Fix testsuite fails on s390x (big-endian) #7309

stliibm · 2025-09-18T14:51:46Z

On big-endian machines, there are plenty of test fails due to missing/wrong byteswaps. I've created a couple of patches to fix those fails, in numpy_helper.py::from_array() and helper.py::make_tensor() and in various testcases.

Some of the fixed Testcases test_make_tensor_raw are also mentioned in the issue from 2024-07-14 created by tehbone:
s390x test failures - test_make_tensor_raw #6181

Especially the last commit "Adjust remaining tests which are using tobytes()." is not really required as all corresponding tests did not fail before. But it makes the tests more consistent as before.

In the end I'm down to 7 remaining fails:

7 failed, 5314 passed, 3660 skipped, 31 warnings
FAILED onnx/test/helper_test.py::TestHelperTensorFunctions::test_make_bfloat16_tensor_raw
FAILED onnx/test/test_backend_reference.py::OnnxBackendNodeModelTest::test_cast_BFLOAT16_to_FLOAT_cpu
FAILED onnx/test/test_backend_reference.py::OnnxBackendNodeModelTest::test_cast_FLOAT_to_BFLOAT16_cpu
FAILED onnx/test/test_backend_reference.py::OnnxBackendNodeModelTest::test_castlike_BFLOAT16_to_FLOAT_cpu
FAILED onnx/test/test_backend_reference.py::OnnxBackendNodeModelTest::test_castlike_BFLOAT16_to_FLOAT_expanded_cpu
FAILED onnx/test/test_backend_reference.py::OnnxBackendNodeModelTest::test_castlike_FLOAT_to_BFLOAT16_cpu
FAILED onnx/test/test_backend_reference.py::OnnxBackendNodeModelTest::test_castlike_FLOAT_to_BFLOAT16_expanded_cpu

Those fails happen as the ml_dtypes-bfloat16 datatype does not byteswap. I've already filed an issue:
"bfloat16: byteswap does not swap bytes #308"
There is already a pull-request fixing this issue, but the assigned reviewer has not yet responded:
"Fix bfloat16 byteswap #311"
With a locally fixed ml_dtypes, the remaining tests are also passing.

How are those dependency issues solved in onnx?
Is the required version bumped as soon as fixed and released in ml_dtypes?
Or shall we implement a workaround with a different approach to swap bytes for bfloat16 datatype in onnx-sources?

