correct the format,and removed some unnecessary changes to the pass.

examples/BuddyMatmul/linalg-transposematmulb-f32.mlir

@@ -72,4 +72,4 @@ func.func @main(){
   call @printMemrefF32(%printed_m5) : (memref<*xf32>) -> ()
 
   return
-}
+}

frontend/Python/graph/graph.py

@@ -164,50 +164,6 @@ def add_node(self, node: Op):
         self._body.append(node)
         self.node_table[node.name] = node
 
-    def check_deletenode(self, node : Op) -> bool:
-        #node : graphnode.name
-        if (not(node.name in self.node_table) ):
-            raise KeyError("node{0} not in graph".format(node.name))
-
-        if (len(node._children)==0):
-            return True
-        return False; 
-
-    def delete_node(self, node: Op,parents : List[Op]):
-        for i in parents:
-            i._children.remove(node.name)
-        node.args.clear()
-        node.kwargs.clear()
-        node._children.clear()
-        self._body.remove(node)
-        self.node_table.pop(node.name)
-
-    def displace_node(self,node: Op,newnode:Op):
-        newnode._arguments = node.args
-        newnode._keyword_arguments = node.kwargs
-        newnode._tensor_meta = node.tensor_meta
-        newnode._op_type = node._op_type
-
-        #deal with users/childrens
-        for i in node._children:
-            newnode.add_children(i)
-        users = [self.node_table[i] for i in node._children]
-        for user in users:
-            user._parents[user._parents.index(node.name)]=newnode.name
-            user.args[user.args.index(node.name)]=newnode.name
-        node._children.clear()
-        #deal with parents+args
-        for i in node._parents:
-            newnode.add_parent(i)
-        parents = [self.node_table[i] for i in node._parents]
-        for parent in parents:
-            parent._children[parent._children.index(node.name)]=newnode.name
-        node._parents.clear()
-        #update node table
-        self._body[self._body.index(node)] = newnode
-        self.node_table.pop(node.name)
-        self.node_table[newnode.name] = newnode
-
     def init_op_group(self):
         """
         Initializes operation groups within the graph.
@@ -223,16 +179,6 @@ def init_op_group(self):
             self.group_map_device[subgraph_name] = DeviceType.UNKNOW
             self.op_groups[subgraph_name] = group
 
-    def check_classicfusetype(self,op:Op):
-        pattern = None
-        if isinstance(op,MatmulOp):
-            parentop = [ self.node_table[str(i)] for i in op._parents]
-            for target in parentop:
-                if (isinstance(target,PermuteOp) ):
-                    pattern = target,parentop,"transpose+mamtmul2D"
-        #TODO : other patterns can be fused
-        return pattern
-
     def fuse_ops(self, pattern_list: List[FunctionType]):
         """
         Fuse operations in the graph based on provided fusion patterns.

frontend/Python/graph/operation.py

@@ -153,10 +153,6 @@ def __init__(self) -> None:
         super().__init__()
         self._op_type = OpType.ReduceType
 
-class transpose_Matmul_fusedOp(Op):
-    def __init__(self) -> None:
-        super().__init__()
-        self._op_type = OpType.ReduceType
 
 class GetItemOp(Op):
     def __init__(self) -> None:

frontend/Python/graph/transform/fuse_ops.py

@@ -28,45 +28,6 @@
 # OP_TYPE_FUSABLE_BY_SPECIFIC_PASS = []
 # ANCHOR_OP_TYPE = [] 
 
-classicfuse_register = {
-    "transpose+mamtmul2D": transpose_Matmul_fusedOp
-}
-
-def classic_fuse(graph : Graph):
-    for op in graph.body:
-        pattern = graph.check_classicfusetype(op)
-        if (pattern):
-            do_classicfusion(graph,op,pattern[0],pattern[1],pattern[2])
-        else:
-            continue
-
-def do_classicfusion(graph : Graph,node,target : Op,parents : List[Op],pattern : str):
-    """
-    Function to fuse some typical operations into one operation.
-    Such as transpose + matmul
-
-    Args:
-    - graph (Graph): The input graph to be simplified.
-
-    Returns:
-    - None: Modifies the input graph in place.
-    """
-
-    fusedop = classicfuse_register.get(pattern)()
-    fusedop.name = "fused"+node.name
-    graph.displace_node(node,fusedop)
-    fusedop.args.pop(fusedop.args.index(target.name))
-    fusedop._parents.pop(fusedop._parents.index(target.name))
-    fusedop.args.extend(target.args)
-    fusedop._parents.extend(target._parents)
-    targets_parent = [graph.node_table[i] for i in target._parents]
-    for i in targets_parent:
-        i.add_children(fusedop.name)
-    target._children.pop(target._children.index(fusedop.name))
-
-    if(graph.check_deletenode(target)):
-        graph.delete_node(target,targets_parent)
-
 def simply_fuse(graph: Graph):
     """
     Function to fuse all operations into one graph.
@@ -79,8 +40,6 @@ def simply_fuse(graph: Graph):
     """
     new_op_group = []
     device = DeviceType.UNKNOW
-    classic_fuse(graph)
-
     for op in graph.body:
         if isinstance(op, PlaceholderOp):
             continue

frontend/Python/ops/linalg.py

@@ -1150,24 +1150,6 @@ def matmul_op(
     op = linalg.matmul(input1, input2, outs=[matmul_result_buffer])
     return op
 
-def transpose_matmul_fused_op(
-    node: transpose_Matmul_fusedOp, 
-    symbol_table:Dict[Tuple[str, int], ir.Operation]
-    ):
-    input1 = symbol_table.get((str(node.args[0]),0))
-    input2 = symbol_table.get((str(node.args[1]),0))
-
-    if input1 is None or input2 is None:
-        return
-    output_shape = list(node.tensor_meta["shape"])
-    dtype = node.tensor_meta["dtype"]
-    mlir_dtype = mlir_element_type_get(dtype)
-    tensor_type = ir.RankedTensorType.get(output_shape, mlir_dtype)
-    element = mlir_element_attr_get(dtype, 0.0)
-    attr = ir.DenseElementsAttr.get_splat(tensor_type, element)
-    result_buffer = arith.ConstantOp(tensor_type, attr).result
-    op = linalg.matmul_transpose_b(input1, input2, outs=[result_buffer])
-    return op
 
 def transpose_op(
     node: TransposeOp,
@@ -1986,7 +1968,6 @@ def gt_op(node: GtOp, symbol_table):
 
 ops_registry = {
     "MatmulOp": matmul_op,
-    "transpose_Matmul_fusedOp": transpose_matmul_fused_op,
     "ArangeOp": arange_op,
     "UnsqueezeOp": unsqueeze_op,
     "ViewOp": view_op,

midend/lib/Conversion/MatMulOptimization/MatMulTransposeBVec.cpp

@@ -54,7 +54,7 @@ class MatMulTransposeBVecPattern : public ConversionPattern{
         Value B = op->getOperand(1);
         Value C = op->getOperand(2);
 
-        // Get shape of input and output
+        // Get shape of input and output.
         ShapedType ATy = A.getType().cast<ShapedType>();
         Type eleTy = ATy.getElementType();
 

tools/buddy-opt/buddy-opt.cpp

@@ -80,7 +80,6 @@ void registerLowerSchePass();
 void registerFuncBufferizeDynamicOffsetPass();
 void registerConvertMemcpyToGPUPass();
 void registerLegalizeShmemOutliningPass();
-void registerMatMul_TransposeB_VecPass();
 void registerMatMulTransposeBVecPass();
 void registerConvertMemcpyToGPUPass();
 void registerLegalizeShmemOutliningPass();