On Fri, 9 Aug 2024 16:45:50 +0100 Pavel Begunkov wrote:

quoted

I think this is good, and it doesn't seem hacky to me, because we can
check the page_pools of the netdev while we hold rtnl, so we can be
sure nothing is messing with the pp configuration in the meantime.
Like you say below it does validate the driver rather than rely on the
driver saying it's doing the right thing. I'll look into putting this
in the next version.  

Why not have a flag set by the driver and advertising whether it
supports providers or not, which should be checked for instance in
netdev_rx_queue_restart()? If set, the driver should do the right
thing. That's in addition to a new pp_params flag explicitly telling
if pp should use providers. It's more explicit and feels a little
less hacky.

You mean like I suggested in the previous two emails? :)

Given how easy the check is to implement, I think it's worth
adding as a sanity check. But the flag should be the main API,
if the sanity check starts to be annoying we'll ditch it.