See https://ci.linaro.org/job/ldcg-python-manylinux-tensorflow-onednn-nightly/836/display/redirect

Changes:

------------------------------------------ [...truncated 1.38 MiB...] Stats over 10 runs: max = 59.8s, min = 34.0s, avg = 42.8s, dev = 6.5s //tensorflow/compiler/tests:fft_test_cpu PASSED in 18.6s Stats over 12 runs: max = 18.6s, min = 7.2s, avg = 13.3s, dev = 4.3s //tensorflow/python/data/experimental/kernel_tests:group_by_reducer_test PASSED in 17.7s Stats over 12 runs: max = 17.7s, min = 4.5s, avg = 10.1s, dev = 4.1s //tensorflow/python/data/kernel_tests:choose_from_datasets_test PASSED in 15.5s Stats over 12 runs: max = 15.5s, min = 4.9s, avg = 7.5s, dev = 3.0s //tensorflow/python/data/kernel_tests:memory_cleanup_test_cpu PASSED in 12.3s Stats over 12 runs: max = 12.3s, min = 4.3s, avg = 6.9s, dev = 2.1s //tensorflow/python/distribute:moving_averages_test_2gpu PASSED in 23.8s Stats over 12 runs: max = 23.8s, min = 13.5s, avg = 16.7s, dev = 2.7s //tensorflow/python/distribute:moving_averages_test_cpu PASSED in 22.3s Stats over 12 runs: max = 22.3s, min = 13.2s, avg = 16.2s, dev = 2.3s //tensorflow/python/eager/polymorphic_function:polymorphic_function_test_cpu PASSED in 33.2s Stats over 15 runs: max = 33.2s, min = 13.3s, avg = 19.5s, dev = 5.6s //tensorflow/python/kernel_tests/linalg:linear_operator_low_rank_update_test_cpu PASSED in 195.8s Stats over 15 runs: max = 195.8s, min = 144.3s, avg = 158.5s, dev = 12.9s //tensorflow/python/kernel_tests/nn_ops:rnn_cell_test_cpu PASSED in 71.0s Stats over 15 runs: max = 71.0s, min = 6.7s, avg = 18.0s, dev = 17.2s //tensorflow/python/data/experimental/kernel_tests/service:dynamic_sharding_test PASSED in 13.5s Stats over 16 runs: max = 13.5s, min = 4.9s, avg = 8.5s, dev = 3.0s //tensorflow/python/data/kernel_tests:snapshot_test PASSED in 29.5s Stats over 16 runs: max = 29.5s, min = 9.3s, avg = 18.8s, dev = 6.0s //tensorflow/python/kernel_tests/control_flow:control_flow_ops_py_test_cpu PASSED in 28.5s Stats over 16 runs: max = 28.5s, min = 5.1s, avg = 9.4s, dev = 5.5s //tensorflow/python/kernel_tests/linalg:matrix_exponential_op_test PASSED in 10.1s Stats over 16 runs: max = 10.1s, min = 4.7s, avg = 6.7s, dev = 1.6s //tensorflow/python/kernel_tests/signal:dct_ops_test_cpu PASSED in 26.1s Stats over 16 runs: max = 26.1s, min = 8.1s, avg = 13.8s, dev = 4.1s //tensorflow/python/ops:image_ops_test_cpu PASSED in 31.2s Stats over 16 runs: max = 31.2s, min = 9.2s, avg = 16.7s, dev = 5.2s //tensorflow/python/data/experimental/kernel_tests/service:distributed_save_ft_test PASSED in 73.6s Stats over 17 runs: max = 73.6s, min = 11.4s, avg = 31.9s, dev = 20.9s //tensorflow/python/data/kernel_tests:map_test PASSED in 77.7s Stats over 19 runs: max = 77.7s, min = 26.8s, avg = 44.9s, dev = 10.4s //tensorflow/compiler/tests:pooling_ops_3d_test_cpu PASSED in 12.8s Stats over 20 runs: max = 12.8s, min = 4.1s, avg = 6.0s, dev = 1.9s //tensorflow/compiler/tests:pooling_ops_3d_test_cpu_mlir_bridge_test PASSED in 13.0s Stats over 20 runs: max = 13.0s, min = 4.4s, avg = 5.8s, dev = 1.9s //tensorflow/compiler/tests:pooling_ops_test_cpu PASSED in 11.6s Stats over 20 runs: max = 11.6s, min = 4.1s, avg = 5.7s, dev = 1.9s //tensorflow/compiler/tests:pooling_ops_test_cpu_mlir_bridge_test PASSED in 11.5s Stats over 20 runs: max = 11.5s, min = 4.5s, avg = 5.9s, dev = 2.0s //tensorflow/compiler/tests:stochastic_cast_op_test_cpu PASSED in 11.2s Stats over 20 runs: max = 11.2s, min = 5.9s, avg = 7.4s, dev = 1.4s //tensorflow/dtensor/python/tests:rng_test_cpu PASSED in 15.5s Stats over 20 runs: max = 15.5s, min = 6.8s, avg = 8.9s, dev = 1.9s //tensorflow/python/autograph/tests:loop_control_flow_test PASSED in 32.7s Stats over 20 runs: max = 32.7s, min = 13.3s, avg = 17.2s, dev = 3.8s //tensorflow/python/kernel_tests:metrics_test PASSED in 42.5s Stats over 20 runs: max = 42.5s, min = 8.3s, avg = 22.1s, dev = 10.8s //tensorflow/python/kernel_tests/array_ops:matrix_band_part_op_test_cpu PASSED in 13.1s Stats over 20 runs: max = 13.1s, min = 4.0s, avg = 5.1s, dev = 2.0s //tensorflow/python/kernel_tests/data_structures:barrier_ops_test PASSED in 13.2s Stats over 20 runs: max = 13.2s, min = 3.9s, avg = 6.4s, dev = 2.9s //tensorflow/python/kernel_tests/linalg:eig_op_test PASSED in 70.1s Stats over 20 runs: max = 70.1s, min = 5.0s, avg = 18.6s, dev = 19.0s //tensorflow/python/kernel_tests/linalg:linalg_grad_test_cpu PASSED in 128.8s Stats over 20 runs: max = 128.8s, min = 44.6s, avg = 81.3s, dev = 26.2s //tensorflow/python/kernel_tests/linalg:norm_op_test_cpu PASSED in 13.4s Stats over 20 runs: max = 13.4s, min = 5.4s, avg = 7.9s, dev = 1.7s //tensorflow/python/kernel_tests/linalg:normalize_op_test_cpu PASSED in 19.4s Stats over 20 runs: max = 19.4s, min = 7.2s, avg = 12.6s, dev = 3.3s //tensorflow/python/kernel_tests/linalg:qr_op_test_cpu PASSED in 186.8s Stats over 20 runs: max = 186.8s, min = 46.4s, avg = 104.4s, dev = 43.3s //tensorflow/python/kernel_tests/linalg:self_adjoint_eig_op_test_cpu PASSED in 29.5s Stats over 20 runs: max = 29.5s, min = 5.2s, avg = 13.7s, dev = 7.8s //tensorflow/python/kernel_tests/math_ops:batch_matmul_op_test_cpu PASSED in 25.2s Stats over 20 runs: max = 25.2s, min = 7.1s, avg = 14.3s, dev = 4.2s //tensorflow/python/kernel_tests/math_ops:matmul_op_test_cpu PASSED in 39.3s Stats over 20 runs: max = 39.3s, min = 17.4s, avg = 26.2s, dev = 5.2s //tensorflow/python/kernel_tests/math_ops:tensordot_op_test_cpu PASSED in 98.3s Stats over 20 runs: max = 98.3s, min = 6.7s, avg = 41.8s, dev = 29.1s //tensorflow/python/kernel_tests/nn_ops:embedding_ops_test_cpu PASSED in 41.3s Stats over 20 runs: max = 41.3s, min = 12.9s, avg = 19.3s, dev = 6.3s //tensorflow/compiler/tests:unary_ops_test_cpu FAILED in 3 out of 22 in 27.5s Stats over 22 runs: max = 27.5s, min = 4.3s, avg = 10.5s, dev = 6.8s /home/buildslave/.cache/bazel/_bazel_buildslave/fbac33eb30dbfb6b11b15a7ff5ac830d/execroot/org_tensorflow/bazel-out/aarch64-opt/testlogs/tensorflow/compiler/tests/unary_ops_test_cpu/shard_18_of_20/test.log /home/buildslave/.cache/bazel/_bazel_buildslave/fbac33eb30dbfb6b11b15a7ff5ac830d/execroot/org_tensorflow/bazel-out/aarch64-opt/testlogs/tensorflow/compiler/tests/unary_ops_test_cpu/shard_18_of_20/test_attempts/attempt_1.log /home/buildslave/.cache/bazel/_bazel_buildslave/fbac33eb30dbfb6b11b15a7ff5ac830d/execroot/org_tensorflow/bazel-out/aarch64-opt/testlogs/tensorflow/compiler/tests/unary_ops_test_cpu/shard_18_of_20/test_attempts/attempt_2.log //tensorflow/compiler/tests:unary_ops_test_cpu_mlir_bridge_test FAILED in 3 out of 22 in 28.6s Stats over 22 runs: max = 28.6s, min = 4.4s, avg = 11.2s, dev = 7.2s /home/buildslave/.cache/bazel/_bazel_buildslave/fbac33eb30dbfb6b11b15a7ff5ac830d/execroot/org_tensorflow/bazel-out/aarch64-opt/testlogs/tensorflow/compiler/tests/unary_ops_test_cpu_mlir_bridge_test/shard_18_of_20/test.log /home/buildslave/.cache/bazel/_bazel_buildslave/fbac33eb30dbfb6b11b15a7ff5ac830d/execroot/org_tensorflow/bazel-out/aarch64-opt/testlogs/tensorflow/compiler/tests/unary_ops_test_cpu_mlir_bridge_test/shard_18_of_20/test_attempts/attempt_1.log /home/buildslave/.cache/bazel/_bazel_buildslave/fbac33eb30dbfb6b11b15a7ff5ac830d/execroot/org_tensorflow/bazel-out/aarch64-opt/testlogs/tensorflow/compiler/tests/unary_ops_test_cpu_mlir_bridge_test/shard_18_of_20/test_attempts/attempt_2.log //tensorflow/python/data/kernel_tests:interleave_test PASSED in 32.5s Stats over 24 runs: max = 32.5s, min = 9.7s, avg = 19.2s, dev = 6.4s //tensorflow/python/data/kernel_tests:sample_from_datasets_test PASSED in 29.7s Stats over 24 runs: max = 29.7s, min = 4.3s, avg = 11.2s, dev = 6.7s //tensorflow/dtensor/python/tests:multi_device_spmd_test_cpu PASSED in 40.2s Stats over 25 runs: max = 40.2s, min = 25.6s, avg = 32.5s, dev = 3.9s //tensorflow/python/kernel_tests/nn_ops:conv_ops_3d_test_cpu PASSED in 61.7s Stats over 30 runs: max = 61.7s, min = 4.2s, avg = 17.0s, dev = 13.0s //tensorflow/python/data/experimental/kernel_tests/service:data_service_ops_test PASSED in 23.6s Stats over 32 runs: max = 23.6s, min = 4.5s, avg = 10.1s, dev = 4.3s //tensorflow/python/data/experimental/kernel_tests/service:worker_tags_test PASSED in 18.1s Stats over 32 runs: max = 18.1s, min = 4.3s, avg = 10.1s, dev = 4.1s //tensorflow/python/distribute:multi_process_runner_test_2gpu PASSED in 219.5s Stats over 35 runs: max = 219.5s, min = 4.9s, avg = 23.8s, dev = 37.6s //tensorflow/python/distribute:multi_process_runner_test_cpu PASSED in 218.2s Stats over 35 runs: max = 218.2s, min = 5.2s, avg = 24.2s, dev = 37.7s //tensorflow/core/kernels:stochastic_cast_op_test PASSED in 1.8s Stats over 48 runs: max = 1.8s, min = 0.4s, avg = 0.5s, dev = 0.3s //tensorflow/compiler/mlir/quantization/tensorflow/python:quantize_model_test PASSED in 58.4s Stats over 50 runs: max = 58.4s, min = 30.7s, avg = 44.9s, dev = 6.8s //tensorflow/compiler/tests:sort_ops_test_cpu PASSED in 19.1s Stats over 50 runs: max = 19.1s, min = 4.1s, avg = 11.7s, dev = 4.0s //tensorflow/compiler/tests:sort_ops_test_cpu_mlir_bridge_test PASSED in 20.8s Stats over 50 runs: max = 20.8s, min = 4.4s, avg = 11.2s, dev = 4.5s //tensorflow/python/kernel_tests/linalg:linear_operator_circulant_test_cpu PASSED in 89.0s Stats over 50 runs: max = 89.0s, min = 51.9s, avg = 71.4s, dev = 9.5s //tensorflow/python/kernel_tests/linalg/sparse:csr_sparse_matrix_dense_mat_mul_grad_test_cpu PASSED in 22.5s Stats over 50 runs: max = 22.5s, min = 6.9s, avg = 11.9s, dev = 3.8s //tensorflow/python/kernel_tests/linalg/sparse:csr_sparse_matrix_dense_mat_mul_onednn_grad_test PASSED in 24.2s Stats over 50 runs: max = 24.2s, min = 5.1s, avg = 11.9s, dev = 3.9s //tensorflow/python/kernel_tests/math_ops:cwise_ops_binary_test_cpu PASSED in 44.1s Stats over 50 runs: max = 44.1s, min = 9.0s, avg = 23.0s, dev = 8.2s //tensorflow/python/kernel_tests/math_ops:cwise_ops_test_cpu PASSED in 13.7s Stats over 50 runs: max = 13.7s, min = 4.1s, avg = 6.6s, dev = 2.0s