On big-endian systems like s390x, the already existing byteswapping approach does not work as expected. If the array is accessed directly, it works fine: print (array[0][0]) -29045=0x8E8B array = array.view(array.dtype.newbyteorder("<")) print (array[0][0]) -29810= 0x8B8E But tobytes() returns the same bytes before and after the existing approach: print(array.tobytes().hex()) 8e8b... Similar to to_array(), the new approach is using byteswap() and tobytes() is really returning the swapped bytes. Furthermore, I've removed _IS_LITTLE_ENDIAN and used sys.byteorder directly as also done in to_array(). This fixes those tests on s390x: onnx/test/numpy_helper_test.py::TestNumpyHelper::test_int16 FAILED onnx/test/numpy_helper_test.py::TestNumpyHelper::test_uint16 FAILED onnx/test/numpy_helper_test.py::TestNumpyHelper::test_int32 FAILED onnx/test/numpy_helper_test.py::TestNumpyHelper::test_int64 FAILED onnx/test/numpy_helper_test.py::TestNumpyHelper::test_complex128 FAILED onnx/test/numpy_helper_test.py::TestNumpyHelper::test_complex64 FAILED onnx/test/numpy_helper_test.py::TestNumpyHelper::test_float FAILED onnx/test/numpy_helper_test.py::TestNumpyHelper::test_float16 FAILED onnx/test/numpy_helper_test.py::TestNumpyHelper::test_from_dict_values_are_np_arrays_of_float FAILED onnx/test/numpy_helper_test.py::TestNumpyHelper::test_from_dict_values_are_np_arrays_of_int FAILED onnx/test/numpy_helper_test.py::TestNumpyHelper::test_from_dict_values_are_np_arrays_of_ints FAILED onnx/test/numpy_helper_test.py::TestNumpyHelper::test_to_array_from_array_00_FLOAT FAILED onnx/test/numpy_helper_test.py::TestNumpyHelper::test_to_array_from_array_03_UINT16 FAILED onnx/test/numpy_helper_test.py::TestNumpyHelper::test_to_array_from_array_04_INT16 FAILED onnx/test/numpy_helper_test.py::TestNumpyHelper::test_to_array_from_array_05_INT32 FAILED onnx/test/numpy_helper_test.py::TestNumpyHelper::test_to_array_from_array_06_INT64 FAILED onnx/test/numpy_helper_test.py::TestNumpyHelper::test_to_array_from_array_08_FLOAT16 FAILED onnx/test/numpy_helper_test.py::TestNumpyHelper::test_to_array_from_array_09_DOUBLE FAILED onnx/test/numpy_helper_test.py::TestNumpyHelper::test_to_array_from_array_10_UINT32 FAILED onnx/test/numpy_helper_test.py::TestNumpyHelper::test_to_array_from_array_11_UINT64 FAILED onnx/test/numpy_helper_test.py::TestNumpyHelper::test_to_array_from_array_12_COMPLEX64 FAILED onnx/test/numpy_helper_test.py::TestNumpyHelper::test_to_array_from_array_13_COMPLEX128 FAILED onnx/test/test_external_data.py::TestLoadExternalData_0_protobuf::test_load_external_data FAILED onnx/test/test_external_data.py::TestLoadExternalData_0_protobuf::test_load_external_data_for_model FAILED onnx/test/test_external_data.py::TestLoadExternalData_0_protobuf::test_save_external_data FAILED onnx/test/test_external_data.py::TestLoadExternalData_1_textproto::test_load_external_data FAILED onnx/test/test_external_data.py::TestLoadExternalData_1_textproto::test_load_external_data_for_model FAILED onnx/test/test_external_data.py::TestLoadExternalData_1_textproto::test_save_external_data FAILED onnx/test/test_external_data.py::TestLoadExternalDataSingleFile_0_protobuf::test_load_external_single_file_data FAILED onnx/test/test_external_data.py::TestLoadExternalDataSingleFile_0_protobuf::test_save_external_single_file_data FAILED onnx/test/test_external_data.py::TestLoadExternalDataSingleFile_1_textproto::test_load_external_single_file_data FAILED onnx/test/test_external_data.py::TestLoadExternalDataSingleFile_1_textproto::test_save_external_single_file_data FAILED onnx/test/test_external_data.py::TestSaveAllTensorsAsExternalData_0_protobuf::test_convert_model_to_external_data_converts_attribute_values FAILED onnx/test/test_external_data.py::TestSaveAllTensorsAsExternalData_0_protobuf::test_convert_model_to_external_data_from_one_file_with_location FAILED onnx/test/test_external_data.py::TestSaveAllTensorsAsExternalData_0_protobuf::test_convert_model_to_external_data_without_size_threshold FAILED onnx/test/test_external_data.py::TestSaveAllTensorsAsExternalData_0_protobuf::test_save_model_does_convert_and_saves_the_model FAILED onnx/test/test_external_data.py::TestSaveAllTensorsAsExternalData_0_protobuf::test_save_model_without_loading_external_data FAILED onnx/test/test_external_data.py::TestSaveAllTensorsAsExternalData_1_textproto::test_convert_model_to_external_data_converts_attribute_values FAILED onnx/test/test_external_data.py::TestSaveAllTensorsAsExternalData_1_textproto::test_convert_model_to_external_data_from_one_file_with_location FAILED onnx/test/test_external_data.py::TestSaveAllTensorsAsExternalData_1_textproto::test_convert_model_to_external_data_without_size_threshold FAILED onnx/test/test_external_data.py::TestSaveAllTensorsAsExternalData_1_textproto::test_save_model_does_convert_and_saves_the_model FAILED onnx/test/test_external_data.py::TestSaveAllTensorsAsExternalData_1_textproto::test_save_model_without_loading_external_data FAILED onnx/test/inference_function_test.py::TestInferenceFunctionCall::test_reshape_inference FAILED onnx/test/model_container_refeval_test.py::TestLargeOnnxReferenceEvaluator::test_large_multi_files FAILED onnx/test/model_container_refeval_test.py::TestLargeOnnxReferenceEvaluator::test_large_one_weight_file FAILED onnx/test/model_container_refeval_test.py::TestLargeOnnxReferenceEvaluator::test_large_onnx_no_large_initializer FAILED onnx/test/reference_evaluator_model_test.py::TestReferenceEvaluatorModel::test_loop_fft FAILED onnx/test/reference_evaluator_test.py::TestReferenceEvaluator::test_function_attribute_nested_nested_graph FAILED onnx/test/reference_evaluator_test.py::TestReferenceEvaluator::test_if FAILED onnx/test/reference_evaluator_test.py::TestReferenceEvaluator::test_if_function FAILED onnx/test/reference_evaluator_test.py::TestReferenceEvaluator::test_op_reduce_00 FAILED onnx/test/reference_evaluator_test.py::TestReferenceEvaluator::test_op_reduce_01 FAILED onnx/test/reference_evaluator_test.py::TestReferenceEvaluator::test_op_reduce_02 FAILED onnx/test/reference_evaluator_test.py::TestReferenceEvaluator::test_op_reduce_03 FAILED onnx/test/reference_evaluator_test.py::TestReferenceEvaluator::test_op_reduce_04 FAILED onnx/test/reference_evaluator_test.py::TestReferenceEvaluator::test_op_reduce_05 FAILED onnx/test/reference_evaluator_test.py::TestReferenceEvaluator::test_op_reduce_06 FAILED onnx/test/reference_evaluator_test.py::TestReferenceEvaluator::test_op_reduce_07 FAILED onnx/test/reference_evaluator_test.py::TestReferenceEvaluator::test_op_reduce_08 FAILED onnx/test/reference_evaluator_test.py::TestReferenceEvaluator::test_op_reduce_09 FAILED onnx/test/reference_evaluator_test.py::TestReferenceEvaluator::test_op_reduce_10 FAILED onnx/test/reference_evaluator_test.py::TestReferenceEvaluator::test_op_reduce_11 FAILED onnx/test/reference_evaluator_test.py::TestReferenceEvaluator::test_op_reduce_12 FAILED onnx/test/reference_evaluator_test.py::TestReferenceEvaluator::test_op_reduce_13 FAILED onnx/test/reference_evaluator_test.py::TestReferenceEvaluator::test_op_reduce_14 FAILED onnx/test/reference_evaluator_test.py::TestReferenceEvaluator::test_op_reduce_15 FAILED onnx/test/reference_evaluator_test.py::TestReferenceEvaluator::test_op_reduce_16 FAILED onnx/test/reference_evaluator_test.py::TestReferenceEvaluator::test_op_reduce_17 FAILED onnx/test/reference_evaluator_test.py::TestReferenceEvaluator::test_reference_evaluator_lr_clip FAILED onnx/test/tools_test.py::TestToolsFunctions::test_replace_constant FAILED onnx/test/tools_test.py::TestToolsFunctions::test_replace_constant_function FAILED onnx/test/tools_test.py::TestToolsFunctions::test_replace_constant_graph FAILED onnx/test/tools_test.py::TestToolsFunctions::test_replace_initializer FAILED onnx/test/tools_test.py::TestToolsFunctions::test_replace_range FAILED onnx/test/tools_test.py::TestToolsFunctions::test_replace_range_function FAILED onnx/test/tools_test.py::TestToolsFunctions::test_replace_range_graph FAILED Signed-off-by: Stefan Liebler <stli@linux.ibm.com>

