AMDMIGraphX
AMDMIGraphX copied to clipboard
ROCmSoftwarePlatform
wait() failing for the default stream 0
This PR fixes these two issues when using the default stream on the GPU: i.e. via MIGRAPHX_ENABLE_NULL_STREAM=1:
-
Trying to
verifyon a model, (resnet50v2 here) gives this exception:context.hpp:365: get_elapsed_ms: Failed hipEventElapsedTime: device not ready -
Trying to run
perfon a model with the default stream gives more optimistic results, as the gpu-sync-up is essentially bypassed, incorrectly.
![azure-pipelines[bot] avatar](https://avatars.githubusercontent.com/in/9426?v=4 "Published by a pub.dev verified publisher"){kind=small}
Codecov Report
All modified and coverable lines are covered by tests :white_check_mark:
Additional details and impacted files
@@ Coverage Diff @@
## develop #3873 +/- ##
========================================
Coverage 92.04% 92.04%
========================================
Files 531 531
Lines 24526 24526
========================================
Hits 22573 22573
Misses 1953 1953
:rocket: New features to boost your workflow:
- :snowflake: Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
![codecov[bot] avatar](https://avatars.githubusercontent.com/in/254?v=4 "Published by a pub.dev verified publisher"){kind=small}
Since this relies on undocumented behavior, a unit test needs to be added.
https://rocm.docs.amd.com/projects/HIP/en/docs-develop/doxygen/html/group___stream.html#gabbfb9f573a6ebe8c478605ecb5504a74
This is not an undocumented behavior.
| Test | Batch | Rate new 2c990d |
Rate old 090b5b |
Diff | Compare |
|---|---|---|---|---|---|
| torchvision-resnet50 | 64 | 3,240.05 | 3,236.02 | 0.12% | :white_check_mark: |
| torchvision-resnet50_fp16 | 64 | 6,886.10 | 6,878.18 | 0.12% | :white_check_mark: |
| torchvision-densenet121 | 32 | 2,441.46 | 2,442.41 | -0.04% | :white_check_mark: |
| torchvision-densenet121_fp16 | 32 | 4,182.04 | 4,186.75 | -0.11% | :white_check_mark: |
| torchvision-inceptionv3 | 32 | 1,620.49 | 1,618.68 | 0.11% | :white_check_mark: |
| torchvision-inceptionv3_fp16 | 32 | 2,708.44 | 2,707.37 | 0.04% | :white_check_mark: |
| cadene-inceptionv4 | 16 | 755.45 | 755.87 | -0.06% | :white_check_mark: |
| cadene-resnext64x4 | 16 | 814.14 | 814.28 | -0.02% | :white_check_mark: |
| slim-mobilenet | 64 | 7,438.34 | 7,434.38 | 0.05% | :white_check_mark: |
| slim-nasnetalarge | 64 | 208.67 | 208.54 | 0.06% | :white_check_mark: |
| slim-resnet50v2 | 64 | 3,334.91 | 3,330.30 | 0.14% | :white_check_mark: |
| bert-mrpc-onnx | 8 | 1,141.85 | 1,140.97 | 0.08% | :white_check_mark: |
| bert-mrpc-tf | 1 | 460.03 | 459.23 | 0.17% | :white_check_mark: |
| pytorch-examples-wlang-gru | 1 | 345.08 | 343.39 | 0.49% | :white_check_mark: |
| pytorch-examples-wlang-lstm | 1 | 473.25 | 472.19 | 0.23% | :white_check_mark: |
| torchvision-resnet50_1 | 1 | 799.39 | 792.95 | 0.81% | :white_check_mark: |
| cadene-dpn92_1 | 1 | 414.31 | 412.79 | 0.37% | :white_check_mark: |
| cadene-resnext101_1 | 1 | 392.20 | 392.73 | -0.14% | :white_check_mark: |
| onnx-taau-downsample | 1 | 394.56 | 395.29 | -0.18% | :white_check_mark: |
| dlrm-criteoterabyte | 1 | 32.21 | 32.20 | 0.01% | :white_check_mark: |
| dlrm-criteoterabyte_fp16 | 1 | 51.27 | 51.30 | -0.05% | :white_check_mark: |
| agentmodel | 1 | 10,166.97 | 10,374.99 | -2.01% | :white_check_mark: |
| unet_fp16 | 2 | 59.40 | 59.43 | -0.05% | :white_check_mark: |
| resnet50v1_fp16 | 1 | 1,042.28 | 1,040.55 | 0.17% | :white_check_mark: |
| resnet50v1_int8 | 1 | 1,063.18 | 1,073.64 | -0.97% | :white_check_mark: |
| bert_base_cased_fp16 | 64 | 1,170.99 | 1,170.80 | 0.02% | :white_check_mark: |
| bert_large_uncased_fp16 | 32 | 356.47 | 356.49 | -0.00% | :white_check_mark: |
| bert_large_fp16 | 1 | 201.98 | 199.69 | 1.15% | :white_check_mark: |
