how can I implement BF16 = BF16 * int9 + int9

[Muzaffer Kal]

Hi all,

This is part of an operation which I would like to implement in RTL so I thought I'd give XLS a try. I looked at the FP32 to int conversion which is reasonable but I'd like to ask your opinion on how I should go about implementing the above expression as I don't want to reimplement all floating point operations necessary. 

Are there any libraries which might be helpful for rounding, normalization etc.
Any suggestions or guidance would be appreciated.

Best

-K