type, char>{}); 
        
           }

[format.py] _{reported by reviewdog 🐶}

AMDMIGraphX/test/gpu/kernels/rocm/utility/integer_sequence.cpp

Lines 47 to 49 in 71e9d44

    
           { 
        
               EXPECT(rocm::is_same<rocm::integer_sequence<rocm::size_t, 0, 1>::value_type, rocm::size_t>{}); 
        
           }

[format.py] _{reported by reviewdog 🐶}

AMDMIGraphX/test/gpu/kernels/rocm/utility/integer_sequence.cpp

Lines 58 to 60 in 71e9d44

    
           { 
        
               static_assert(rocm::integer_sequence<int, 10, 20, 30>::size() == 3); 
        
           }

[format.py] _{reported by reviewdog 🐶}

AMDMIGraphX/test/gpu/kernels/rocm/utility/integer_sequence.cpp

Lines 63 to 65 in 71e9d44

    
           { 
        
               static_assert(rocm::integer_sequence<int, 0, 1, 2, 3, 4>::size() == 5); 
        
           }

[format.py] _{reported by reviewdog 🐶}

AMDMIGraphX/test/gpu/kernels/rocm/utility/integer_sequence.cpp

Lines 76 to 78 in 71e9d44

    
           { 
        
               EXPECT(rocm::is_same<rocm::index_sequence<>, rocm::integer_sequence<rocm::size_t>>{}); 
        
           }

[format.py] _{reported by reviewdog 🐶}

AMDMIGraphX/test/gpu/kernels/rocm/utility/integer_sequence.cpp

Lines 83 to 85 in 71e9d44

    
           { 
        
               EXPECT(rocm::is_same<rocm::index_sequence<0>::value_type, rocm::size_t>{}); 
        
           }

[format.py] _{reported by reviewdog 🐶}

AMDMIGraphX/test/gpu/kernels/rocm/utility/integer_sequence.cpp

Lines 90 to 92 in 71e9d44

    
           { 
        
               EXPECT(rocm::is_same<rocm::make_integer_sequence<int, 0>, rocm::integer_sequence<int>>{}); 
        
           }

[format.py] _{reported by reviewdog 🐶}

AMDMIGraphX/test/gpu/kernels/rocm/utility/integer_sequence.cpp

Lines 95 to 97 in 71e9d44

    
           { 
        
               EXPECT(rocm::is_same<rocm::make_integer_sequence<int, 1>, rocm::integer_sequence<int, 0>>{}); 
        
           }

[format.py] _{reported by reviewdog 🐶}

AMDMIGraphX/test/gpu/kernels/rocm/utility/integer_sequence.cpp

Lines 126 to 128 in 71e9d44

    
           { 
        
               EXPECT(rocm::is_same<rocm::make_index_sequence<0>, rocm::index_sequence<>>{}); 
        
           }

[format.py] _{reported by reviewdog 🐶}

AMDMIGraphX/test/gpu/kernels/rocm/utility/integer_sequence.cpp

Lines 131 to 133 in 71e9d44

    
           { 
        
               EXPECT(rocm::is_same<rocm::make_index_sequence<1>, rocm::index_sequence<0>>{}); 
        
           }

[format.py] _{reported by reviewdog 🐶}

AMDMIGraphX/test/gpu/kernels/rocm/utility/integer_sequence.cpp

Lines 136 to 138 in 71e9d44

    
           { 
        
               EXPECT(rocm::is_same<rocm::make_index_sequence<4>, rocm::index_sequence<0, 1, 2, 3>>{}); 
        
           }

[format.py] _{reported by reviewdog 🐶}

AMDMIGraphX/test/gpu/kernels/rocm/utility/integer_sequence.cpp

Lines 145 to 147 in 71e9d44

    
           { 
        
               EXPECT(rocm::is_same<rocm::index_sequence_for<>, rocm::index_sequence<>>{}); 
        