On big-endian systems like s390x, those tests are failing as creating a numpy-array via numpy_helper.to_array(tensor) assumes that raw_data within the tensor is formatted as little-endian and then performs a byte swap on big-endian systems. This fixes those tests on s390x: onnx/test/helper_test.py::test_make_tensor_raw[TensorProto.INT16] FAILED onnx/test/helper_test.py::test_make_tensor_raw[TensorProto.INT32] FAILED onnx/test/helper_test.py::test_make_tensor_raw[TensorProto.INT64] FAILED onnx/test/helper_test.py::test_make_tensor_raw[TensorProto.FLOAT16] FAILED onnx/test/helper_test.py::test_make_tensor_raw[TensorProto.FLOAT] FAILED onnx/test/helper_test.py::test_make_tensor_raw[TensorProto.DOUBLE] FAILED onnx/test/helper_test.py::test_make_tensor_raw[TensorProto.COMPLEX64] FAILED onnx/test/helper_test.py::test_make_tensor_raw[TensorProto.COMPLEX128] FAILED onnx/test/helper_test.py::test_make_tensor_raw[TensorProto.UINT16] FAILED onnx/test/helper_test.py::test_make_tensor_raw[TensorProto.UINT32] FAILED Note: Some of those tests are also mentioned in the following issue from 2024-01-14: s390x test failures - test_make_tensor_raw onnx#6181 The same applies for test_make_bfloat16_tensor_raw. Note that this test did not fail before as byteswapping ml_dtypes.bfloat16 is currently not byteswapping at all. The test will pass as soon as either ml_dtypes.bfloat16 is fixed or an alternative approach for byteswapping is used in numpy_helper.to_array(). See ml_dtypes issue: "bfloat16: byteswap does not swap bytes onnx#308" jax-ml/ml_dtypes#308 Signed-off-by: Stefan Liebler <stli@linux.ibm.com>