| distilgpt2_fp16 | 16 | 2,224.43 | 2,229.36 | -0.22% | :white_check_mark: |
| yolov5s | 1 | 545.58 | 541.91 | 0.68% | :white_check_mark: |
| tinyllama | 1 | 43.65 | 43.69 | -0.09% | :white_check_mark: |
| vicuna-fastchat | 1 | 44.79 | 44.89 | -0.22% | :white_check_mark: |
| whisper-tiny-encoder | 1 | 418.12 | 417.80 | 0.08% | :white_check_mark: |
| whisper-tiny-decoder | 1 | 401.84 | 409.80 | -1.94% | :white_check_mark: |
| llama2_7b | 1 | 19.07 | 19.04 | 0.19% | :white_check_mark: |
| qwen1.5-7b | 1 | 23.42 | 23.43 | -0.06% | :white_check_mark: |
| phi3-3.8b | 1 | 26.54 | 26.55 | -0.03% | :white_check_mark: |
| mask-rcnn | 1 | 12.68 | 12.72 | -0.35% | :white_check_mark: |
| llama3-8b | 1 | 21.61 | 21.65 | -0.19% | :white_check_mark: |
| whisper-large-encoder | 1 | 10.18 | 10.18 | -0.02% | :white_check_mark: |
| whisper-large-decoder | 1 | 101.29 | 101.24 | 0.04% | :white_check_mark: |
| mistral-7b | 1 | 23.68 | 23.68 | -0.01% | :white_check_mark: |
| FLUX.1-schnell | 1 | 770.83 | 771.09 | -0.03% | :white_check_mark: |
| nan | nan | nan | nan | nan% | :x: |
This build is not recommended to merge :red_circle:
:x:bert-mrpc-tf: ERROR - check error output
2025-06-10 00:28:55.048454: I tensorflow/core/platform/cpu_feature_guard.cc:210] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
To enable the following instructions: SSE3 SSE4.1 SSE4.2 AVX AVX2 FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
WARNING: All log messages before absl::InitializeLog() is called are written to STDERR
I0000 00:00:1749533340.438932 182934 gpu_device.cc:2022] Created device /job:localhost/replica:0/task:0/device:GPU:0 with 62973 MB memory: -> device: 0, name: AMD Instinct MI250X/MI250, pci bus id: 0000:32:00.0
WARNING: All log messages before absl::InitializeLog() is called are written to STDERR
I0000 00:00:1749533341.302674 182934 mlir_graph_optimization_pass.cc:401] MLIR V1 optimization pass is not enabled
2025-06-10 00:29:09.621814: E external/local_xla/xla/service/gpu/llvm_gpu_backend/gpu_backend_lib.cc:250] bitcode module is required by this HLO module but was not found at ./opencl.bc
2025-06-10 00:29:09.621985: E external/local_xla/xla/service/gpu/llvm_gpu_backend/gpu_backend_lib.cc:250] bitcode module is required by this HLO module but was not found at ./opencl.bc
2025-06-10 00:29:09.622024: E external/local_xla/xla/service/gpu/llvm_gpu_backend/gpu_backend_lib.cc:250] bitcode module is required by this HLO module but was not found at ./opencl.bc
2025-06-10 00:29:09.622068: E external/local_xla/xla/service/gpu/llvm_gpu_backend/gpu_backend_lib.cc:250] bitcode module is required by this HLO module but was not found at ./opencl.bc
2025-06-10 00:29:09.622110: E external/local_xla/xla/service/gpu/llvm_gpu_backend/gpu_backend_lib.cc:250] bitcode module is required by this HLO module but was not found at ./opencl.bc
2025-06-10 00:29:09.622150: E external/local_xla/xla/service/gpu/llvm_gpu_backend/gpu_backend_lib.cc:250] bitcode module is required by this HLO module but was not found at ./opencl.bc
2025-06-10 00:29:09.622188: E external/local_xla/xla/service/gpu/llvm_gpu_backend/gpu_backend_lib.cc:250] bitcode module is required by this HLO module but was not found at ./opencl.bc
2025-06-10 00:29:09.622228: E external/local_xla/xla/service/gpu/llvm_gpu_backend/gpu_backend_lib.cc:250] bitcode module is required by this HLO module but was not found at ./opencl.bc
error: Failure when generating HSACO
error: Failure when generating HSACO
error: Failure when generating HSACO
error: Failure when generating HSACO
error: Failure when generating HSACO
error: Failure when generating HSACO
error: Failure when generating HSACO
error: Failure when generating HSACO
2025-06-10 00:29:09.623293: E tensorflow/compiler/mlir/tools/kernel_gen/tf_framework_c_interface.cc:228] INTERNAL: Generating device code failed.
