func.func @main_dispatch_18_matmul_2304x144x24_dispatch_0_matmul_2304x144x24_dispatch_0_matmul_2304x144x24() {
  %c8 = arith.constant 8 : index
  %c32 = arith.constant 32 : index
  %c256 = arith.constant 256 : index
  %c72 = arith.constant 72 : index
  %c144 = arith.constant 144 : index
  %c2304 = arith.constant 2304 : index
  %c0 = arith.constant 0 : index
  %c12_i32 = arith.constant 12 : i32
  %c-128_i32 = arith.constant -128 : i32
  %c127_i32 = arith.constant 127 : i32
  %c0_i32 = arith.constant 0 : i32
  %0 = hal.interface.binding.subspan set(0) binding(0) type(storage_buffer) alignment(64) offset(%c0) flags(ReadOnly) : !flow.dispatch.tensor<readonly:tensor<2304x24xi8>>
  %1 = hal.interface.binding.subspan set(0) binding(1) type(storage_buffer) alignment(64) offset(%c0) flags(ReadOnly) : !flow.dispatch.tensor<readonly:tensor<24x144xi8>>
  %2 = hal.interface.binding.subspan set(0) binding(2) type(storage_buffer) alignment(64) offset(%c0) flags(ReadOnly) : !flow.dispatch.tensor<readonly:tensor<144xi32>>
  %3 = hal.interface.binding.subspan set(0) binding(3) type(storage_buffer) alignment(64) offset(%c0) flags(ReadOnly) : !flow.dispatch.tensor<readonly:tensor<144xi32>>
  %4 = hal.interface.binding.subspan set(0) binding(4) type(storage_buffer) alignment(64) offset(%c0) flags(ReadOnly) : !flow.dispatch.tensor<readonly:tensor<144xi32>>
  %5 = hal.interface.binding.subspan set(0) binding(5) type(storage_buffer) alignment(64) offset(%c0) flags(ReadOnly) : !flow.dispatch.tensor<readonly:tensor<144xi8>>
  %6 = hal.interface.binding.subspan set(0) binding(6) type(storage_buffer) alignment(64) offset(%c0) : !flow.dispatch.tensor<writeonly:tensor<2304x144xi8>>
  %workgroup_id_x = hal.interface.workgroup.id[0] : index
  %workgroup_count_x = hal.interface.workgroup.count[0] : index
  %workgroup_id_y = hal.interface.workgroup.id[1] : index
  %workgroup_count_y = hal.interface.workgroup.count[1] : index
  %7 = affine.apply affine_map<()[s0] -> (s0 * 256)>()[%workgroup_id_y]
  %8 = affine.apply affine_map<()[s0] -> (s0 * 256)>()[%workgroup_count_y]
  %9 = affine.apply affine_map<()[s0] -> (s0 * 72)>()[%workgroup_id_x]
  %10 = affine.apply affine_map<()[s0] -> (s0 * 72)>()[%workgroup_count_x]
  scf.for %arg0 = %7 to %c2304 step %8 {
    %11 = flow.dispatch.tensor.load %0, offsets = [%arg0, 0], sizes = [256, 24], strides = [1, 1] : !flow.dispatch.tensor<readonly:tensor<2304x24xi8>> -> tensor<256x24xi8>
    scf.for %arg1 = %9 to %c144 step %10 {
      %12 = flow.dispatch.tensor.load %6, offsets = [%arg0, %arg1], sizes = [256, 72], strides = [1, 1] : !flow.dispatch.tensor<writeonly:tensor<2304x144xi8>> -> tensor<256x72xi8>
      %13 = flow.dispatch.tensor.load %2, offsets = [%arg1], sizes = [72], strides = [1] : !flow.dispatch.tensor<readonly:tensor<144xi32>> -> tensor<72xi32>
      %14 = flow.dispatch.tensor.load %1, offsets = [0, %arg1], sizes = [24, 72], strides = [1, 1] : !flow.dispatch.tensor<readonly:tensor<24x144xi8>> -> tensor<24x72xi8>
      %15 = flow.dispatch.tensor.load %3, offsets = [%arg1], sizes = [72], strides = [1] : !flow.dispatch.tensor<readonly:tensor<144xi32>> -> tensor<72xi32>
      %16 = flow.dispatch.tensor.load %4, offsets = [%arg1], sizes = [72], strides = [1] : !flow.dispatch.tensor<readonly:tensor<144xi32>> -> tensor<72xi32>