           }

[format.py] _{reported by reviewdog 🐶}

AMDMIGraphX/test/gpu/kernels/rocm/utility/integer_sequence.cpp

Lines 150 to 152 in 71e9d44

    
           { 
        
               EXPECT(rocm::is_same<rocm::index_sequence_for<int>, rocm::index_sequence<0>>{}); 
        
           }

[format.py] _{reported by reviewdog 🐶}

AMDMIGraphX/test/gpu/kernels/rocm/utility/integer_sequence.cpp

Lines 161 to 163 in 71e9d44

    
           { 
        
               static_assert(rocm::index_sequence_for<int, int, int, int>::size() == 4); 
        
           }

[format.py] _{reported by reviewdog 🐶}

AMDMIGraphX/test/gpu/kernels/rocm/utility/move.cpp

Lines 33 to 35 in 71e9d44

    
           { 
        
               return true; 
        
           }

[format.py] _{reported by reviewdog 🐶}

AMDMIGraphX/test/gpu/kernels/rocm/utility/move.cpp

Lines 39 to 41 in 71e9d44

    
           { 
        
               return false; 
        
           }

[format.py] _{reported by reviewdog 🐶}

AMDMIGraphX/test/gpu/kernels/rocm/utility/move.cpp

Lines 72 to 74 in 71e9d44

    
           { 
        
               EXPECT(rocm::is_rvalue_reference<decltype(rocm::move(42))>{}); 
        
           }

github-actions

Remaining comments which cannot be posted as a review comment to avoid GitHub Rate Limit

format.py

[format.py] _{reported by reviewdog 🐶}

AMDMIGraphX/test/gpu/kernels/rocm/utility/move.cpp

Lines 72 to 74 in b3a907d

    
           { 
        
               EXPECT(rocm::is_rvalue_reference<decltype(rocm::move(42))>{}); 
        
           }

This reverts commit 443fcfe.

gh-app-migraphx-bot-pr-write · 2026-06-24T14:50:02Z

Test	Batch	New Rate (53e9a1)	Old Rate (5c50d0)*	Diff	Status
torchvision-resnet50	64	3,273.91	3,278.37	-0.14%	✅
torchvision-resnet50_fp16	64	7,607.22	7,611.38	-0.05%	✅
torchvision-densenet121	32	2,502.19	2,501.41	0.03%	✅
torchvision-densenet121_fp16	32	5,043.66	5,004.64	0.78%	✅
torchvision-inceptionv3	32	2,061.62	2,059.27	0.11%	✅
torchvision-inceptionv3_fp16	32	4,419.90	4,417.65	0.05%	✅
cadene-inceptionv4	16	822.31	819.80	0.31%	✅
cadene-resnext64x4	16	783.11	783.52	-0.05%	✅
slim-mobilenet	64	8,431.47	8,394.63	0.44%	✅
slim-nasnetalarge	64	229.39	197.60	16.08%	🔆
slim-resnet50v2	64	3,198.38	3,185.87	0.39%	✅
bert-mrpc-onnx	8	1,166.71	1,161.65	0.44%	✅
bert-mrpc-tf	1	490.76	486.30	0.92%	✅
pytorch-examples-wlang-gru	1	487.91	469.65	3.89%	✅
pytorch-examples-wlang-lstm	1	391.39	379.72	3.07%	✅
torchvision-resnet50_1	1	1,038.22	1,023.59	1.43%	✅
cadene-dpn92_1	1	477.69	436.69	9.39%	🔆
cadene-resnext101_1	1	364.57	364.25	0.09%	✅
onnx-taau-downsample	1	848.90	845.97	0.35%	✅
dlrm-criteoterabyte	1	32.68	32.54	0.42%	✅
dlrm-criteoterabyte_fp16	1	52.58	52.59	-0.02%	✅
agentmodel	1	11,504.92	9,452.06	21.72%	🔆
unet_fp16	2	59.16	59.01	0.24%	✅
resnet50v1_fp16	1	1,376.97	1,378.39	-0.10%	✅
resnet50v1_int8	1	1,839.90	1,626.43	13.12%	🔆
bert_base_cased_fp16	64	1,102.10	1,102.97	-0.08%	✅
bert_large_uncased_fp16	32	346.77	346.63	0.04%	✅
bert_large_fp16	1	206.26	205.51	0.36%	✅
distilgpt2_fp16	16	2,093.73	2,098.82	-0.24%	✅
yolov5s	1	561.13	562.00	-0.15%	✅
tinyllama	1	45.98	45.93	0.11%	✅
vicuna-fastchat	1	44.12	44.06	0.13%	✅
whisper-tiny-encoder	1	414.49	415.56	-0.26%	✅
whisper-tiny-decoder	1	409.14	410.79	-0.40%	✅
llama2_7b	1	20.96	21.01	-0.25%	✅
qwen1.5-7b	1	23.65	23.60	0.20%	✅
phi3-3.8b	1	26.89	26.80	0.33%	✅
llama3-8b	1	21.84	21.70	0.63%	✅
whisper-large-encoder	1	10.21	10.22	-0.05%	✅
whisper-large-decoder	1	104.02	104.52	-0.48%	✅
mistral-7b	1	23.86	23.74	0.50%	✅
FLUX.1-schnell	1	770.93	723.54	6.55%	🔆

