TaoLv commented on issue #13362: Add NHWC layout support to Pooling (cuDNN only) URL: https://github.com/apache/incubator-mxnet/pull/13362#issuecomment-444771242 Tested this PR on CPU w/ and w/o MKL-DNN and got error message as below. w/o MKL-DNN: ``` Traceback (most recent call last): File "test_pool_nhwc.py", line 28, in <module> test_pooling_nhwc() File "test_pool_nhwc.py", line 21, in test_pooling_nhwc out_nhwc = mx.nd.Pooling(data=inp_nhwc, kernel=kernel, pool_type='max', global_pool=False, stride=stride, pad=pad, layout='NHWC') File "<string>", line 132, in Pooling File "/home/lvtao/Workspace/mxnet/python/mxnet/_ctypes/ndarray.py", line 92, in _imperative_invoke ctypes.byref(out_stypes))) File "/home/lvtao/Workspace/mxnet/python/mxnet/base.py", line 252, in check_call raise MXNetError(py_str(_LIB.MXGetLastError())) mxnet.base.MXNetError: [12:35:52] src/operator/nn/./pooling-inl.h:167: Check failed: param_.layout.value() == kNCW || param_.layout.value() == kNCHW || param_.layout.value() == kNCDHW Need CuDNN for layout support ``` w/o MKL-DNN: ``` terminate called after throwing an instance of 'dmlc::Error' what(): [12:28:11] src/engine/./threaded_engine.h:380: std::exception A fatal error occurred in asynchronous engine operation. If you do not know what caused this error, you can try set environment variable MXNET_ENGINE_TYPE to NaiveEngine and run with debugger (i.e. gdb). This will force all operations to be synchronous and backtrace will give you the series of calls that lead to this error. Remember to set MXNET_ENGINE_TYPE back to empty after debugging. Stack trace returned 10 entries: [bt] (0) /home/lvtao/Workspace/mxnet/python/mxnet/../../lib/libmxnet.so(dmlc::StackTrace()+0x42) [0x7f547889afb6] [bt] (1) /home/lvtao/Workspace/mxnet/python/mxnet/../../lib/libmxnet.so(dmlc::LogMessageFatal::~LogMessageFatal()+0x1b) [0x7f547889b233] [bt] (2) /home/lvtao/Workspace/mxnet/python/mxnet/../../lib/libmxnet.so(mxnet::engine::ThreadedEngine::ExecuteOprBlock(mxnet::RunContext, mxnet::engine::OprBlock*)+0x66e) [0x7f547b7ad24c] [bt] (3) /home/lvtao/Workspace/mxnet/python/mxnet/../../lib/libmxnet.so(void mxnet::engine::ThreadedEnginePerDevice::CPUWorker<(dmlc::ConcurrentQueueType)0>(mxnet::Context, mxnet::engine::ThreadedEnginePerDevice::ThreadWorkerBlock<(dmlc::ConcurrentQueueType)0>*, std::shared_ptr<dmlc::ManualEvent> const&)+0x8b) [0x7f547b7bfe9d] [bt] (4) /home/lvtao/Workspace/mxnet/python/mxnet/../../lib/libmxnet.so(mxnet::engine::ThreadedEnginePerDevice::PushToExecute(mxnet::engine::OprBlock*, bool)::{lambda()#1}::operator()() const::{lambda(std::shared_ptr<dmlc::ManualEvent>)#1}::operator()(dmlc::ManualEvent) const+0x33) [0x7f547b7be22b] [bt] (5) /home/lvtao/Workspace/mxnet/python/mxnet/../../lib/libmxnet.so(std::_Function_handler<void (std::shared_ptr<dmlc::ManualEvent>), mxnet::engine::ThreadedEnginePerDevice::PushToExecute(mxnet::engine::OprBlock*, bool)::{lambda()#1}::operator()() const::{lambda(std::shared_ptr<dmlc::ManualEvent>)#1}>::_M_invoke(std::_Any_data const&, std::shared_ptr<dmlc::ManualEvent>)+0x4a) [0x7f547b7c2751] [bt] (6) /home/lvtao/Workspace/mxnet/python/mxnet/../../lib/libmxnet.so(std::function<void (std::shared_ptr<dmlc::ManualEvent>)>::operator()(std::shared_ptr<dmlc::ManualEvent>) const+0x5c) [0x7f547b7b6142] [bt] (7) /home/lvtao/Workspace/mxnet/python/mxnet/../../lib/libmxnet.so(void std::_Bind_simple<std::function<void (std::shared_ptr<dmlc::ManualEvent>)> (std::shared_ptr<dmlc::ManualEvent>)>::_M_invoke<0ul>(std::_Index_tuple<0ul>)+0x56) [0x7f547b7b6036] [bt] (8) /home/lvtao/Workspace/mxnet/python/mxnet/../../lib/libmxnet.so(std::_Bind_simple<std::function<void (std::shared_ptr<dmlc::ManualEvent>)> (std::shared_ptr<dmlc::ManualEvent>)>::operator()()+0x1b) [0x7f547b7b5ee1] [bt] (9) /home/lvtao/Workspace/mxnet/python/mxnet/../../lib/libmxnet.so(std::thread::_Impl<std::_Bind_simple<std::function<void (std::shared_ptr<dmlc::ManualEvent>)> (std::shared_ptr<dmlc::ManualEvent>)> >::_M_run()+0x1c) [0x7f547b7b5e30] Aborted (core dumped) ```
---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [email protected] With regards, Apache Git Services
