On Sat, Mar 08, 2014 at 09:47:38AM -0800, Greg Kroah-Hartman wrote: > On Sun, Mar 09, 2014 at 01:27:48AM +0800, clanlab.proj wrote: > > Hi Greg, Dave and Felipe, > > > > On Sun, Mar 9, 2014 at 12:45 AM, Greg Kroah-Hartman > > <gre...@linuxfoundation.org> wrote: > > > On Sat, Mar 08, 2014 at 03:39:48PM +0800, clanlab.proj wrote: > > >> Hi Dave and Greg, > > >> > > >> >> Signed-off-by: Weinn Jheng <clanlab.p...@gmail.com> > > >> >> Cc: David Brownell <dbrown...@users.sourceforge.net> > > >> >> Cc: David S. Miller <da...@davemloft.net> > > >> > > > >> > Dave, does this look ok from NAPI point of view ? > > >> > > >> I've found the another patch NAPI version has been taken by Greg. > > > > > > I did? What git commit id is it? > > > > Ha! I found I saw before should be the merge notice. > > > > The patch may be actually taken by Felipe > > commit 808855f0615b2a5c8fd916e4988f3b9e748aac73 (patch) > > tree 7911e8bed2a7b99bbc5950d9cc67724999be7f13 > > "has been applied to my tree and can be found at: > > http://bit.ly/1oxoXwd > > " > > Neither of those commits are in my USB tree, sorry.
here's the correct commit 716fb91dfe1777bd6d5e598f3d3572214b3ed296. NAPI should be the way to go, using a plain workqueue is just wrong, let the networking schedule skbs the way it wants. Also, the throughput difference is only *slightly* better with workqueue. -- balbi
signature.asc
Description: Digital signature
