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;
MIME-Version: 1.0
References: <20191108210236.1296047-1-arnd@arndb.de> <20191108211323.1806194-12-arnd@arndb.de>
 <alpine.DEB.2.21.1911132306070.2507@nanos.tec.linutronix.de>
In-Reply-To: <alpine.DEB.2.21.1911132306070.2507@nanos.tec.linutronix.de>
From:   Arnd Bergmann <arnd@arndb.de>
Date:   Thu, 14 Nov 2019 11:51:15 +0100
Message-ID: <CAK8P3a27OV864GfvLK_wjO7dK__r59dZ_dNQACp4G00gJrAwMw@mail.gmail.com>
Subject: Re: [PATCH 21/23] y2038: itimer: change implementation to timespec64
To:     Thomas Gleixner <tglx@linutronix.de>
Cc:     y2038 Mailman List <y2038@lists.linaro.org>,
        Steven Rostedt <rostedt@goodmis.org>,
        Ingo Molnar <mingo@redhat.com>,
        "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
        Anna-Maria Gleixner <anna-maria@linutronix.de>,
        Frederic Weisbecker <frederic@kernel.org>,
        "Peter Zijlstra (Intel)" <peterz@infradead.org>,
        Kees Cook <keescook@chromium.org>
Content-Type: text/plain; charset="UTF-8"
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Wed, Nov 13, 2019 at 11:28 PM Thomas Gleixner <tglx@linutronix.de> wrote:
> On Fri, 8 Nov 2019, Arnd Bergmann wrote:
> > @@ -321,12 +321,12 @@ TRACE_EVENT(itimer_state,
> >               __entry->which          = which;
> >               __entry->expires        = expires;
> >               __entry->value_sec      = value->it_value.tv_sec;
> > -             __entry->value_usec     = value->it_value.tv_usec;
> > +             __entry->value_usec     = value->it_value.tv_nsec / NSEC_PER_USEC;
> >               __entry->interval_sec   = value->it_interval.tv_sec;
> > -             __entry->interval_usec  = value->it_interval.tv_usec;
> > +             __entry->interval_usec  = value->it_interval.tv_nsec / NSEC_PER_USEC;
>
> Hmm, having a division in a tracepoint is clearly suboptimal.

Ok, moving it to the TP_printk() as Steven suggested.

> > -     TP_printk("which=%d expires=%llu it_value=%ld.%ld it_interval=%ld.%ld",
> > +     TP_printk("which=%d expires=%llu it_value=%ld.%06ld it_interval=%ld.%06ld",
>
> We print only 6 digits after the . so that would be even correct w/o a
> division. But it probably does not matter much.

This is just a cosmetic fix, it can be a separate patch if you care. The idea
is to print the numbers as normal decimal representation, e.g. 0.001000
for a millisecond instead of the nonstandard 0.1000.

> > @@ -197,19 +207,13 @@ static void set_cpu_itimer(struct task_struct *tsk, unsigned int clock_id,
> >  #define timeval_valid(t) \
> >       (((t)->tv_sec >= 0) && (((unsigned long) (t)->tv_usec) < USEC_PER_SEC))
>
> Hrm, why do we have yet another incarnation of timeval_valid()?

No idea, you have to ask the author of commit 7d99b7d634d8 ("[PATCH]
Validate and
sanitze itimer timeval from userspace") ;-)

> Can we please have only one (the inline version)?

I'm removing the inline version in a later patch along with most of the rest of
include/linux/time32.h.

Having the macro version is convenient for this patch, since I'm using it
on two different structures (itimerval/__kernel_old_timeval and
old_itimerval32/old_timeval32), neither of which is the type used in the
inline function.

I could use two local inline functions instead of the macro, or just open
code both call sites if you prefer that.

      Arnd