Upgrade MKL-DNN dependency to v1.0

[TaoLv]

Description

This is an effort from the whole Intel MXNet team. Credits belong to everyone in the team.

This PR upgrades the 3rdparty/mkldnn dependency to it's v1.0.x release which has many API breaking changes and is not backward compatible to the previous v0.x versions. Hence this PR also changes the integration code in operators while keep the integration methodology and software architecture the same. Development has been done on the feature branch mkldnn-v1.0 and tracked via github project: https://github.com/apache/incubator-mxnet/projects/16.

Please see the discussion here: https://lists.apache.org/thread.html/f46ab920f18795496eafe713e6e9e561c684e06189085cec17b401dc@%3Cdev.mxnet.apache.org%3E
Please see MKL-DNN v1.0 RFC here: https://github.com/intel/mkl-dnn/tree/rfc-api-changes-v1.0/doc/rfc/api-v1.0

As mentioned in the dev@ thread, this PR also removes the mklml and iomp5 library which previously are distributed along with MXNet pip package. So it also fixes the license issue in #15544. It's a requirement for 1.6.0 release.

Checklist

Essentials

Please feel free to remove inapplicable items for your PR.

• The PR title starts with [MXNET-$JIRA_ID], where $JIRA_ID refers to the relevant JIRA issue created (except PRs with tiny changes)
• Changes are complete (i.e. I finished coding on this PR)
• All changes have test coverage:
• Unit tests are added for small changes to verify correctness (e.g. adding a new operator)
• Nightly tests are added for complicated/long-running ones (e.g. changing distributed kvstore)
• Build tests will be added for build configuration changes (e.g. adding a new build option with NCCL)
• Code is well-documented:
• For user-facing API changes, API doc string has been updated.
• For new C++ functions in header files, their functionalities and arguments are documented.
• For new examples, README.md is added to explain the what the example does, the source of the dataset, expected performance on test set and reference to the original paper if applicable
• Check the API doc at https://mxnet-ci-doc.s3-accelerate.dualstack.amazonaws.com/PR-$PR_ID/$BUILD_ID/index.html
• To the my best knowledge, examples are either not affected by this change, or have been fixed to be compatible with this change

Changes

• Feature1, tests, (and when applicable, API doc)
• Feature2, tests, (and when applicable, API doc)

Comments

• If this change is a backward incompatible change, why must this change be made.
• Interesting edge cases to note here

[pengzhao-intel]

RNN has inconsistent output in training mode. The following script prints a nonzero array:

import mxnet as mx

sym = mx.sym.RNN(
	mx.symbol.Variable('input'),
	mx.symbol.Variable('weights'),
	mx.symbol.Variable('state'),
	state_size = 4,
	num_layers = 1,
	mode = 'gru'
)
ex = sym.bind(mx.cpu(),
	{
		"input": mx.ndarray.random.uniform(low = 0, high = 1, shape = (3, 1, 2)),
		"weights": mx.ndarray.random.uniform(low = 0, high = 1, shape = (96)),
		"state": mx.nd.zeros([1, 1, 4]),
	}
)

ex.forward(is_train=True)
res1 = ex.outputs[0].asnumpy()
ex.forward(is_train=True)
res2 = ex.outputs[0].asnumpy()

print(res1 - res2)

This also happens if MKL-DNN is disabled by env var, but it doesn't happen on master (built with MKL-DNN)

@zixuanweeei please take a look for this issue

[zixuanweeei]

deeply appreciative of your work @matteosal. I can reproduce the inconsistency with "gru" mode. "lstm" and "rnn_*" could give consistent results. I will look into this. Thanks again. @pengzhao-intel

[zixuanweeei]

@matteosal #16663 could fix the inconsistent of GRU.

[pengzhao-intel]

@ALL, only one issue is left at now and we will fix it soon. I prefer to merge this PR in one or two days so that we have buffer time to test by the master or nightly build.

Please take a final review and feel free to raise the issue or concern :)

[pengzhao-intel]

The only known issue is fixed by #16672

[pengzhao-intel]

LGTM

[pengzhao-intel]

CI failure is not related to the change. Please retrigger the CI.

[TaoLv]

@samskalicky @apeforest @marcoabreu @mseth10 I hope your comments have been addressed. If so, could you please approve?

[mseth10]

LGTM

[samskalicky]

LGTM thanks for all of your hard work!

[pengzhao-intel]

Thanks all of the efforts from the team :)
Merging now.