Fix coredumps

[leezu]

Fix MXNet trapping the segfault signal and preventing coredumps being generated.

Edit: Comparison of user experience during segfault with and without this PR:

[ins] In [1]: import gc
         ...: import objgraph
         ...: gc.set_debug(gc.DEBUG_SAVEALL)
         ...: import mxnet as mx
         ...: net = mx.gluon.nn.Dense(10, in_units=10)
         ...: net.initialize()
         ...: objgraph.show_refs([net.weight._data[0]], filename='weight_array.png')
         ...: del net
         ...: print(gc.collect())
Graph written to /tmp/objgraph-2ypakr86.dot (28 nodes)
Image generated as weight_array.png
38

[ins] In [2]: gc.garbage
Out[2]:
Segmentation fault: 11

%  

ipython crashed and user is back to shell. With this PR

[ins] In [1]: import gc
         ...: import objgraph
         ...: gc.set_debug(gc.DEBUG_SAVEALL)
         ...: import mxnet as mx
         ...: net = mx.gluon.nn.Dense(10, in_units=10)
         ...: net.initialize()
         ...: objgraph.show_refs([net.weight._data[0]], filename='weight_array.png')
         ...: del net
         ...: print(gc.collect())
Graph written to /tmp/objgraph-b0i9rl27.dot (28 nodes)
Image generated as weight_array.png
38

[ins] In [2]: gc.garbage
Out[2]:
Segmentation fault: 11

zsh: abort      ipython
% echo $?                                                                                                                                                          9s ~/src/mxnet-master/build fixcoredumps
134

After this PR, if the user set ulimit -c unlimited prior to running the segfault inducing program, they will get a coredump for analysis in gdb.

[mxnet-bot]

Hey @leezu , Thanks for submitting the PR
All tests are already queued to run once. If tests fail, you can trigger one or more tests again with the following commands:

• To trigger all jobs: @mxnet-bot run ci [all]
• To trigger specific jobs: @mxnet-bot run ci [job1, job2]

---

CI supported jobs: [miscellaneous, clang, website, unix-gpu, windows-gpu, edge, unix-cpu, centos-gpu, windows-cpu, sanity, centos-cpu]

---

Note:
Only following 3 categories can trigger CI :PR Author, MXNet Committer, Jenkins Admin.
All CI tests must pass before the PR can be merged.

[leezu]

@mxnet-bot run ci [unix-cpu, centos-cpu]

[mxnet-bot]

Jenkins CI successfully triggered : [unix-cpu, centos-cpu]

[leezu]

@mxnet-bot run ci [unix-cpu]

[mxnet-bot]

Jenkins CI successfully triggered : [unix-cpu]

[leezu]

@mxnet-bot run ci [unix-cpu, centos-cpu]

[mxnet-bot]

Jenkins CI successfully triggered : [centos-cpu, unix-cpu]

[leezu]

@mxnet-bot run ci [unix-cpu]

[mxnet-bot]

Jenkins CI successfully triggered : [unix-cpu]

[leezu]

@mxnet-bot run ci [unix-cpu, centos-gpu]

[mxnet-bot]

Jenkins CI successfully triggered : [centos-gpu, unix-cpu]