// CHECK:           return %[[VAL_2]] : tensor<10x8xi64>
// CHECK:         }
func.func @convert_remainder_for_int64(%arg0: tensor<10x8xi64>, %arg1: tensor<10x8xi64>) -> tensor<10x8xi64> {
  %0 = mhlo.remainder %arg0, %arg1 : tensor<10x8xi64>
  func.return %0 : tensor<10x8xi64>
}

// CHECK-LABEL:   func @not_convert_remainder_for_int16(
// CHECK-NOT:       "tf.Mod"

    %0 = stablehlo.constant dense<500> : tensor<100xi64>
    %1 = stablehlo.constant dense<0> : tensor<100xi64>
    %2 = stablehlo.compare  LT, %arg2, %1 : (tensor<100xi64>, tensor<100xi64>) -> tensor<100xi1>
    %3 = stablehlo.add %arg2, %0 : tensor<100xi64>
    %4 = stablehlo.select %2, %3, %arg2 : tensor<100xi1>, tensor<100xi64>
    %5 = stablehlo.reshape %4 : (tensor<100xi64>) -> tensor<100x1xi64>
    %6 = "stablehlo.scatter"(%arg0, %5, %arg3) ({

    %arg0: tensor<2048xi64> {tf.device = "/job:localhost/replica:0/task:0/device:CPU:0"},
    %arg1: tensor<2048xi64> {tf.device = "/job:localhost/replica:0/task:0/device:CPU:0"}) -> (tensor<2048xi32>, tensor<2048xi32>) {
    %cst = "tf.Const"() {value = dense<1024> : tensor<i32>} : () -> tensor<i32>
    %0:2 = "tf_device.launch"() ({
	  %1 = "tf.Cast"(%arg0) {Truncate = false} : (tensor<2048xi64>) -> tensor<2048xi32>

   func.func @main(
      %arg0: tensor<2048xi64> {tf.device = "/job:localhost/replica:0/task:0/device:CPU:0"},
      %arg1: tensor<2048xi64> {tf.device = "/job:localhost/replica:0/task:0/device:CPU:0"}) -> tensor<2048xi32> {
      %cst = "tf.Const"() {value = dense<1024> : tensor<i32>} : () -> tensor<i32>
      %0:2 = "tf_device.launch"() ({
            %1 = "tf.Cast"(%arg0) {Truncate = false} : (tensor<2048xi64>) -> tensor<2048xi32>

// CHECK-DAG: %[[cst_1:.*]] = "tf.Const"() <{value = dense<-1> : tensor<1xi64>}> : () -> tensor<1xi64>
// CHECK: %[[arg1_i64:.*]] = "tf.Cast"(%arg1) <{Truncate = false}> : (tensor<1xi32>) -> tensor<1xi64>
// CHECK: %[[tensor_scatter_update:.*]] = "tf.TensorScatterUpdate"(%[[cst]], %[[cst_0]], %[[arg1_i64]]) : (tensor<2xi64>, tensor<1x1xi64>, tensor<1xi64>) -> tensor<2xi64>

  // CHECK: "mhlo.all_reduce"
  // CHECK-SAME: channel_handle = #mhlo.channel_handle<handle = 10001, type = 1>
  // CHECK-SAME{LITERAL}: replica_groups = dense<[[0, 1]]> : tensor<1x2xi64>
  // CHECK: mhlo.add
  // CHECK: mhlo.return
  %0 = "tf.CollectiveReduceV2"(%input, %group_size, %group_key, %instance_key) {merge_op = "Add", final_op = "Id"} : (tensor<f32>, tensor<i32>, tensor<i32>, tensor<i32>) -> tensor<f32>

  func.func @xla_spmd_shard_to_full_shape(%arg0: tensor<1x2xi64>) -> (tensor<2x2xi64>) {
    // CHECK: %[[SHARDING:.*]] = mhlo.custom_call @Sharding(%arg0) {backend_config = "", mhlo.sharding = "{manual}"} : (tensor<1x2xi64>) -> tensor<1x2xi64>
    // CHECK: %[[FULL:.*]] = mhlo.custom_call @SPMDShardToFullShape(%[[SHARDING]]) {backend_config = "", mhlo.sharding = "{devices=[2,1]0,1}"} : (tensor<1x2xi64>) -> tensor<2x2xi64>
    // CHECK: return %[[FULL]]

  %padding = "tf.Const"() { value = dense<[[1, 2]]> : tensor<1x2xi64> } : () -> tensor<1x2xi64>
  // CHECK-DAG: [[PAD:%.+]] = "tf.Const"() {{.+}} -> tensor<1x2xi64>
  // CHECK-DAG: [[CST:%.+]] = "tf.Const"() <{value = dense<0.000000e+00> : tensor<f32>}>
  // CHECK: "tf.PadV2"(%arg0, [[PAD]], [[CST]])
  %0 = "tf.Pad"(%arg0, %padding) : (tensor<3xf32>, tensor<1x2xi64>) -> tensor<6xf32>
  func.return %0 : tensor<6xf32>
}

  %padding = "tf.Const"() { value = dense<[[1, 2]]> : tensor<1x2xi64> } : () -> tensor<1x2xi64>
  // CHECK: "mhlo.pad"(%arg0, %arg1) <{
  // CHECK-SAME: edge_padding_high = dense<2> : tensor<1xi64>,
  // CHECK-SAME: edge_padding_low = dense<1> : tensor<1xi64>,
  // CHECK-SAME: interior_padding = dense<0> : tensor<1xi64>
  %1 = "tf.PadV2"(%arg0, %padding, %arg1) : (tensor<3xf32>, tensor<1x2xi64>, tensor<f32>) -> tensor<6xf32>

    // CHECK:             %[[A_SHARD:.+]] = "tf.XlaSpmdFullToShardShape"(%[[A]]) <{dim = -1 : i64, manual_sharding = "\08\03\1A\02\02\01\22\02\00\01", unspecified_dims = []}> : (tensor<2x2xi64>) -> tensor<1x2xi64>
    // CHECK:             %[[B:.+]] = "tf._XlaHostComputeMlir"(%[[A_SHARD]])
    // CHECK-SAME:          manual_sharding = true
    // CHECK-SAME:          recv_key = "host_compute_channel_0_retvals"