Executed 3086 out of 3086 tests: 3084 tests pass and 2 fail locally. There were tests whose specified size is too big. Use the --test_verbose_timeout_warnings command line option to see which ones these are. stdout_lines: <omitted> ...ignoring

TASK [tensorflow_in_container : (Python 3.9) collect list of installed Python packages] *** Wednesday 19 June 2024 04:30:55 +0100 (2:04:16.200) 3:20:50.722 ******** changed: [localhost]

TASK [tensorflow_in_container : (Python 3.9) save list of installed Python packages] *** Wednesday 19 June 2024 04:30:56 +0100 (0:00:01.695) 3:20:52.418 ******** changed: [localhost]

TASK [tensorflow_in_container : (Python 3.9) save tests output to file] ******** Wednesday 19 June 2024 04:30:57 +0100 (0:00:00.682) 3:20:53.100 ******** changed: [localhost]

TASK [tensorflow_in_container : (Python 3.9) save tests stderr output to file] *** Wednesday 19 June 2024 04:30:58 +0100 (0:00:00.694) 3:20:53.794 ******** changed: [localhost]

TASK [tensorflow_in_container : (Python 3.9) Extract number of tests run] ****** Wednesday 19 June 2024 04:30:58 +0100 (0:00:00.746) 3:20:54.541 ******** ok: [localhost]

TASK [tensorflow_in_container : (Python 3.9) Show number of tests run] ********* Wednesday 19 June 2024 04:30:58 +0100 (0:00:00.084) 3:20:54.626 ******** ok: [localhost] => ut_tests_line: 'Executed 3086 out of 3086 tests: 3084 tests pass and 2 fail locally.'

