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:   Wed, 18 Sep 2019 07:32:37 -0700
From:   Alexei Starovoitov <alexei.starovoitov@gmail.com>
To:     Yonghong Song <yhs@fb.com>
Cc:     "jinshan.xiong@gmail.com" <jinshan.xiong@gmail.com>,
        "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
        "rostedt@goodmis.org" <rostedt@goodmis.org>,
        "mingo@redhat.com" <mingo@redhat.com>,
        "jinshan.xiong@uber.com" <jinshan.xiong@uber.com>,
        "bpf@vger.kernel.org" <bpf@vger.kernel.org>
Subject: Re: [PATCH] staging: tracing/kprobe: filter kprobe based perf event
Message-ID: <20190918143235.kpclo45eo7qye7fs@ast-mbp.dhcp.thefacebook.com>
References: <20190918052406.21385-1-jinshan.xiong@gmail.com>
 <5302836c-a6a1-c160-2de2-6a5b3d2c4828@fb.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <5302836c-a6a1-c160-2de2-6a5b3d2c4828@fb.com>
User-Agent: NeoMutt/20180223
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Wed, Sep 18, 2019 at 05:51:10AM +0000, Yonghong Song wrote:
> 
> Adding cc to bpf@vger.kernel.org mailing list since this is really
> bpf related.
> 
> On 9/17/19 10:24 PM, jinshan.xiong@gmail.com wrote:
> > From: Jinshan Xiong <jinshan.xiong@gmail.com>
> > 
> > Invoking bpf program only if kprobe based perf_event has been added into
> > the percpu list. This is essential to make event tracing for cgroup to work
> > properly.
> 
> The issue actually exists for bpf programs with kprobe, uprobe, 
> tracepoint and trace_syscall_enter/exit.
> 
> In all these places, bpf program is called regardless of
> whether there are perf events or not on this cpu.
> This provides bpf programs more opportunities to see
> the events. I guess this is by design.
> Alexei/Daniel, could you clarify?

Yes. It is by design.
When bpf is attached to kprobe it will fire on all cpus.
Per-cpu or per-task or per-cgroup filtering is already done
inside bpf programs.
We cannot make this change now it will break all users.