On s390x, I get those "RuntimeWarning: invalid value encountered in cast" in onnx/test/helper_test.py for functions test_make_tensor_vals and test_make_tensor_raw and the UINT8, UINT16, UINT32, UINT64 data types. Fixed those warnings by using only positive floats for those datatypes. See similar commit a90d702 Use randint instead of randn to prevent undefined cast (float to uint) (onnx#4804) Signed-off-by: Stefan Liebler <stli@linux.ibm.com>

…) on s390x On big-endian systems like s390x some external tests are failing because the tensors are created with big instead of little endian bytes. After saving and loading the model, the tensors are converted to a np.ndarray where the bytes are byteswapped on big-endian machines from little to big-endian. Thus this patch creates the tensors in the correct byte order. This fixes those tests on s390x: onnx/test/test_external_data.py::TestExternalDataToArray_0_protobuf::test_save_model_with_external_data_multiple_times FAILED onnx/test/test_external_data.py::TestExternalDataToArray_0_protobuf::test_to_array_with_external_data FAILED onnx/test/test_external_data.py::TestExternalDataToArray_1_textproto::test_save_model_with_external_data_multiple_times FAILED onnx/test/test_external_data.py::TestExternalDataToArray_1_textproto::test_to_array_with_external_data FAILED Signed-off-by: Stefan Liebler <stli@linux.ibm.com>

If a numpy array is passed to helper.py::make_tensor(raw=True), it currently just uses tobytes(). Compared to numpy_helper.py:from_array() this is wrong for the 4bit datatypes as those needs to be packed. Furthermore the bytes needs to be stored in little endian byte order (only needed on big endian machines like s390x). Thus this patch adds the byteswap and the packing. Furthermore the testcase onnx/test/helper_test.py::test_make_tensor_raw() is extended to also check passing numpy arrays. Signed-off-by: Stefan Liebler <stli@linux.ibm.com>

On big endian machines, we have to ensure that the bytes stored in a tensor are converted to little endian. While searching in the sources, those are the remaining places - all testcases - where tobytes() is used. This patch either add a byteswap or is now passing a numpy array to make_tensor, which then performs the byteswap, instead of using tobytes(). Note that all corresponding tests did not fail before. Thus in theory we do not need those changes. But it is more consistent as before. Signed-off-by: Stefan Liebler <stli@linux.ibm.com>

onnx/numpy_helper.py

justinchuby

Thanks for making the fixes!

justinchuby · 2025-09-18T20:20:41Z

Is the required version bumped as soon as fixed and released in ml_dtypes?

Yes we can bump that for s390x systems.

codecov · 2025-09-18T20:25:58Z

Codecov Report

❌ Patch coverage is 92.85714% with 2 lines in your changes missing coverage. Please review.
✅ Project coverage is 54.43%. Comparing base (05b413e) to head (0b9908a).
⚠️ Report is 82 commits behind head on main.
✅ All tests successful. No failed tests found.

Files with missing lines	Patch %	Lines
onnx/numpy_helper.py	60.00%	1 Missing and 1 partial ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #7309      +/-   ##
==========================================
+ Coverage   54.38%   54.43%   +0.04%     
==========================================
  Files         511      511              
  Lines       31916    31930      +14     
  Branches     2870     2873       +3     
==========================================
+ Hits        17359    17381      +22     
+ Misses      13767    13760       -7     
+ Partials      790      789       -1

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

andife · 2025-09-19T08:11:06Z

@cyyever Maybe we should run the test for s390x (big-endian) in some docker container once a week? So far, we haven't tested anything, neither the build process nor the tests?

cyyever · 2025-09-19T08:52:35Z

@andife I agree

cyyever · 2025-09-19T08:58:49Z

We should fix the serialisation endian and record it in documentation.

stliibm · 2025-09-19T13:52:51Z

Is the required version bumped as soon as fixed and released in ml_dtypes?

