On Mon, Nov 04, 2024 at 11:43:17AM -0700, Jonathan Corbet wrote: > Joe Damato <jdamato@xxxxxxxxxx> writes: > > > On Mon, Nov 04, 2024 at 05:52:52PM +0700, Bagas Sanjaya wrote: > >> On Sun, Nov 03, 2024 at 05:24:09AM +0000, Joe Damato wrote: > >> > +It is important to note that choosing a large value for ``gro_flush_timeout`` > >> > +will defer IRQs to allow for better batch processing, but will induce latency > >> > +when the system is not fully loaded. Choosing a small value for > >> > +``gro_flush_timeout`` can cause interference of the user application which is > >> > +attempting to busy poll by device IRQs and softirq processing. This value > >> > +should be chosen carefully with these tradeoffs in mind. epoll-based busy > >> > +polling applications may be able to mitigate how much user processing happens > >> > +by choosing an appropriate value for ``maxevents``. > >> > + > >> > +Users may want to consider an alternate approach, IRQ suspension, to help deal > >> to help dealing > >> > +with these tradeoffs. > >> > + > > > > Thanks for the careful review. I read this sentence a few times and > > perhaps my English grammar isn't great, but I think it should be > > one of: > > > > Users may want to consider an alternate approach, IRQ suspension, to > > help deal with these tradeoffs. (the original) > > The original is just fine here. Bagas, *please* do not bother our > contributors with this kind of stuff, it does not help. Thanks for the feedback. I had been preparing a v6 based on Bagas' comments below where you snipped about in the documentation, etc. Should I continue to prepare a v6? It would only contain documentation changes in this patch; I can't really tell if a v6 is necessary or not.