Check flagged results 🔆

* No develop baseline was found for this PR's branch point; compared against the latest available develop run instead.

gh-app-migraphx-bot-pr-write · 2026-06-24T14:50:04Z

Test	Status	Result
bert-mrpc-onnx	✅	PASSED: MIGraphX meets tolerance
bert-mrpc-tf	✅	PASSED: MIGraphX meets tolerance
pytorch-examples-wlang-gru	✅	PASSED: MIGraphX meets tolerance
pytorch-examples-wlang-lstm	✅	PASSED: MIGraphX meets tolerance
dlrm-criteoterabyte	✅	PASSED: MIGraphX meets tolerance
agentmodel	✅	PASSED: MIGraphX meets tolerance
unet	✅	PASSED: MIGraphX meets tolerance
resnet50v1	✅	PASSED: MIGraphX meets tolerance
bert_base_cased_fp16	✅	PASSED: MIGraphX meets tolerance
bert_large_uncased_fp16	🔴	FAILED: MIGraphX is not within tolerance - check verbose output
bert_large	✅	PASSED: MIGraphX meets tolerance
yolov5s	✅	PASSED: MIGraphX meets tolerance
tinyllama	✅	PASSED: MIGraphX meets tolerance
vicuna-fastchat	✅	PASSED: MIGraphX meets tolerance
whisper-tiny-encoder	✅	PASSED: MIGraphX meets tolerance
whisper-tiny-decoder	✅	PASSED: MIGraphX meets tolerance
distilgpt2_fp16	🔴	FAILED: MIGraphX is not within tolerance - check verbose output
llama2_7b	✅	PASSED: MIGraphX meets tolerance
qwen1.5-7b	✅	PASSED: MIGraphX meets tolerance
phi3-3.8b	✅	PASSED: MIGraphX meets tolerance
llama3-8b	✅	PASSED: MIGraphX meets tolerance
whisper-large-encoder	✅	PASSED: MIGraphX meets tolerance
whisper-large-decoder	✅	PASSED: MIGraphX meets tolerance
mistral-7b	✅	PASSED: MIGraphX meets tolerance
FLUX.1-schnell	✅	PASSED: MIGraphX meets tolerance

Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>

Copilot

Pull request overview

This PR introduces a rocm::-namespaced, device-friendly subset of common C++ standard-library components (type traits, utility, iterators, functional, algorithms, limits, etc.) under the GPU kernel headers, and adds/expands GPU kernel tests to validate the new APIs. It also updates the GPU kernel test harness and build wiring to discover tests recursively and to propagate the originating source filename into HIP compilation for improved diagnostics.

