Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
MIME-Version: 1.0
References: <20230802121116.324604-1-zegao@tencent.com> <20230802121116.324604-6-zegao@tencent.com>
 <20230802110701.5227346d@gandalf.local.home>
In-Reply-To: <20230802110701.5227346d@gandalf.local.home>
From:   Ze Gao <zegao2021@gmail.com>
Date:   Thu, 3 Aug 2023 10:38:43 +0800
Message-ID: <CAD8CoPDUkUOc8_p8Ygnoy6xfQZtPhiDDPgnhTMfoEtmrQFRwMw@mail.gmail.com>
Subject: Re: [RFC PATCH v4 5/7] sched, tracing: add to report task state in
 symbolic chars
To:     Steven Rostedt <rostedt@goodmis.org>
Cc:     Adrian Hunter <adrian.hunter@intel.com>,
        Alexander Shishkin <alexander.shishkin@linux.intel.com>,
        Arnaldo Carvalho de Melo <acme@kernel.org>,
        Ian Rogers <irogers@google.com>,
        Ingo Molnar <mingo@redhat.com>, Jiri Olsa <jolsa@kernel.org>,
        Mark Rutland <mark.rutland@arm.com>,
        Masami Hiramatsu <mhiramat@kernel.org>,
        Namhyung Kim <namhyung@kernel.org>,
        Peter Zijlstra <peterz@infradead.org>,
        linux-kernel@vger.kernel.org, linux-perf-users@vger.kernel.org,
        linux-trace-kernel@vger.kernel.org,
        linux-trace-devel@vger.kernel.org, Ze Gao <zegao@tencent.com>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

On Wed, Aug 2, 2023 at 11:07=E2=80=AFPM Steven Rostedt <rostedt@goodmis.org=
> wrote:
>
> On Wed,  2 Aug 2023 08:10:00 -0400
> Ze Gao <zegao2021@gmail.com> wrote:
>
> > From: Ze Gao <zegao2021@gmail.com>
> >
> > Internal representations of task state are likely to be changed
> > or ordered, and reporting them to userspace without exporting
> > them as part of API is basically wrong, which can easily break
> > a userspace observability tool as kernel evolves. For example,
> > perf suffers from this and still reports wrong states as of this
> > writing.
> >
> > OTOH, some masqueraded states like TASK_REPORT_IDLE and
> > TASK_REPORT_MAX are also reported inadvertently, which confuses
> > things even more and most userspace tools do not even take them
> > into consideration.
> >
> > So add a new variable in company with the old raw value to
> > report task state in symbolic chars, which are self-explaining
> > and no further translation is needed. Of course this does not
> > break any userspace tool.
> >
> > Note for PREEMPT_ACTIVE, we introduce 'p' to report it and use
> > the old conventions for the rest.
> >
> > Signed-off-by: Ze Gao <zegao@tencent.com>
> > Reviewed-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
> > Acked-by: Ian Rogers <irogers@google.com>
> > ---
> >  include/trace/events/sched.h | 44 ++++++++++++++++++++++--------------
> >  1 file changed, 27 insertions(+), 17 deletions(-)
> >
> > diff --git a/include/trace/events/sched.h b/include/trace/events/sched.=
h
> > index 7d34db20b2c6..1c7b94793495 100644
> > --- a/include/trace/events/sched.h
> > +++ b/include/trace/events/sched.h
> > @@ -6,6 +6,7 @@
> >  #define _TRACE_SCHED_H
> >
> >  #include <linux/kthread.h>
> > +#include <linux/sched.h>
> >  #include <linux/sched/numa_balancing.h>
> >  #include <linux/tracepoint.h>
> >  #include <linux/binfmts.h>
> > @@ -214,6 +215,27 @@ static inline int __trace_sched_switch_state(bool =
preempt,
> >
> >       return state ? (1 << (state - 1)) : state;
> >  }
> > +
> > +static inline char __trace_sched_switch_state_char(bool preempt,
> > +                                                unsigned int prev_stat=
e,
> > +                                                struct task_struct *p)
> > +{
> > +     long state;
> > +
> > +#ifdef CONFIG_SCHED_DEBUG
> > +     BUG_ON(p !=3D current);
>
> BUG? Why not WARN_ON()?

I directly copied it from __trace_sched_switch_state since they
are very similar. I had doubt on this too but decided to keep it in
case people want to be 100% sure that the current task is exactly
the one that is being switched, otherwise it's a fatal problem for
scheduler at the point where trace_sched_switch is called.

If you think WARN_ON_ONCE is more appropriate, I can fix both
in v6.

Thoughts?

Regards,
Ze