TASK [tensorflow_in_container : (Python 3.9) Extract failed tests] ************* Wednesday 19 June 2024 04:30:58 +0100 (0:00:00.043) 3:20:54.669 ******** ok: [localhost]

TASK [tensorflow_in_container : (Python 3.9) Show which tests are flaky] ******* Wednesday 19 June 2024 04:30:59 +0100 (0:00:00.163) 3:20:54.833 ******** ok: [localhost] => tests_flaky: []

TASK [tensorflow_in_container : (Python 3.9) Show which tests failed] ********** Wednesday 19 June 2024 04:30:59 +0100 (0:00:00.048) 3:20:54.881 ******** ok: [localhost] => tests_failed: - 'FAILED: //tensorflow/compiler/tests:unary_ops_test_cpu (Summary)' - 'FAILED: //tensorflow/compiler/tests:unary_ops_test_cpu_mlir_bridge_test (Summary)'

TASK [tensorflow_in_container : (Python 3.9) Convert wheel to manylinux2014 compatible one] *** Wednesday 19 June 2024 04:30:59 +0100 (0:00:00.046) 3:20:54.928 ******** fatal: [localhost]: FAILED! => changed=true rc: 1 stderr: |- + for wheel in /tf/pkg/*.whl + echo 'Checking and renaming /tf/pkg/tensorflow_aarch64-2.18.0-cp39-cp39-linux_aarch64.whl...' + python3 -m auditwheel repair --plat manylinux2014_aarch64 /tf/pkg/tensorflow_aarch64-2.18.0-cp39-cp39-linux_aarch64.whl --wheel-dir /tf/pkg + tee check.txt

real 3m0.908s user 2m12.749s sys 0m19.733s ++ grep --extended-regexp --only-matching '/tf/pkg/\S+.whl' check.txt + new_wheel=/tf/pkg/tensorflow_aarch64-2.18.0-cp39-cp39-manylinux_2_17_aarch64.manylinux2014_aarch64.whl + [[ /tf/pkg/tensorflow_aarch64-2.18.0-cp39-cp39-manylinux_2_17_aarch64.manylinux2014_aarch64.whl != /\t\f/\p\k\g/\t\e\n\s\o\r\f\l\o\w_\a\a\r\c\h\6\4-\2.\1\8.\0-\c\p\3\9-\c\p\3\9-\l\i\n\u\x_\a\a\r\c\h\6\4.\w\h\l ]] + rm /tf/pkg/tensorflow_aarch64-2.18.0-cp39-cp39-linux_aarch64.whl + wheel=/tf/pkg/tensorflow_aarch64-2.18.0-cp39-cp39-manylinux_2_17_aarch64.manylinux2014_aarch64.whl + TF_WHEEL=/tf/pkg/tensorflow_aarch64-2.18.0-cp39-cp39-manylinux_2_17_aarch64.manylinux2014_aarch64.whl + bats /usertools/wheel_verification.bats --timing stderr_lines: <omitted> stdout: |- Checking and renaming /tf/pkg/tensorflow_aarch64-2.18.0-cp39-cp39-linux_aarch64.whl... INFO:auditwheel.main_repair:Repairing tensorflow_aarch64-2.18.0-cp39-cp39-linux_aarch64.whl INFO:auditwheel.wheeltools:Previous filename tags: linux_aarch64 INFO:auditwheel.wheeltools:New filename tags: manylinux_2_17_aarch64, manylinux2014_aarch64 INFO:auditwheel.wheeltools:Previous WHEEL info tags: cp39-cp39-linux_aarch64 INFO:auditwheel.wheeltools:New WHEEL info tags: cp39-cp39-manylinux_2_17_aarch64, cp39-cp39-manylinux2014_aarch64 INFO:auditwheel.main_repair: Fixed-up wheel written to /tf/pkg/tensorflow_aarch64-2.18.0-cp39-cp39-manylinux_2_17_aarch64.manylinux2014_aarch64.whl 1..5 ok 1 Wheel is manylinux2014 (manylinux_2_17) compliant in 74313ms # Size of /tf/pkg/tensorflow_aarch64-2.18.0-cp39-cp39-manylinux_2_17_aarch64.manylinux2014_aarch64.whl is 209 / 580 megabytes. ok 2 Wheel conforms to upstream size limitations in 14ms ok 3 Wheel is installable in 61059ms ok 4 TensorFlow is importable in 4255ms not ok 5 TensorFlow has Keras in 3801ms # (in test file /usertools/wheel_verification.bats, line 72) # `python3 -c 'import sys; import tensorflow as tf; sys.exit(0 if "keras" in tf.keras.__name__ else 1)'' failed # Traceback (most recent call last): # File "/tf/venv/lib/python3.9/site-packages/tensorflow/python/util/lazy_loader.py", line 147, in _initialize # import keras # pylint: disable=g-import-not-at-top # File "/tf/venv/lib/python3.9/site-packages/keras/__init__.py", line 23, in <module> # from keras.api import _tf_keras # ImportError: cannot import name '_tf_keras' from 'keras.api' (/tf/venv/lib/python3.9/site-packages/keras/api/__init__.py) # # During handling of the above exception, another exception occurred: # # Traceback (most recent call last): # File "<string>", line 1, in <module> # File "/tf/venv/lib/python3.9/site-packages/tensorflow/python/util/lazy_loader.py", line 181, in __getattr__ # self._initialize() # File "/tf/venv/lib/python3.9/site-packages/tensorflow/python/util/lazy_loader.py", line 161, in _initialize # raise ImportError( # pylint: disable=raise-missing-from # ImportError: Keras cannot be imported. Check that it is installed. stdout_lines: <omitted>