Changes:

Add new src/targets/gpu/kernels/include/rocm/* headers providing rocm:: equivalents of common std utilities (type traits, utility helpers, iterators, functional ops, algorithms, limits/stdint/stddef, etc.).
Add GPU kernel unit tests under test/gpu/kernels/rocm/** for the new rocm:: components.
Update GPU kernel test/build plumbing (recursive test discovery, suite naming, and hip_compile_options::src_file propagation).

Reviewed changes

Copilot reviewed 76 out of 76 changed files in this pull request and generated 4 comments.

Show a summary per file

File	Description
test/gpu/kernels/rocm/utility/swap.cpp	New GPU kernel tests for `rocm::swap`.
test/gpu/kernels/rocm/utility/move.cpp	New GPU kernel tests for `rocm::move`.
test/gpu/kernels/rocm/utility/integer_sequence.cpp	New GPU kernel tests for `rocm::integer_sequence` and related aliases.
test/gpu/kernels/rocm/utility/forward.cpp	New GPU kernel tests for `rocm::forward`.
test/gpu/kernels/rocm/type_traits_relationships.cpp	New GPU kernel tests covering a variety of `rocm::type_traits` relationship traits.
test/gpu/kernels/rocm/stdint.cpp	New GPU kernel tests validating `rocm::stdint` type sizes.
test/gpu/kernels/rocm/std_numeric_limits.cpp	New GPU kernel tests comparing `rocm::numeric_limits` to `std::numeric_limits` (host path).
test/gpu/kernels/rocm/numeric_limits.cpp	New GPU kernel tests validating `rocm::numeric_limits` API surface and types.
test/gpu/kernels/rocm/iterator/reverse_iterator.cpp	New GPU kernel tests for `rocm::reverse_iterator`.
test/gpu/kernels/rocm/iterator/iterator_traits.cpp	New GPU kernel tests for `rocm::iterator_traits` and iterator tags.
test/gpu/kernels/rocm/integral_constant.cpp	New GPU kernel tests for `rocm::integral_constant` operators and conversions.
test/gpu/kernels/rocm/functional/operations.cpp	New GPU kernel tests for `rocm::functional` operation objects and transparency.
test/gpu/kernels/rocm/enable_if.cpp	New GPU kernel tests for `rocm::enable_if` / `enable_if_t`.
test/gpu/kernels/rocm/.clang-tidy	New clang-tidy overrides for the rocm kernel tests subtree.
test/gpu/kernels/main.cpp	Enhance kernel test discovery (templates/register macros) and suite naming; pass `src_file` into HIP compile options.
test/gpu/kernels/CMakeLists.txt	Switch to recursive discovery of kernel test sources.
src/targets/gpu/kernels/include/rocm/utility/swap.hpp	New `rocm::swap` implementation.
src/targets/gpu/kernels/include/rocm/utility/move.hpp	New `rocm::move` implementation.
src/targets/gpu/kernels/include/rocm/utility/integer_sequence.hpp	New `rocm::integer_sequence` and helpers (`index_sequence`, etc.).
src/targets/gpu/kernels/include/rocm/utility/forward.hpp	New `rocm::forward` implementation.
src/targets/gpu/kernels/include/rocm/utility/declval.hpp	New `rocm::declval` implementation.
src/targets/gpu/kernels/include/rocm/utility.hpp	Umbrella include for rocm utility headers.
src/targets/gpu/kernels/include/rocm/type_traits.hpp	New rocm-flavored type traits and helpers/macros for kernels.
src/targets/gpu/kernels/include/rocm/stdint.hpp	New rocm stdint typedefs (host vs RTC handling).
src/targets/gpu/kernels/include/rocm/stddef.hpp	New rocm stddef typedefs (`size_t`, `ptrdiff_t`, etc.).
src/targets/gpu/kernels/include/rocm/limits.hpp	New `rocm::numeric_limits` and related enums/helpers for kernels.
src/targets/gpu/kernels/include/rocm/iterator/reverse_iterator.hpp	New `rocm::reverse_iterator` implementation.
src/targets/gpu/kernels/include/rocm/iterator/iterator_traits.hpp	New `rocm::iterator_traits` plus iterator tag hierarchy.
src/targets/gpu/kernels/include/rocm/iterator.hpp	Umbrella include for rocm iterator headers.
src/targets/gpu/kernels/include/rocm/integral_constant.hpp	New `rocm::integral_constant` and boolean constants.
src/targets/gpu/kernels/include/rocm/functional/operations.hpp	New rocm functional operation objects (`plus`, `less`, etc.).
src/targets/gpu/kernels/include/rocm/functional.hpp	Umbrella include for rocm functional headers.