      %17 = flow.dispatch.tensor.load %5, offsets = [%arg1], sizes = [72], strides = [1] : !flow.dispatch.tensor<readonly:tensor<144xi8>> -> tensor<72xi8>
      %18 = scf.for %arg2 = %c0 to %c256 step %c8 iter_args(%arg3 = %12) -> (tensor<256x72xi8>) {
        %extracted_slice = tensor.extract_slice %11[%arg2, 0] [8, 24] [1, 1] : tensor<256x24xi8> to tensor<8x24xi8>
        %19 = scf.for %arg4 = %c0 to %c72 step %c32 iter_args(%arg5 = %arg3) -> (tensor<256x72xi8>) {
          %20 = affine.min affine_map<(d0) -> (-d0 + 72, 32)>(%arg4)
          %extracted_slice_0 = tensor.extract_slice %13[%arg4] [%20] [1] : tensor<72xi32> to tensor<?xi32>
          %extracted_slice_1 = tensor.extract_slice %14[0, %arg4] [24, %20] [1, 1] : tensor<24x72xi8> to tensor<24x?xi8>
          %21 = tensor.empty(%20) : tensor<8x?xi32>
          %22 = linalg.fill ins(%c0_i32 : i32) outs(%21 : tensor<8x?xi32>) -> tensor<8x?xi32>
          %23 = linalg.matmul {lowering_config = #iree_codegen.lowering_config<tile_sizes = [[256, 72, 0], [8, 32, 0], [0, 0, 12]]>} ins(%extracted_slice, %extracted_slice_1 : tensor<8x24xi8>, tensor<24x?xi8>) outs(%22 : tensor<8x?xi32>) -> tensor<8x?xi32>
          %extracted_slice_2 = tensor.extract_slice %15[%arg4] [%20] [1] : tensor<72xi32> to tensor<?xi32>
          %extracted_slice_3 = tensor.extract_slice %16[%arg4] [%20] [1] : tensor<72xi32> to tensor<?xi32>
          %extracted_slice_4 = tensor.extract_slice %17[%arg4] [%20] [1] : tensor<72xi8> to tensor<?xi8>
          %extracted_slice_5 = tensor.extract_slice %arg5[%arg2, %arg4] [8, %20] [1, 1] : tensor<256x72xi8> to tensor<8x?xi8>
          %24 = linalg.generic {indexing_maps = [affine_map<(d0, d1) -> (d1)>, affine_map<(d0, d1) -> (d0, d1)>, affine_map<(d0, d1) -> (d1)>, affine_map<(d0, d1) -> (d1)>, affine_map<(d0, d1) -> (d1)>, affine_map<(d0, d1) -> (d0, d1)>], iterator_types = ["parallel", "parallel"]} ins(%extracted_slice_0, %23, %extracted_slice_2, %extracted_slice_3, %extracted_slice_4 : tensor<?xi32>, tensor<8x?xi32>, tensor<?xi32>, tensor<?xi32>, tensor<?xi8>) outs(%extracted_slice_5 : tensor<8x?xi8>) {
          ^bb0(%in: i32, %in_6: i32, %in_7: i32, %in_8: i32, %in_9: i8, %out: i8):
            %25 = arith.muli %in_7, %c12_i32 : i32
            %26 = arith.subi %in_6, %25 : i32
            %27 = arith.addi %in, %26 : i32
            %28 = "tosa.apply_scale"(%27, %in_8, %in_9) {double_round = true} : (i32, i32, i8) -> i32
            %29 = arith.addi %28, %c-128_i32 : i32
            %30 = arith.cmpi slt, %29, %c-128_i32 : i32
            %31 = arith.select %30, %c-128_i32, %29 : i32
            %32 = arith.cmpi sgt, %29, %c127_i32 : i32
            %33 = arith.select %32, %c127_i32, %31 : i32
            %34 = arith.trunci %33 : i32 to i8
            linalg.yield %34 : i8
          } -> tensor<8x?xi8>
          %inserted_slice = tensor.insert_slice %24 into %arg5[%arg2, %arg4] [8, %20] [1, 1] : tensor<8x?xi8> into tensor<256x72xi8>
          scf.yield %inserted_slice : tensor<256x72xi8>
        }
        scf.yield %19 : tensor<256x72xi8>
      }
      flow.dispatch.tensor.store %18, %6, offsets = [%arg0, %arg1], sizes = [256, 72], strides = [1, 1] : tensor<256x72xi8> -> !flow.dispatch.tensor<writeonly:tensor<2304x144xi8>>
    }
  }
  return
}