Incubator-mxnet: Segmentation fault on softmax with length

[eric-haibin-lin]

MXNet version:

pip list | grep mx
mxnet                  1.6.0b20190830

Sample code:

>>> len = mx.nd.array([1,2]); 
>>> x = mx.nd.array([[1,2],[3,4]]); 
>>> mx.nd.softmax(data=x, length=len, use_length=True)

Error message:

Segmentation fault: 11

Stack trace:
  [bt] (0) 1   libmxnet.so                         0x0000000110b51430 mxnet::Storage::Get() + 7968
  [bt] (1) 2   libsystem_platform.dylib            0x00007fffa8e62b3a _sigtramp + 26
  [bt] (2) 3   ???                                 0x75326a9feaf1008c 0x0 + 8444929486347108492
  [bt] (3) 4   libmxnet.so                         0x0000000110d007d6 mxnet::Storage::Get() + 1774278
  [bt] (4) 5   libmxnet.so                         0x000000010e6ab4ee std::__1::__tree<std::__1::__value_type<std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> >, mxnet::NDArrayFunctionReg*>, std::__1::__map_value_compare<std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> >, std::__1::__value_type<std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> >, mxnet::NDArrayFunctionReg*>, std::__1::less<std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> > >, true>, std::__1::allocator<std::__1::__value_type<std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> >, mxnet::NDArrayFunctionReg*> > >::destroy(std::__1::__tree_node<std::__1::__value_type<std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> >, mxnet::NDArrayFunctionReg*>, void*>*) + 1822
  [bt] (5) 6   libmxnet.so                         0x000000010edecba9 mxnet::op::FullyConnectedComputeExCPU(nnvm::NodeAttrs const&, mxnet::OpContext const&, std::__1::vector<mxnet::NDArray, std::__1::allocator<mxnet::NDArray> > const&, std::__1::vector<mxnet::OpReqType, std::__1::allocator<mxnet::OpReqType> > const&, std::__1::vector<mxnet::NDArray, std::__1::allocator<mxnet::NDArray> > const&) + 3755241
  [bt] (6) 7   libmxnet.so                         0x00000001104a70a3 mxnet::imperative::PushFCompute(std::__1::function<void (nnvm::NodeAttrs const&, mxnet::OpContext const&, std::__1::vector<mxnet::TBlob, std::__1::allocator<mxnet::TBlob> > const&, std::__1::vector<mxnet::OpReqType, std::__1::allocator<mxnet::OpReqType> > const&, std::__1::vector<mxnet::TBlob, std::__1::allocator<mxnet::TBlob> > const&)> const&, nnvm::Op const*, nnvm::NodeAttrs const&, mxnet::Context const&, std::__1::vector<mxnet::engine::Var*, std::__1::allocator<mxnet::engine::Var*> > const&, std::__1::vector<mxnet::engine::Var*, std::__1::allocator<mxnet::engine::Var*> > const&, std::__1::vector<mxnet::Resource, std::__1::allocator<mxnet::Resource> > const&, std::__1::vector<mxnet::NDArray*, std::__1::allocator<mxnet::NDArray*> > const&, std::__1::vector<mxnet::NDArray*, std::__1::allocator<mxnet::NDArray*> > const&, std::__1::vector<unsigned int, std::__1::allocator<unsigned int> > const&, std::__1::vector<mxnet::OpReqType, std::__1::allocator<mxnet::OpReqType> > const&)::'lambda'(mxnet::RunContext)::operator()(mxnet::RunContext) const + 739
  [bt] (7) 8   libmxnet.so                         0x00000001104a652d std::__1::__function::__func<mxnet::imperative::PushFCompute(std::__1::function<void (nnvm::NodeAttrs const&, mxnet::OpContext const&, std::__1::vector<mxnet::TBlob, std::__1::allocator<mxnet::TBlob> > const&, std::__1::vector<mxnet::OpReqType, std::__1::allocator<mxnet::OpReqType> > const&, std::__1::vector<mxnet::TBlob, std::__1::allocator<mxnet::TBlob> > const&)> const&, nnvm::Op const*, nnvm::NodeAttrs const&, mxnet::Context const&, std::__1::vector<mxnet::engine::Var*, std::__1::allocator<mxnet::engine::Var*> > const&, std::__1::vector<mxnet::engine::Var*, std::__1::allocator<mxnet::engine::Var*> > const&, std::__1::vector<mxnet::Resource, std::__1::allocator<mxnet::Resource> > const&, std::__1::vector<mxnet::NDArray*, std::__1::allocator<mxnet::NDArray*> > const&, std::__1::vector<mxnet::NDArray*, std::__1::allocator<mxnet::NDArray*> > const&, std::__1::vector<unsigned int, std::__1::allocator<unsigned int> > const&, std::__1::vector<mxnet::OpReqType, std::__1::allocator<mxnet::OpReqType> > const&)::'lambda'(mxnet::RunContext), std::__1::allocator<mxnet::imperative::PushFCompute(std::__1::function<void (nnvm::NodeAttrs const&, mxnet::OpContext const&, std::__1::vector<mxnet::TBlob, std::__1::allocator<mxnet::TBlob> > const&, std::__1::vector<mxnet::OpReqType, std::__1::allocator<mxnet::OpReqType> > const&, std::__1::vector<mxnet::TBlob, std::__1::allocator<mxnet::TBlob> > const&)> const&, nnvm::Op const*, nnvm::NodeAttrs const&, mxnet::Context const&, std::__1::vector<mxnet::engine::Var*, std::__1::allocator<mxnet::engine::Var*> > const&, std::__1::vector<mxnet::engine::Var*, std::__1::allocator<mxnet::engine::Var*> > const&, std::__1::vector<mxnet::Resource, std::__1::allocator<mxnet::Resource> > const&, std::__1::vector<mxnet::NDArray*, std::__1::allocator<mxnet::NDArray*> > const&, std::__1::vector<mxnet::NDArray*, std::__1::allocator<mxnet::NDArray*> > const&, std::__1::vector<unsigned int, std::__1::allocator<unsigned int> > const&, std::__1::vector<mxnet::OpReqType, std::__1::allocator<mxnet::OpReqType> > const&)::'lambda'(mxnet::RunContext)>, void (mxnet::RunContext)>::operator()(mxnet::RunContext&&) + 29
  [bt] (8) 9   libmxnet.so                         0x00000001104213c7 std::__1::enable_if<(__is_forward_iterator<mxnet::NDArray**>::value) && (is_constructible<mxnet::NDArray*, std::__1::iterator_traits<mxnet::NDArray**>::reference>::value), void>::type std::__1::vector<mxnet::NDArray*, std::__1::allocator<mxnet::NDArray*> >::assign<mxnet::NDArray**>(mxnet::NDArray**, mxnet::NDArray**) + 32487

