Hi Jerin, > -----Original Message----- > From: Jerin Jacob <[email protected]> > Sent: Thursday, January 2, 2020 5:52 PM > To: Gavin Hu <[email protected]> > Cc: dpdk-dev <[email protected]>; nd <[email protected]>; David Marchand > <[email protected]>; [email protected]; > [email protected]; [email protected]; [email protected]; > [email protected]; [email protected]; Pavan Nikhilesh > <[email protected]>; Honnappa Nagarahalli > <[email protected]>; Ruifeng Wang > <[email protected]>; Phil Yang <[email protected]>; Joyce Kong > <[email protected]>; Steve Capper <[email protected]> > Subject: Re: [dpdk-dev] [PATCH v2 1/3] eal/arm64: relax the io barrier for > aarch64 > > On Mon, Dec 23, 2019 at 3:46 PM Gavin Hu <[email protected]> wrote: > > > > Hi Jerin, > > > > > -----Original Message----- > > > From: Jerin Jacob <[email protected]> > > > Sent: Monday, December 23, 2019 5:20 PM > > > To: Gavin Hu <[email protected]> > > > Cc: dpdk-dev <[email protected]>; nd <[email protected]>; David Marchand > > > <[email protected]>; [email protected]; > > > [email protected]; [email protected]; > > > [email protected]; [email protected]; [email protected]; > > > Pavan Nikhilesh <[email protected]>; Honnappa Nagarahalli > > > <[email protected]>; Ruifeng Wang > > > <[email protected]>; Phil Yang <[email protected]>; Joyce Kong > > > <[email protected]>; Steve Capper <[email protected]> > > > Subject: Re: [dpdk-dev] [PATCH v2 1/3] eal/arm64: relax the io barrier for > > > aarch64 > > > > > > On Mon, Dec 23, 2019 at 2:44 PM Gavin Hu <[email protected]> wrote: > > > > > > > > Hi Jerin, > > > > > > Hi Gavin, > > > > > > > > > > > I think we are on the same page with regard to the problem, and the > > > situations, thanks for illuminating the historical background of the two > > > barriers. > > > > About the solution, I added inline comments. > > > > > It will be optimization only when if we are changing in the fast path. > > > > > In the slow path, it does not matter. > > > > > I think, the First step should be to use rte_cio_* wherever it is > > > > > coherent memory used in _fast path_. I think, Almost every driver > > > > > fixed that. > > > > > > > > > > I am not against this patch(changing the slow path to use rte_cio* > > > > > from rte_io* and virtio changes associated with that). > > > > > If you are taking that patch, pay attention to all the drivers in the > > > > > tree which is using rte_io* for mixed access in slowpath. > > > > I see 30+ drivers has calling rte_io* directly or indirectly through > > > rte_write/read*. > > > > It is hard for me to figure out all the mixed accesses in these > > > > drivers, and > > > as you said, it makes no sense to change the _slow path_. > > > > > > > > How about we keep the old rte_io as is, and introduce 'fast path' > > > > version > > > of rte_io for new code use? > > > > Then in future, we may merge the two? > > > > Another reason about this proposal is maybe there is rte_io calling in > the > > > fast path, but they are not mixed accesses and rte_cio is not suitable. > > > > > > Could you share more details about the case where fastpath + rte_io > > > needed + rte_cio is not suitable? > > > > Here is an example for i40e, in the fast path, but only a pure io memory > access. > > > https://code.dpdk.org/dpdk/v19.11/source/drivers/net/i40e/i40e_rxtx.c#L12 > 08 > > Yes. That's a performance issue. > > It could be changed to following for the fix that works on x86, arm64 > with existing infra. > > From: > I40E_PCI_REG_WRITE() > > to: > > rte_cio_wmb() > I40E_PCI_REG_WRITE_RELAXED() Yes, this is correct, I will submit a new patch for this. This is an example out of all the cases that I must fix before relaxing the rte_io barriers. My plan is as follows, any comments are welcome! 1. replace rte_*mb and rte_io_*mb with rte_cio_*mb where applicable in the fastpath, this is an optimization, as the barriers are relaxed. 2. replace all the rte_io_*mb with rte_cio_*mb where applicable in the slowpath and control path 3. until *all* the occurrences in the step 1 and 2 are done, then this path can be re-activated.
Please advise if the above approach works from your viewpoint. Maybe I will stop at step 1, step 2 and 3 are not necessary as they are not in the fastpath? > > > > > I wanted two variants of rte_io, because also x86 requires two as indicated > here, one for no-WC and another for WC. > > http://inbox.dpdk.org/dev/20191204151916.12607-1- > [email protected]/T/#ea8bb1b4a378ab09baedbf95b4542bcb92f4a396f > > > > > > > > > > > Any thoughts? > > > > > > > > > > > > > > > But as the case in i40e, we must pay attention to where rte_cio was > > > > > missing but rescued by old rte_io(but not by new rte_io). > > > > > > > > > > > >