src/targets/gpu/kernels/include/rocm/config.hpp	New rocm config macros (inline namespace, host/device annotations, deduction guide workaround hook).
src/targets/gpu/kernels/include/rocm/bit.hpp	New rocm bit utilities (`bit_cast`, `popcount`, rotations, etc.).
src/targets/gpu/kernels/include/rocm/assert.hpp	New rocm assert/check helpers (host/device compatible printing + abort).
src/targets/gpu/kernels/include/rocm/array.hpp	New `rocm::array` and helpers (iterators, swap, to_array).
src/targets/gpu/kernels/include/rocm/algorithm/upper_bound.hpp	New rocm `upper_bound` algorithm.
src/targets/gpu/kernels/include/rocm/algorithm/transform.hpp	New rocm `transform` algorithms (unary/binary).
src/targets/gpu/kernels/include/rocm/algorithm/stable_sort.hpp	New rocm `stable_sort` implementation.
src/targets/gpu/kernels/include/rocm/algorithm/sort.hpp	New rocm `sort` implementation.
src/targets/gpu/kernels/include/rocm/algorithm/search.hpp	New rocm `search` implementation.
src/targets/gpu/kernels/include/rocm/algorithm/rotate.hpp	New rocm `rotate` implementation.
src/targets/gpu/kernels/include/rocm/algorithm/none_of.hpp	New rocm `none_of` implementation.
src/targets/gpu/kernels/include/rocm/algorithm/min_element.hpp	New rocm `min_element` implementation.
src/targets/gpu/kernels/include/rocm/algorithm/merge.hpp	New rocm `merge` implementation.
src/targets/gpu/kernels/include/rocm/algorithm/max_element.hpp	New rocm `max_element` implementation.
src/targets/gpu/kernels/include/rocm/algorithm/lower_bound.hpp	New rocm `lower_bound` implementation.
src/targets/gpu/kernels/include/rocm/algorithm/iter_swap.hpp	New rocm `iter_swap` implementation.
src/targets/gpu/kernels/include/rocm/algorithm/is_sorted.hpp	New rocm `is_sorted` implementation.
src/targets/gpu/kernels/include/rocm/algorithm/is_sorted_until.hpp	New rocm `is_sorted_until` implementation.
src/targets/gpu/kernels/include/rocm/algorithm/iota.hpp	New rocm `iota` implementation.
src/targets/gpu/kernels/include/rocm/algorithm/inner_product.hpp	New rocm `inner_product` implementation.
src/targets/gpu/kernels/include/rocm/algorithm/for_each.hpp	New rocm `for_each` implementation.
src/targets/gpu/kernels/include/rocm/algorithm/find.hpp	New rocm `find` implementation.
src/targets/gpu/kernels/include/rocm/algorithm/find_if.hpp	New rocm `find_if` implementation.
src/targets/gpu/kernels/include/rocm/algorithm/fill.hpp	New rocm `fill` implementation.
src/targets/gpu/kernels/include/rocm/algorithm/equal.hpp	New rocm `equal` implementation.
src/targets/gpu/kernels/include/rocm/algorithm/copy.hpp	New rocm `copy` implementation.
src/targets/gpu/kernels/include/rocm/algorithm/copy_if.hpp	New rocm `copy_if` implementation.
src/targets/gpu/kernels/include/rocm/algorithm/any_of.hpp	New rocm `any_of` implementation.
src/targets/gpu/kernels/include/rocm/algorithm/all_of.hpp	New rocm `all_of` implementation.
src/targets/gpu/kernels/include/rocm/algorithm/accumulate.hpp	New rocm `accumulate` implementation.
src/targets/gpu/kernels/include/rocm/algorithm.hpp	Umbrella include for rocm algorithm headers.
src/targets/gpu/kernels/include/rocm/.clang-tidy	New clang-tidy overrides for rocm kernel headers subtree.
src/targets/gpu/kernels/include/migraphx/kernels/test.hpp	Add no-op template/register test macros for parsing/registration.
src/targets/gpu/include/migraphx/gpu/compile_hip_code_object.hpp	Add `hip_compile_options::src_file` for labeling the main compilation unit.
src/targets/gpu/compile_hip_code_object.cpp	Use `options.src_file` for the main source instead of hardcoding `main.cpp`.
src/targets/gpu/CMakeLists.txt	Broaden kernel header scanning, add deduction-guide capability check, and adjust clang-tidy file-check stubs.
src/compile_src.cpp	Pass full relative source paths to the compiler for better traceability/diagnostics.
CMakeLists.txt	Exclude rocm kernel header/test define-prefix regions from cppcheck definePrefix heuristics.

