[MLIR] Canonicalization pattern for eliminating identity ops (#377)

* Canonicalization pattern for eliminating identity ops

* Add a test for the identity elimination rule

* Remove frontend from test

* Use CHECK-NEXT instead of CHECK

src/compiler/dialect/onnx/gen_doc.py

@@ -264,7 +264,7 @@ def  collect_types(schema, input) :
 
 def gen_schema(schema) :
     ShapeInferenceList=['Add', 'MatMul', 'Gemm']
-    CanonicalList=['Add']
+    CanonicalList=['Add', 'Identity']
     line_indent = '  '
 
     #s = 'def ONNX'+schema.name+str(schema.since_version)+'Op:ONNX_Op<"'+schema.name+'", \n'

src/compiler/dialect/onnx/onnxop.inc

@@ -1021,6 +1021,7 @@ def ONNXHardmaxOp:ONNX_Op<"Hardmax",
 
 def ONNXIdentityOp:ONNX_Op<"Identity", 
     [NoSideEffect]> {
+  let hasCanonicalizer = 1;
   let summary = "ONNX Identity operation";
   let description = [{
     "Identity operator"
@@ -2785,11 +2786,13 @@ def ONNXSliceOp:ONNX_Op<"Slice",
     "Slices uses `starts`, `ends`, `axes` and `steps` inputs to specify the start and end"
     "dimension and step for each axis in the list of axes, it uses this information to"
     "slice the input `data` tensor. If a negative value is passed for any of the"
-    "start or end indices, it represent number of elements before the end of that"
+    "start or end indices, it represents number of elements before the end of that"
     "dimension. If the value passed to start or end is larger than the `n` (the"
     "number of elements in this dimension), it represents `n`. For slicing to the"
-    "end of a dimension with unknown size, it is recommended to pass in `INT_MAX`."
-    "If a negative value is passed for step, it represents slicing backward."
+    "end of a dimension with unknown size, it is recommended to pass in `INT_MAX` "
+    "when sclicing forward and 'INT_MIN' when slicing backward."
+    "If a negative value is passed for step, it represents slicing backward. "
+    "However step value cannot be 0."
     "If `axes` are omitted, they are set to `[0, ..., ndim-1]`."
     "If `steps` are omitted, they are set to `[1, ..., 1]` of length `len(starts)`"
     "Example 1:"

src/compiler/pass/onnx_combine.cpp

@@ -28,3 +28,8 @@ void ONNXAddOp::getCanonicalizationPatterns(
     OwningRewritePatternList& results, MLIRContext* context) {
   results.insert<MulAddToGemmOptPattern>(context);
 }
+/// on the ONNXIdentityOp.
+void ONNXIdentityOp::getCanonicalizationPatterns(
+    OwningRewritePatternList& results, MLIRContext* context) {
+  results.insert<IdentityEliminationPattern>(context);
+}

src/compiler/pass/onnx_combine.td

@@ -35,4 +35,8 @@ def MulAddToGemmOptPattern : Pat<(ONNXAddOp (ONNXMatMulOp:$res $m1, $m2), $m3),
                                  (ONNXFullGemmOp $m1, $m2, $m3),
 				 [(HasOneUse $res)]>;
 
+// ONNX_Op (onnx.Identity (%X)) = ONNX_Op (%X)
+def IdentityEliminationPattern : Pat<(ONNXIdentityOp $arg),
+				     (replaceWithValue $arg)>;
+
 #endif // ONNX_COMBINE

test/mlir/onnx/onnx_canonicalization.mlir

@@ -9,3 +9,12 @@ module {
    "std.return"(%1) : (tensor<10x10xf32>) -> ()
  }
 }
+
+func @test_identity_identity(%a0: tensor<10x10xf32>, %a1: tensor<10x10xf32>) -> tensor<10x10xf32> {
+   // CHECK-LABEL: test_identity_identity
+   // CHECK-NEXT: %{{[0-9]+}} = "onnx.Add"(%{{.*}}, %{{.*}}) : (tensor<10x10xf32>, tensor<10x10xf32>) -> tensor<10x10xf32>
+   %0 = "onnx.Identity"(%a0) : (tensor<10x10xf32>) -> tensor<10x10xf32>
+   %1 = "onnx.Identity"(%a1) : (tensor<10x10xf32>) -> tensor<10x10xf32>
+   %2 = "onnx.Add"(%0, %1) : (tensor<10x10xf32>, tensor<10x10xf32>) -> tensor<10x10xf32>
+   "std.return"(%2) : (tensor<10x10xf32>) -> ()
+}