Yes we can bump that for s390x systems.

Sounds great. Thus we first have to wait until the pull-request for ml_dtypes is reviewed/merged and a new release is created.

justinchuby · 2025-09-19T14:03:59Z

@stliibm do you know what the best of setting up an s390x CI on GitHub is?

This patch introduces a common function numpy_helper.py::tobytes_little_endian() to convert a numpy array to bytes. On big-endian systems it also swaps the bytes which ensures that the returned bytes are in little-endian order. Signed-off-by: Stefan Liebler <stli@linux.ibm.com>

stliibm · 2025-09-19T14:21:56Z

@stliibm do you know what the best of setting up an s390x CI on GitHub is?

From a colleague, I've got an example how to use Github action runner on s390x. Of course in the link below, it was requested for other projects:
"New Project: GGML-ORG / RamaLama #8"

Here is also an onboarding guide:
Welcome to GitHub Actions on Power, Z, and LinuxONE

I hope this helps?

onnx/numpy_helper.py

justinchuby

Thanks!

onnx/numpy_helper.py

This adds a doc-string and the versionadded directive to tobytes_little_endian(). Signed-off-by: Stefan Liebler <stli@linux.ibm.com>

The check when to swap the bytes to little-endian is now more accurate as the byteorder of the dtype is really checked. Furthermore, we avoid using byteswap() but instead use astype() as the latter one also records the correct byteorder. Signed-off-by: Stefan Liebler <stli@linux.ibm.com>

cyyever · 2025-09-26T13:13:53Z

nice

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

stliibm · 2025-09-29T09:28:15Z

@justinchuby @cyyever Thanks a lot for the review and merging the branch. I've just left a comment in the ml_dtypes PR Fix bfloat16 byteswap #311. Hopefully it gets reviewed soon. Then also the remaining bfloat16 testcases will pass.

cyyever · 2025-09-29T09:31:50Z

@stliibm Nice job

In order to get the testsuite clean on s390x (big-endian), we need the already committed onnx-commit: "Fix testsuite fails on s390x (big-endian) (onnx#7309)" onnx@f307a8d AND the ml_dtypes-commit: "Fix bfloat16 byteswap not working" jax-ml/ml_dtypes@fd7fa4a The ml_dtypes one has now landed in ml_dtypes v0.5.4 release: https://github.com/jax-ml/ml_dtypes/releases/tag/v0.5.4 "Fixed bug in byte-swap operation for custom floats (onnx#311)" This patch bumps the requirement for ml_dtypes to v0.5.4 only on s390x as discussed in the original pull-request: onnx#7309 (comment) All other architectures, still require for ml_dtypes v0.5.0 as before.

In order to get the testsuite clean on s390x (big-endian), we need the already committed onnx-commit: "Fix testsuite fails on s390x (big-endian) (onnx#7309)" onnx@f307a8d AND the ml_dtypes-commit: "Fix bfloat16 byteswap not working" jax-ml/ml_dtypes@fd7fa4a The ml_dtypes one has now landed in ml_dtypes v0.5.4 release: https://github.com/jax-ml/ml_dtypes/releases/tag/v0.5.4 "Fixed bug in byte-swap operation for custom floats (onnx#311)" This patch bumps the requirement for ml_dtypes to v0.5.4 only on s390x as discussed in the original pull-request: onnx#7309 (comment) All other architectures, still require for ml_dtypes v0.5.0 as before. Signed-off-by: Stefan Liebler <stli@linux.ibm.com>