+using intmax_t  = int64_t;
+using uintmax_t = uint64_t;
+using intmax_t  = int64_t;


+    constexpr pointer operator->() const
+    {
+        iterator_type tmp = current;
+        --tmp;
+        return tmp;
+    }


+
+} // namespace ROCM_INLINE_NS
+} // namespace rocm
+#endif // MIGRAPHX_GUARD_FUNCTIONAL_OPERATIONS_HPP


+    file(WRITE ${CMAKE_CURRENT_BINARY_DIR}/kernels/include/${KERNEL_DIR}/${KERNEL_BASE_FILE}.cpp "#include <${KERNEL_DIR}/${KERNEL_BASE_FILE}.hpp>\n")
+    target_sources(migraphx_gpu_kernel_file_check PRIVATE ${CMAKE_CURRENT_BINARY_DIR}/kernels/include/${KERNEL_DIR}/${KERNEL_BASE_FILE}.cpp)


pfultz2 added 30 commits February 11, 2026 12:24

Add initial rocm cxx tests

b8d025d

Format

3cfc07e

Add more type traits

89747b1

Format

99cbeb0

Add more type traits tests

e9b8a6a

Format

706b2d5

Move declval

2676ef7

Add operations class

aa49111

Format

c0acd27

Add reverse iterator

5764ba7

Format

4330fb6

Add array class

032c06d

Format

140a37d

Add algorithm header

3bc95ac

Format

b488046

Update algorithm header

8fb398a

Format

825515b

Add remove_pointer

8cb326a

Format

8abb772

Add remove_cvref

0abfa7d

FOrmat

13e37c5

Add is_pointer trait

9583af1

Format

7ae6f56

Use builtin

a491154

Add bit_cast

1e0c7ef

Format

b228e3b

Add more bit functions

c9f2730

Format

dbf5b7d

Simplify rotr

63f025f

Add more type traits

9ddf29f

Add workaround macro

71e9d44

github-actions Bot reviewed May 12, 2026

View reviewed changes

pfultz2 added 3 commits May 12, 2026 15:08

Fix tests

0f542c8

Format

6b944a9

Add clang-tidy customization

b3a907d

github-actions Bot reviewed May 12, 2026

View reviewed changes

pfultz2 and others added 9 commits May 12, 2026 17:28

Apply fixits

207681c

Format

ddf5120

Fix unused function warning

0c3571a

Format

cdccdbb

Fix tidy warnings

0958e9d

Fix macro prefix

06ed1fd

Merge branch 'develop' into rocm-cxx

9c0aec5

Fix format issue

443fcfe

Revert "Fix format issue"

cce9144

This reverts commit 443fcfe.

