> On 10/12/21 3:04 AM, Dmitry Kozlyuk wrote: > > Data path performance can benefit if the PMD knows which memory it > > will need to handle in advance, before the first mbuf is sent to the PMD. > [...] > > I'd like to understand why the API is internal (not experimental). I think > reasons > should be clear from function description.
My reasoning was that PMDs need this API while applications don't. PMDs may need to deal with any mempools and don't control their creation, while the application knows which mempools it creates and doesn't care about internal mempools PMDs might create. But maybe I was wrong and there are applications that want to use mbufs from those internal mempools for non-DPDK IO, something SPDK-like. I'll add a note about that in the description and make them experimental.
