Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Date:   Fri, 20 Dec 2019 11:10:04 -0600
From:   Dennis Zhou <dennis@kernel.org>
To:     Christopher Lameter <cl@linux.com>
Cc:     Tejun Heo <tj@kernel.org>,
        Jesper Dangaard Brouer <brouer@redhat.com>,
        =?iso-8859-1?Q?Bj=F6rn_T=F6pel?= <bjorn.topel@gmail.com>,
        bpf <bpf@vger.kernel.org>, LKML <linux-kernel@vger.kernel.org>
Subject: Re: Percpu variables, benchmarking, and performance weirdness
Message-ID: <20191220171004.GA8596@dennisz-mbp.dhcp.thefacebook.com>
References: <CAJ+HfNgNAzvdBw7gBJTCDQsne-HnWm90H50zNvXBSp4izbwFTA@mail.gmail.com>
 <20191220103420.6f9304ab@carbon>
 <20191220151239.GE2914998@devbig004.ftw2.facebook.com>
 <alpine.DEB.2.21.1912201536120.16819@www.lameter.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <alpine.DEB.2.21.1912201536120.16819@www.lameter.com>
User-Agent: Mutt/1.12.2 (2019-09-21)
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Fri, Dec 20, 2019 at 03:36:51PM +0000, Christopher Lameter wrote:
> On Fri, 20 Dec 2019, Tejun Heo wrote:
> 
> > On Fri, Dec 20, 2019 at 10:34:20AM +0100, Jesper Dangaard Brouer wrote:
> > > > So, my question to the uarch/percpu folks out there: Why are percpu
> > > > accesses (%gs segment register) more expensive than regular global
> > > > variables in this scenario.
> > >
> > > I'm also VERY interested in knowing the answer to above question!?
> > > (Adding LKML to reach more people)
> >
> > No idea.  One difference is that percpu accesses are through vmap area
> > which is mapped using 4k pages while global variable would be accessed
> > through the fault linear mapping.  Maybe you're getting hit by tlb
> > pressure?

bpf_redirect_info is static so that should be accessed via the linear
mapping as well if we're embedding the first chunk.

> 
> And there are some accesses from remote processors to per cpu ares of
> other cpus. If those are in the same cacheline then those will cause
> additional latencies.
> 

I guess we could pad out certain structs like bpf_redirect_info, but
that isn't really ideal.