+	    for(j=0,fw=0.0;j<=jx;j++) fw += x[j]*f[jx+i-j];
+	    q[i] = fw;
 	}
 
 	jz = jk;
diff --git a/sysdeps/ieee754/flt-32/k_rem_pio2f.c b/sysdeps/ieee754/flt-32/k_rem_pio2f.c
index e54a067..215b0e0 100644
--- a/sysdeps/ieee754/flt-32/k_rem_pio2f.c
+++ b/sysdeps/ieee754/flt-32/k_rem_pio2f.c
@@ -65,7 +65,8 @@ int __kernel_rem_pio2f(float *x, float *y, int e0, int nx, int prec, const int32
 

  DLTensor* dltensor_in = &dlm_in.dl_tensor;
  dltensor_in->data = data.data();
  dltensor_in->device = {kDLCPU, 0};
  dltensor_in->ndim = static_cast<int32_t>(shape.size());
  dltensor_in->dtype = {kDLFloat, 32, 1};
  dltensor_in->shape = shape.data();
  dltensor_in->strides = strides.data();
  TFE_TensorHandle* handle = TFE_HandleFromDLPack(&dlm_in, status, ctx);
  ASSERT_NE(handle, nullptr)

      immediate_execution_ctx_.reset(ctx_raw);
    }

    // Computing numerical gradients with TensorFloat-32 is numerically
    // unstable. Some forward pass tests also fail with TensorFloat-32 due to
    // low tolerances
    enable_tensor_float_32_execution(false);
  }

  AbstractContextPtr immediate_execution_ctx_;
  GradientRegistry registry_;
  Status status_;

 public:

  };
  tf_gcs_filesystem::RamFileBlockCache cache1(0, 0, 0, fetcher);
  tf_gcs_filesystem::RamFileBlockCache cache2(16, 0, 0, fetcher);
  tf_gcs_filesystem::RamFileBlockCache cache3(0, 32, 0, fetcher);
  tf_gcs_filesystem::RamFileBlockCache cache4(16, 32, 0, fetcher);

  EXPECT_FALSE(cache1.IsCacheEnabled());
  EXPECT_FALSE(cache2.IsCacheEnabled());
  EXPECT_FALSE(cache3.IsCacheEnabled());
  EXPECT_TRUE(cache4.IsCacheEnabled());
}

        "//tensorflow/c/experimental/ops:nn_ops",
        "//tensorflow/core:test",
        "//tensorflow/core:test_main",
        "//tensorflow/core/platform:tensor_float_32_utils",
    ] + if_libtpu(
        if_false = ["//tensorflow/compiler/mlir/tensorflow/c:mlir_c_api_registration"],
        if_true = [],
    ),
)

tf_cuda_cc_test(
    name = "math_grad_test",

      immediate_execution_ctx_.reset(ctx_raw);
    }

    // Computing numerical gradients with TensorFloat-32 is numerically
    // unstable. Some forward pass tests also fail with TensorFloat-32 due to
    // low tolerances
    enable_tensor_float_32_execution(false);
  }

  AbstractContextPtr immediate_execution_ctx_;
  GradientRegistry registry_;
  Status status_;

 public:

      immediate_execution_ctx_.reset(ctx_raw);
    }

    // Computing numerical gradients with TensorFloat-32 is numerically
    // unstable. Some forward pass tests also fail with TensorFloat-32 due to
    // low tolerances
    enable_tensor_float_32_execution(false);
  }

  AbstractContextPtr immediate_execution_ctx_;
  GradientRegistry registry_;
  Status status_;

 public:

      ASSERT_EQ(errors::OK, s.code()) << s.message();
      ctx_.reset(ctx_raw);
    }

    // Computing numerical gradients with TensorFloat-32 is numerically
    // unstable. Some forward pass tests also fail with TensorFloat-32 due to
    // low tolerances
    enable_tensor_float_32_execution(false);
  }

  AbstractContextPtr ctx_;

 public:
  bool UseMlir() const { return strcmp(std::get<0>(GetParam()), "mlir") == 0; }

        case 8:
          *tf_dtype = TF_DataType::TF_UINT8;
          return absl::OkStatus();
        case 16:
          *tf_dtype = TF_DataType::TF_UINT16;
          return absl::OkStatus();
        case 32:
          *tf_dtype = TF_DataType::TF_UINT32;
          return absl::OkStatus();
        case 64:
          *tf_dtype = TF_DataType::TF_UINT64;
          return absl::OkStatus();
        default:

        | Select NextHop).NextHop; \
    $ifIndex = (Get-NetAdapter -InterfaceDescription \"Hyper-V Virtual Ethernet*\" | Sort-Object \
        | Select ifIndex).ifIndex; \
    New-NetRoute -DestinationPrefix 169.254.169.254/32 -InterfaceIndex $ifIndex -NextHop $gateway

# Enable Long Paths for Win32 File/Folder APIs.
RUN New-ItemProperty -Path HKLM:\SYSTEM\CurrentControlSet\Control\FileSystem \