2025-06-10 00:29:09.624389: W tensorflow/core/framework/op_kernel.cc:1829] UNKNOWN: JIT compilation failed.
2025-06-10 00:29:09.624410: I tensorflow/core/framework/local_rendezvous.cc:405] Local rendezvous is aborting with status: UNKNOWN: JIT compilation failed.
[[{{node import/bert/embeddings/LayerNorm/moments/SquaredDifference}}]]
2025-06-10 00:29:09.624421: I tensorflow/core/framework/local_rendezvous.cc:405] Local rendezvous is aborting with status: UNKNOWN: JIT compilation failed.
[[{{node import/bert/embeddings/LayerNorm/moments/SquaredDifference}}]]
[[import/loss/output/_21]]
2025-06-10 00:29:09.624437: I tensorflow/core/framework/local_rendezvous.cc:424] Local rendezvous recv item cancelled. Key hash: 11217777527359497193
Traceback (most recent call last):
File "/usr/local/lib/python3.10/dist-packages/tensorflow/python/client/session.py", line 1407, in _do_call
return fn(*args)
File "/usr/local/lib/python3.10/dist-packages/tensorflow/python/client/session.py", line 1390, in _run_fn
return self._call_tf_sessionrun(options, feed_dict, fetch_list,
File "/usr/local/lib/python3.10/dist-packages/tensorflow/python/client/session.py", line 1483, in _call_tf_sessionrun
return tf_session.TF_SessionRun_wrapper(self._session, options, feed_dict,
tensorflow.python.framework.errors_impl.UnknownError: 2 root error(s) found.
(0) UNKNOWN: JIT compilation failed.
[[{{node import/bert/embeddings/LayerNorm/moments/SquaredDifference}}]]
[[import/loss/output/_21]]
(1) UNKNOWN: JIT compilation failed.
[[{{node import/bert/embeddings/LayerNorm/moments/SquaredDifference}}]]
0 successful operations.
0 derived errors ignored.
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "/src/AMDMIGraphX/tools/accuracy/accuracy_checker.py", line 340, in
main()
File "/src/AMDMIGraphX/tools/accuracy/accuracy_checker.py", line 324, in main
y_out = sess.run(y, feed_dict=tf_dict)
File "/usr/local/lib/python3.10/dist-packages/tensorflow/python/client/session.py", line 977, in run
result = self._run(None, fetches, feed_dict, options_ptr,
File "/usr/local/lib/python3.10/dist-packages/tensorflow/python/client/session.py", line 1220, in _run
results = self._do_run(handle, final_targets, final_fetches,
File "/usr/local/lib/python3.10/dist-packages/tensorflow/python/client/session.py", line 1400, in _do_run
return self._do_call(_run_fn, feeds, fetches, targets, options,
File "/usr/local/lib/python3.10/dist-packages/tensorflow/python/client/session.py", line 1426, in _do_call
raise type(e)(node_def, op, message) # pylint: disable=no-value-for-parameter
tensorflow.python.framework.errors_impl.UnknownError: Graph execution error:
Detected at node 'import/bert/embeddings/LayerNorm/moments/SquaredDifference' defined at (most recent call last):
Node: 'import/bert/embeddings/LayerNorm/moments/SquaredDifference'
Detected at node 'import/bert/embeddings/LayerNorm/moments/SquaredDifference' defined at (most recent call last):
Node: 'import/bert/embeddings/LayerNorm/moments/SquaredDifference'
2 root error(s) found.
(0) UNKNOWN: JIT compilation failed.
[[{{node import/bert/embeddings/LayerNorm/moments/SquaredDifference}}]]
[[import/loss/output/_21]]
(1) UNKNOWN: JIT compilation failed.
[[{{node import/bert/embeddings/LayerNorm/moments/SquaredDifference}}]]
0 successful operations.
0 derived errors ignored.
Original stack trace for 'import/bert/embeddings/LayerNorm/moments/SquaredDifference'::red_circle:unet: FAILED: MIGraphX is not within tolerance - check verbose output
:red_circle:bert_large_uncased_fp16: FAILED: MIGraphX is not within tolerance - check verbose output
:red_circle:mask-rcnn: FAILED: MIGraphX is not within tolerance - check verbose output