In order to get the testsuite clean on s390x (big-endian), we need the already committed onnx-commit: "Fix testsuite fails on s390x (big-endian) (#7309)" f307a8d AND the ml_dtypes-commit: "Fix bfloat16 byteswap not working" jax-ml/ml_dtypes@fd7fa4a The ml_dtypes one has now landed in ml_dtypes v0.5.4 release: https://github.com/jax-ml/ml_dtypes/releases/tag/v0.5.4 "Fixed bug in byte-swap operation for custom floats (#311)" This patch bumps the requirement for ml_dtypes to v0.5.4 only on s390x as discussed in the original pull-request: #7309 (comment) All other architectures, still require for ml_dtypes v0.5.0 as before. Signed-off-by: Stefan Liebler <stli@linux.ibm.com>

stliibm added 6 commits September 18, 2025 15:49

stliibm requested review from a team as code owners September 18, 2025 14:51

github-project-automation bot added this to PR Tracker Sep 18, 2025

github-project-automation bot moved this to In progress in PR Tracker Sep 18, 2025

justinchuby reviewed Sep 18, 2025

View reviewed changes

onnx/numpy_helper.py Show resolved Hide resolved

justinchuby reviewed Sep 18, 2025

View reviewed changes

onnx/numpy_helper.py Outdated Show resolved Hide resolved

justinchuby reviewed Sep 18, 2025

View reviewed changes

justinchuby self-assigned this Sep 19, 2025

justinchuby reviewed Sep 19, 2025

View reviewed changes

onnx/numpy_helper.py Show resolved Hide resolved

justinchuby approved these changes Sep 19, 2025

View reviewed changes

github-project-automation bot moved this from In progress to Reviewer approved in PR Tracker Sep 19, 2025

justinchuby reviewed Sep 19, 2025

View reviewed changes

onnx/numpy_helper.py Outdated Show resolved Hide resolved

cyyever suggested changes Sep 20, 2025

View reviewed changes

onnx/numpy_helper.py Outdated Show resolved Hide resolved

github-project-automation bot moved this from Reviewer approved to Review in progress in PR Tracker Sep 20, 2025

Add doc-string to function numpy_helper.py::tobytes_little_endian()

3e09bda

This adds a doc-string and the versionadded directive to tobytes_little_endian(). Signed-off-by: Stefan Liebler <stli@linux.ibm.com>

cyyever approved these changes Sep 26, 2025

View reviewed changes

github-project-automation bot moved this from Review in progress to Reviewer approved in PR Tracker Sep 26, 2025

cyyever added the run release CIs Use this label to trigger release tests in CI label Sep 26, 2025

justinchuby added the auto update doc Generate md/proto files automatically using the CI pipeline label Sep 26, 2025

justinchuby and others added 2 commits September 26, 2025 07:52

Merge branch 'main' into stli-f42-testsuite-basecommit20250918-20250918

1619771

CI:apply auto updated documentation/backend test data

056db08

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

justinchuby removed the auto update doc Generate md/proto files automatically using the CI pipeline label Sep 26, 2025

Merge branch 'main' into stli-f42-testsuite-basecommit20250918-20250918

0b9908a

justinchuby enabled auto-merge (squash) September 26, 2025 15:08

justinchuby merged commit f307a8d into onnx:main Sep 26, 2025
34 checks passed

github-project-automation bot moved this from Reviewer approved to Done in PR Tracker Sep 26, 2025

stliibm mentioned this pull request Sep 29, 2025

Fix bfloat16 byteswap jax-ml/ml_dtypes#311

Merged

titaiwangms added the module: CI pipelines Issues related to the CI pipeline label Oct 30, 2025

stliibm mentioned this pull request Nov 24, 2025

Require ml_dtypes>=0.5.4 on s390x #7484

Merged

Fix testsuite fails on s390x (big-endian) #7309

Fix testsuite fails on s390x (big-endian) #7309

Uh oh!

Conversation

stliibm commented Sep 18, 2025

Uh oh!

Uh oh!

Uh oh!

justinchuby left a comment

Choose a reason for hiding this comment

Uh oh!

justinchuby commented Sep 18, 2025

Uh oh!

codecov bot commented Sep 18, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

andife commented Sep 19, 2025

Uh oh!

cyyever commented Sep 19, 2025

Uh oh!

cyyever commented Sep 19, 2025

Uh oh!

stliibm commented Sep 19, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

justinchuby commented Sep 19, 2025

Uh oh!

stliibm commented Sep 19, 2025

Uh oh!

Uh oh!

justinchuby left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

cyyever commented Sep 26, 2025

Uh oh!

Uh oh!

stliibm commented Sep 29, 2025

Uh oh!

cyyever commented Sep 29, 2025

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

codecov bot commented Sep 18, 2025 •

edited

Loading

stliibm commented Sep 19, 2025 •

edited

Loading