[PATCH] nvme/tcp: Add support to set the tcp worker cpu affinity

[Li Feng]

> 2023年4月18日 上午11:58，Chaitanya Kulkarni <chaitanyak at nvidia.com> 写道：
> 
> 
>> - It is not backed by performance improvements, but more importantly
>> does not cover any potential regressions in key metrics (bw/iops/lat)
>> or lack there of.
>> 
> 
> I've already asked for this, without seeing performance numbers
> and any regression for general usecase it is hard to justify this.
> 
> -ck
> 
> 
Hi ck,
Thanks for your comment.

In the previous mail, just paste the 256k read result(1GB/s to 1.4GB/s), I will add
more io pattern result asap when the machine environment is available.