[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20200918003219.GC3049@jagdpanzerIV.localdomain>
Date: Fri, 18 Sep 2020 09:32:19 +0900
From: Sergey Senozhatsky <sergey.senozhatsky@...il.com>
To: Marc Zyngier <maz@...nel.org>
Cc: Leo Yan <leo.yan@...aro.org>,
Sergey Senozhatsky <sergey.senozhatsky@...il.com>,
Arnaldo Carvalho de Melo <acme@...nel.org>,
Mark Rutland <mark.rutland@....com>,
Peter Zijlstra <peterz@...radead.org>,
Will Deacon <will@...nel.org>,
John Garry <john.garry@...wei.com>,
Mathieu Poirier <mathieu.poirier@...aro.org>,
Namhyung Kim <namhyung@...nel.org>,
Suleiman Souhlal <suleiman@...gle.com>,
linux-kernel@...r.kernel.org, linux-arm-kernel@...ts.infradead.org
Subject: Re: [PATCHv3] perf kvm: add kvm-stat for arm64
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?
(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.
-ss
Powered by blists - more mailing lists