* [MLIR] Bump MLIR repo to commit 59167c2, 07/25/2019.

MLIR commit:
Author: River Riddle <riverriddle@google.com>
Date:   Wed Jul 24 16:41:11 2019 -0700

    NFC: Use ValueOfRange instead of T in Diagnostic::appendRange.

        For iterator_range, T is often the name of another iterator type
        and not the the value of the range.

LLVM commit:
Author: Marshall Clow <mclow.lists@gmail.com>
Date:   Thu Jul 25 03:26:05 2019 +0000

    Implement change #4 of P1466: Change weekday to accept both 0 and 7
    as Sunday. Add accessors 'c_encoding' and 'iso_encoding' to provide
    different interpretations of the weekday. Remove 'operator unsigned'

* style

* Move MLIR/LLVM repos a bit more forward

* [MLIR] Enable affine dialect loop fusion

Enable affine dialect loop fusion in nGraph pipeline. It also adds an
opt flag to enable/disable it when ngraph-opt is in place. Fusion seems
to work for simple cases. It wasn't able to fuse dot + add, though, at
least in my test case. One example that worked:

Input:
  %6 = alloc() : memref<2500x2500xf32>
  affine.for %i3 = 0 to 2500 {
    affine.for %i4 = 0 to 2500 {
      %7 = load %arg0[%i3, %i4] : memref<2500x2500xf32>
      %8 = load %0[%i3, %i4] : memref<2500x2500xf32>
      %9 = addf %8, %7 : f32
      store %9, %6[%i3, %i4] : memref<2500x2500xf32>
    }
  }
  %10 = alloc() : memref<2500x2500xf32>
  affine.for %i5 = 0 to 2500 {
    affine.for %i6 = 0 to 2500 {
      %11 = load %arg2[%i5, %i6] : memref<2500x2500xf32>
      %12 = load %0[%i5, %i6] : memref<2500x2500xf32>
      %13 = addf %12, %11 : f32
      store %13, %10[%i5, %i6] : memref<2500x2500xf32>
    }
  }
  %14 = alloc() : memref<2500x2500xf32>
  affine.for %i7 = 0 to 2500 {
    affine.for %i8 = 0 to 2500 {
      %15 = load %10[%i7, %i8] : memref<2500x2500xf32>
      %16 = load %6[%i7, %i8] : memref<2500x2500xf32>
      %17 = addf %16, %15 : f32
      store %17, %14[%i7, %i8] : memref<2500x2500xf32>
    }
  }

Output:
  %8 = alloc() : memref<2500x2500xf32>
  affine.for %i3 = 0 to 2500 {
    affine.for %i4 = 0 to 2500 {
      %9 = load %arg2[%i3, %i4] : memref<2500x2500xf32>
      %10 = load %2[%i3, %i4] : memref<2500x2500xf32>
      %11 = addf %10, %9 : f32
      %12 = affine.apply #map2(%i3, %i4, %i3, %i4)
      %13 = affine.apply #map3(%i3, %i4, %i3, %i4)
      store %11, %0[%12, %13] : memref<1x1xf32>
      %14 = load %arg0[%i3, %i4] : memref<2500x2500xf32>
      %15 = load %2[%i3, %i4] : memref<2500x2500xf32>
      %16 = addf %15, %14 : f32
      %17 = affine.apply #map2(%i3, %i4, %i3, %i4)
      %18 = affine.apply #map3(%i3, %i4, %i3, %i4)
      store %16, %1[%17, %18] : memref<1x1xf32>
      %19 = affine.apply #map2(%i3, %i4, %i3, %i4)
      %20 = affine.apply #map3(%i3, %i4, %i3, %i4)
      %21 = load %0[%19, %20] : memref<1x1xf32>
      %22 = affine.apply #map2(%i3, %i4, %i3, %i4)
      %23 = affine.apply #map3(%i3, %i4, %i3, %i4)
      %24 = load %1[%22, %23] : memref<1x1xf32>
      %25 = addf %24, %21 : f32
      store %25, %8[%i3, %i4] : memref<2500x2500xf32>
    }
  }

* Rename MLIR_LLVM_OPTIONS to NGRAPH_MLIR_OPTIONS

Something like this works now:
NGRAPH_MLIR_OPTIONS="--enable-affine-loop-fusion=false"

* Disable loop fusion by default and fix typo

* [MLIR] Fix naming convention in MLIR files

Add naming convention note per file to state which files should use
nGraph naming convention and which MLIR naming convention and align
naming convention in those files with such a note.

* Remove m-prefix

* WIP

* All but two unit tests passing

* Explanatory comment

* Cleanup

* A bit of cleanup stemming from review comments

* Rewrite to use LoopNestBuilder

* Remove unnecessary check from CompiledKernel

* Removed some dead-ish code I missed

* Switch to camelCase in lowerer.cpp

* Fix assignment of resIndexHandles that was triggering an assert

* Add in some safety checks

* dyn_cast -> cast

* - templatize computing binary elementwise
- added lowering support for Add, Sub, Multiply, Divide

* - Added Support for Greater and less Op

* -Add support for Minimum and Maximum

* use edsc::intrinsics::select instead of terenary operator

* Addressed PR comments

* - return after the conditional check

* Support MLIR lowering of Relu

* Use EDSC comparison

* style-apply

* Use .inc file for Conversion classes list

* Disable i32 Relu for plaidml

Improvements and changes in dialect conversion APIs. Type conversion
and op rewriting are now a bit more independent mechanisms. Dialect
conversion now requires a target object with the legal transformations
to be applied. Legality checks to detect illegal operations after
dialect conversion have been strengthened significantly.

MLIR LLVM commit:
commit 82d5084cf28bca7a0eebec283cdeae118dfa1d11 (HEAD)
Author: Jacques Pienaar <jpienaar@google.com>
Date:   Wed Jun 26 11:12:40 2019 -0700

    Split out TranslateClParser and add new parse method that reuses
    SourceMgr.

MLIR commit:
commit 82d5084cf28bca7a0eebec283cdeae118dfa1d11 (HEAD)
Author: Jacques Pienaar <jpienaar@google.com>
Date:   Wed Jun 26 11:12:40 2019 -0700

    Split out TranslateClParser and add new parse method that reuses
    SourceMgr.

Add support for any rank ArgMin. Hacky workaround for IndexType tensors by doing type-conversion after operation

It refactor MLIR dump code in MLIRCompiler and adds missing dumps for
some steps of the lowering. It also removes some redundant dump for
affine dialect.

MLIR Compile only supported ops

It introduces end-to-end support for sub-graphs with multiple nodes. It
adds a test with a dot+add ops.
Co-authored-by: Diego Caballero <diego.caballero@intel.com>
Co-authored-by: Nagy Mostafa <nagy.mostafa@gmail.com>

It extends dot definition to be able to deal with operands that are not
square tensors. It also fixes a bug in the lowerer related to that.