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;
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII;
 format=flowed
Content-Transfer-Encoding: 7bit
Date:   Fri, 18 Sep 2020 09:20:37 +0100
From:   Marc Zyngier <maz@kernel.org>
To:     Sergey Senozhatsky <sergey.senozhatsky@gmail.com>
Cc:     Leo Yan <leo.yan@linaro.org>,
        Arnaldo Carvalho de Melo <acme@kernel.org>,
        Mark Rutland <mark.rutland@arm.com>,
        Peter Zijlstra <peterz@infradead.org>,
        Will Deacon <will@kernel.org>,
        John Garry <john.garry@huawei.com>,
        Mathieu Poirier <mathieu.poirier@linaro.org>,
        Namhyung Kim <namhyung@kernel.org>,
        Suleiman Souhlal <suleiman@google.com>,
        linux-kernel@vger.kernel.org, linux-arm-kernel@lists.infradead.org
Subject: Re: [PATCHv3] perf kvm: add kvm-stat for arm64
In-Reply-To: <20200918003219.GC3049@jagdpanzerIV.localdomain>
References: <20200917003645.689665-1-sergey.senozhatsky@gmail.com>
 <20200917100950.GC12548@leoy-ThinkPad-X240s>
 <20200917101219.GD12548@leoy-ThinkPad-X240s>
 <652f10660f09bd608b825233713f775a@kernel.org>
 <20200917114231.GE12548@leoy-ThinkPad-X240s>
 <ca309fcda71944455a8c6c1b308886ba@kernel.org>
 <20200918003219.GC3049@jagdpanzerIV.localdomain>
User-Agent: Roundcube Webmail/1.4.8
Message-ID: <f60758eeacb4e94db698d98d6d447939@kernel.org>
Precedence: bulk

On 2020-09-18 01:32, Sergey Senozhatsky wrote:
> On (20/09/17 12:53), Marc Zyngier wrote:
>> Feel free to add a *new* tracepoint instead.
> 
> Wouldn't we want a whole bunch of new tracepoints in this case?

Yes. I don't have a better solution as long as tracepoints are ABI.
Get someone to sign-off on it, and I'll happily change them.

> (almost all of the existing ones with the extra vcpu_id field).
> Right now we have 3 types of events:
> - events with no vcpu at all        // nil
> - events with vcpu_pc               // "0x%016lx", __entry->vcpu_pc
> - events with (void *)vcpu          // "vcpu: %p", __entry->vcpu
> 
> It might be helpful if we could filter out events by vcpu_id.
> But this, basically, doubles the number of events in the ringbuffer.

Only if you enable them both, right? You define new tracepoints that
do whatever you need them to do (hopefully in a cross-architecture
compliant way), and have perf to only use the new ones on arm64.
How would that double the number of events in the buffer?

         M.
-- 
Jazz is not dead. It just smells funny...