Casting floating point numbers fails for Decimal64 but works for other variants

[AdamGS]

Describe the bug
Casting a floating point number like 1.1_f64 to decimal64 fails.

To Reproduce
The following test fails, but only for Decimal64.

    #[test]
    fn test_cast_floating_to_decimals() {
        for output_type in [
            DataType::Decimal32(9, 3),
            DataType::Decimal64(9, 3),
            DataType::Decimal128(9, 3),
            DataType::Decimal256(9, 3),
        ] {
            let input_type = DataType::Float64;
            assert!(can_cast_types(&input_type, &output_type));

            let array = vec![Some(1.1_f64)];
            let array = PrimitiveArray::<Float64Type>::from_iter(array);
            let result = cast_with_options(
                &array,
                &output_type,
                &CastOptions {
                    safe: false,
                    format_options: FormatOptions::default(),
                },
            );
            assert!(
                result.is_ok(),
                "Failed to cast to {output_type} with: {}",
                result.unwrap_err().to_string()
            );
        }
    }

Expected behavior
This cast operation should work for all decimal types.

Additional context

[AdamGS]

take