PLAY RECAP ********************************************************************* localhost : ok=39 changed=26 unreachable=0 failed=1 skipped=7 rescued=0 ignored=1

Wednesday 19 June 2024 04:36:26 +0100 (0:05:27.535) 3:26:22.464 ******** =============================================================================== tensorflow_in_container : (Python 3.9) run TensorFlow-oneDNN-ACL git tests 7456.20s tensorflow_in_container : (Python 3.9) build TensorFlow-oneDNN-ACL git 3470.02s tensorflow_in_container : (Python 3.9) start build container --------- 1045.00s tensorflow_in_container : (Python 3.9) Convert wheel to manylinux2014 compatible one - 327.54s Cleanup after previous build ------------------------------------------- 32.73s tensorflow : Fetch TensorFlow-oneDNN-ACL from git ---------------------- 22.72s tensorflow : Remove TensorFlow-oneDNN-ACL git sources from previous build --- 4.41s Gathering Facts --------------------------------------------------------- 2.64s tensorflow_in_container : (Python 3.9) collect list of installed Python packages for build --- 1.87s tensorflow_in_container : (Python 3.9) collect list of installed Python packages --- 1.70s tensorflow_in_container : (Python 3.9) build Python versioned container --- 1.67s Create cache dirs ------------------------------------------------------- 1.43s tensorflow_in_container : (Python 3.9) create buildslave bazel cache dir --- 1.17s tensorflow : Save git commit hash to file ------------------------------- 1.14s tensorflow_in_container : (Python 3.9) cleanup Bazel install directory --- 0.90s tensorflow_in_container : (Python 3.9) Copy built wheel ----------------- 0.90s tensorflow_in_container : (Python 3.9) create link for bazelisk cache --- 0.81s tensorflow_in_container : (Python 3.9) create link for pip cache -------- 0.80s tensorflow_in_container : (Python 3.9) create link for bazel cache ------ 0.79s tensorflow_in_container : (Python 3.9) save tests stderr output to file --- 0.75s Build step 'Execute shell' marked build as failure