@haojin2

[mxnet-label-bot]

Hey, this is the MXNet Label Bot.
Thank you for submitting the issue! I will try and suggest some labels so that the appropriate MXNet community members can help resolve it.
Here are my recommended label(s): Bug

[eric-haibin-lin]

There is also no documentation for example inputs/outputs.

[ptrendx]

Is the problem here that the length is fp32 and not one of integer types?

[haojin2]

@eric-haibin-lin
This is what I got on my end with your code

>>> import mxnet as mx 
>>> len = mx.nd.array([1,2]); 
>>> x = mx.nd.array([[1,2],[3,4]])
>>> mx.nd.softmax(data=x, length=len, use_length=True)
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/home/ubuntu/incubator-mxnet/python/mxnet/ndarray/ndarray.py", line 234, in __repr__
    return '\n%s\n<%s %s @%s>' % (str(self.asnumpy()),
  File "/home/ubuntu/incubator-mxnet/python/mxnet/ndarray/ndarray.py", line 2504, in asnumpy
    ctypes.c_size_t(data.size)))
  File "/home/ubuntu/incubator-mxnet/python/mxnet/base.py", line 254, in check_call
    raise MXNetError(py_str(_LIB.MXGetLastError()))
mxnet.base.MXNetError: [05:19:12] src/operator/nn/./softmax-inl.h:825: This operation only support integer types, not float32

If I change your code to:

import mxnet as mx
len = mx.nd.array([1,2], dtype='int32'); 
x = mx.nd.array([[1,2],[3,4]]); 
mx.nd.softmax(data=x, length=len, use_length=True)

I get:

[[1.         0.        ]
 [0.26894143 0.7310586 ]]
<NDArray 2x2 @cpu(0)>

Which seems like a correct answer to me.

[eric-haibin-lin]

@haojin2 thanks. I'm not sure why it does not show the intended error msg on my. I was using Mac. Were you also using macOs?
Also - in the doc shall we include an example of using use_len=True, and that the length must be int types?