pfultz2 and others added 8 commits July 14, 2026 14:53

Merge branch 'develop' into rocm-cxx

f1b5b5e

Merge branch 'develop' into rocm-cxx

25556d8

Define assert for cppcheck

f3a8d13

Fix cppcheck warnings

3a83a73

Format

e836859

Update year

899f29c

Fix tidy warnings

53e9a14

Update src/targets/gpu/kernels/include/rocm/limits.hpp

caa9859

Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>

pfultz2 marked this pull request as ready for review July 24, 2026 14:16

pfultz2 requested a review from causten as a code owner July 24, 2026 14:16

Copilot AI review requested due to automatic review settings July 24, 2026 14:16

Copilot started reviewing on behalf of pfultz2 July 24, 2026 14:18 View session

Copilot AI reviewed Jul 24, 2026

View reviewed changes

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Add std C++ components to rocm namespace and add unit tests#4752

Add std C++ components to rocm namespace and add unit tests#4752
pfultz2 wants to merge 65 commits into
developfrom
rocm-cxx

pfultz2 commented Apr 8, 2026

Uh oh!

github-actions Bot left a comment

Uh oh!

github-actions Bot left a comment

Uh oh!

gh-app-migraphx-bot-pr-write Bot commented Jun 24, 2026 •

edited

Loading

Uh oh!

gh-app-migraphx-bot-pr-write Bot commented Jun 24, 2026 •

edited

Loading

Uh oh!

Copilot AI left a comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

	{
	EXPECT(rocm::is_same<rocm::iterator_traits<volatile int*>::reference, volatile int&>{});
	}

	{
	EXPECT(rocm::is_same<rocm::integer_sequence<int, 0, 1, 2>::value_type, int>{});
	}

	{
	static_assert(rocm::integer_sequence<int, 10, 20, 30>::size() == 3);
	}

	{
	EXPECT(rocm::is_same<rocm::index_sequence<>, rocm::integer_sequence<rocm::size_t>>{});
	}

	{
	EXPECT(rocm::is_same<rocm::make_integer_sequence<int, 0>, rocm::integer_sequence<int>>{});
	}

	{
	EXPECT(rocm::is_same<rocm::make_index_sequence<0>, rocm::index_sequence<>>{});
	}

	{
	static_assert(rocm::index_sequence_for<int, int, int, int>::size() == 4);
	}

	{
	EXPECT(rocm::is_rvalue_reference<decltype(rocm::move(42))>{});
	}

		file(WRITE ${CMAKE_CURRENT_BINARY_DIR}/kernels/include/${KERNEL_DIR}/${KERNEL_BASE_FILE}.cpp "#include <${KERNEL_DIR}/${KERNEL_BASE_FILE}.hpp>\n")
		target_sources(migraphx_gpu_kernel_file_check PRIVATE ${CMAKE_CURRENT_BINARY_DIR}/kernels/include/${KERNEL_DIR}/${KERNEL_BASE_FILE}.cpp)

Uh oh!

Conversation

pfultz2 commented Apr 8, 2026

Motivation

Technical Details

Changelog Category

Uh oh!

github-actions Bot left a comment

Choose a reason for hiding this comment

Uh oh!

github-actions Bot left a comment

Choose a reason for hiding this comment

Uh oh!

gh-app-migraphx-bot-pr-write Bot commented Jun 24, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

gh-app-migraphx-bot-pr-write Bot commented Jun 24, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

gh-app-migraphx-bot-pr-write Bot commented Jun 24, 2026 •

edited

Loading

gh-app-migraphx-bot-pr-write Bot commented Jun 24, 2026 •

edited

Loading