* [ONNX] LRN op

* Review fix pt. 1

* Unsqueeze Fix

* Pooling fix

* set global line ending options for repo

* Normalize line endings

* add newline at eof

* IntelGPU backend. Compile function moved to use switch for operations.

* PR1626. IntelGPU backend. Test unhandled_op fixed.

* PR1626. Comments addressed

* Fix output element type for selection functions

* style

* Use MKLDNN for non-strided slice ops.

* Remove unused variables.

* Add utility function is_strided to graph_util.
Add datatype check for slice in cpu_assignment.

* Create blocked mkldnn memory desc for output when input format is blocked.

* Add slice converLayout fusion to cpu_post_layout_optimizations.

* Rename tensor_view.
Fix a bug.

* Rename a variable.

* Use the Eigen OpenMP thread pool device

* CMake: Work around ancient non-compliant compilers used by the CI system

* Work around the broken CI system by making OpenMP optional

* Add support for Quantized Convolution op via mkldnn for IA backend (codegen + DEX)

* Use call_with_validate

* Style fix

* Fix clang compile errors

* fix Sphinx version for Jenkins and update nG versioning

* Try stable verison of Sphinx 1.7 instead

* postpone adding dropout op to make CI building happy

* sort op list

* use op_tbl

* throw unsupported_op exception when appropriate

* remove dead code

* Add more use of NGRAPH_OP macro to remove boilerplate definitions/implementations

* revert moving class out of namespace

* change from switch to dispatcher map

* [ONNX] Squeeze op

* Review fix pt. 1

* Type change

* [ONNX] Transpose op

* Review fix pt. 1

* Review fix pt. 2

* [ONNX] Slice op

* Review fix pt. 1

* Review fix pt. 2

[ONNX CI] nGraph-ONNX CI Hotfix

* refactor distributed code

* add test file changes

* add distributed class

* update test file

bug fix for nvgpu goe

* pass args instead of pointer to array

* add 3d tiled reshpae

* working version

* add shared mem version of 2d, 3d reshape

* remove unused code

* style

* resolve commits

* add test for 3D reshape, some 3D reshape will be treat as 2D

* Add support for Quantize op via mkldnn for IA backend (codegen + DEX)

* PR feedback addressed

* Use call_with_validate

[ONNX] Unary ops

* add very basic test for Coordinate

* add correct license header

* remove contains_key()

* fix logic error

* wip

* wip

* serializer update to use op_tbl

* fix deserialize

* remove Remainder

* Move to MKLDNN v0.16. Git hash picks up some bug fixes post v0.16

* Temporary workaround for an MKLDNN bug

* reset workspace manager after use

* use unique_ptr for pass memory manager

* Add Reshape passthrough when reshape is logical only.

* Move address conditional to RT in case reference is different but both pointer values are the same.

* Add exclusion to common function collection pass so that tensor name check
is applied during emission for ops that pass inputs through.

* Add propagate_in_place_input/output to handle
propagating inputs and outputs through the non-destructive ops in the graph.

* Remove references to direct execution.

* Remove checks for in-place transpose.

* Formatting.

* elementwise updated

* add support for constructing any order arg list

* add comment for using node_names

* use array instead of vector for args/output

* fix per review comment

* remove dead code