Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Date:   Fri, 10 Sep 2021 21:08:04 +0200
From:   Peter Zijlstra <peterz@infradead.org>
To:     Song Liu <songliubraving@fb.com>
Cc:     "bpf@vger.kernel.org" <bpf@vger.kernel.org>,
        "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
        "acme@kernel.org" <acme@kernel.org>,
        "mingo@redhat.com" <mingo@redhat.com>,
        "kjain@linux.ibm.com" <kjain@linux.ibm.com>,
        Kernel Team <Kernel-team@fb.com>,
        John Fastabend <john.fastabend@gmail.com>,
        Andrii Nakryiko <andrii@kernel.org>
Subject: Re: [PATCH v6 bpf-next 1/3] perf: enable branch record for software
 events
Message-ID: <20210910190804.GS4323@worktop.programming.kicks-ass.net>
References: <20210907202802.3675104-1-songliubraving@fb.com>
 <20210907202802.3675104-2-songliubraving@fb.com>
 <YTs2MpaI7iofckJI@hirez.programming.kicks-ass.net>
 <YTtjeyfJXXiDielu@hirez.programming.kicks-ass.net>
 <96445733-055E-41E3-986B-5E1DC04ADEFA@fb.com>
 <20210910184027.GQ4323@worktop.programming.kicks-ass.net>
 <20210910185003.GC5106@worktop.programming.kicks-ass.net>
 <6830FC62-995A-4282-BD30-76E2506ED993@fb.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <6830FC62-995A-4282-BD30-76E2506ED993@fb.com>
Precedence: bulk

On Fri, Sep 10, 2021 at 07:00:08PM +0000, Song Liu wrote:

> Hmm.. not really. We call migrate_disable() before entering the BPF program. 
> And the helper calls snapshot_branch_stack() inside the BPF program. To move
> it to before migrate_disable(), we will have to add a "whether to snapshot
> branch stack" check before entering the BPF program. This check, while is
> cheap, is added to all BPF programs on this hook, even when the program does 
> not use snapshot at all. So we would rather keep all logic inside the helper, 
> and not touch the common path. 

Moo :/ Because I also really don't want to expose struct rq, it's
currently nicely squirelled away in kernel/sched/ and doesn't get
anywhere near include/.

A well, maybe we can do something clever with migrate_disable() itself.
I'll put it on this endless todo list ;-)