Skip site navigation (1)Skip section navigation (2)
Date:      Sun, 29 Mar 2015 01:16:21 +0300
From:      Slawa Olhovchenkov <slw@zxy.spb.ru>
To:        Adrian Chadd <adrian@freebsd.org>
Cc:        "freebsd-hackers@freebsd.org" <freebsd-hackers@freebsd.org>
Subject:   Re: irq cpu binding
Message-ID:  <20150328221621.GG23643@zxy.spb.ru>
In-Reply-To: <CAJ-Vmo=wecgoVYcS14gsOnT86p=HEMdao65aXTi7jLfVVyOELg@mail.gmail.com>
References:  <20150328181026.GB23643@zxy.spb.ru> <CAJ-Vmok=0=C%2BHLvLXEM5pbYp_zP6C498wudAfEA40s4AVpGa_w@mail.gmail.com> <20150328183147.GC23643@zxy.spb.ru> <CAJ-VmokPdnOSASLcrzHtqs_r6GfUW6Q5ap7r9Tnp1OvKhDh20A@mail.gmail.com> <20150328192505.GD23643@zxy.spb.ru> <CAJ-Vmo=R5J=LT9ntgZJ4xHZwn8_WEAUC=3wSqPKn3=z8ki_N6A@mail.gmail.com> <20150328194959.GE23643@zxy.spb.ru> <CAJ-Vmo=1rzB%2BYNGNuV9s=FnSse7FL7S42OSS4u-PzUs74b850A@mail.gmail.com> <20150328201219.GF23643@zxy.spb.ru> <CAJ-Vmo=wecgoVYcS14gsOnT86p=HEMdao65aXTi7jLfVVyOELg@mail.gmail.com>

next in thread | previous in thread | raw e-mail | index | archive | help
On Sat, Mar 28, 2015 at 02:52:49PM -0700, Adrian Chadd wrote:

> On 28 March 2015 at 13:12, Slawa Olhovchenkov <slw@zxy.spb.ru> wrote:
> > On Sat, Mar 28, 2015 at 12:54:05PM -0700, Adrian Chadd wrote:
> >
> >> This is why I started fleshing out more of the RSS stuff - the
> >> defaults aren't .. all that useful.
> >
> > Do you review implementing RSS as accpet_filter?
> 
> If someone has a hack to do that, sure. Thing is, I'm not all that
> interested in trying to support a hack to make things transparent as
> there are other things to get "right" for multicore / NUMA
> performance, and those already require much cpu pinning and such.
> 
> I'd rather finish off what I've started - expose this stuff via some
> freebsd libraries to make application developers lives easier, and not
> try to do more hacks. :)

accpet_filter allow use RSS in nginx w/o any modification, I think.

> >> ixgbe should be calling bus_bind_intr().
> >
> > And how to re-bind this later?..
> 
> That's a question for john. I thought it was being rebound when you
> reset the irq interrupt via cpuset, but maybe I'm wrong.

Do you forward this to john?



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20150328221621.GG23643>