Received-SPF: pass (google.com: domain of linux-kernel+bounces-170720-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:4601:e00::3 as permitted sender) client-ip=2604:1380:4601:e00::3;
Precedence: bulk
MIME-Version: 1.0
References: <20240506-b4-sio-ntp-c-v1-1-a01281aa01ba@google.com>
In-Reply-To: <20240506-b4-sio-ntp-c-v1-1-a01281aa01ba@google.com>
From: John Stultz <jstultz@google.com>
Date: Mon, 6 May 2024 23:02:17 -0700
Message-ID: <CANDhNCps8qE1a+Vp2HNgU=-vA5RwhrLd2_D0EMZiNv1j_OOYWA@mail.gmail.com>
Subject: Re: [PATCH] ntp: safeguard against time_constant overflow case
To: Justin Stitt <justinstitt@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>, Stephen Boyd <sboyd@kernel.org>, 
	Nathan Chancellor <nathan@kernel.org>, Bill Wendling <morbo@google.com>, linux-kernel@vger.kernel.org, 
	llvm@lists.linux.dev
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

On Mon, May 6, 2024 at 3:01=E2=80=AFPM Justin Stitt <justinstitt@google.com=
> wrote:
>
> Nonetheless, let's slightly rework the logic surrounding time_constant
> and how it is incremented such that we avoid unintentional wrap-around
> (even though it is extremely unlikely to be hit in non-fuzzing scenarios)=
.
>
> [1]: https://github.com/llvm/llvm-project/pull/82432
>
> Signed-off-by: Justin Stitt <justinstitt@google.com>
> ---
>  kernel/time/ntp.c | 11 +++++++----
>  1 file changed, 7 insertions(+), 4 deletions(-)
>
> diff --git a/kernel/time/ntp.c b/kernel/time/ntp.c
> index 406dccb79c2b..a9f039601968 100644
> --- a/kernel/time/ntp.c
> +++ b/kernel/time/ntp.c
> @@ -65,6 +65,9 @@ static s64                    time_offset;
>  /* pll time constant:                                                  *=
/
>  static long                    time_constant =3D 2;
>
> +/* pll time constant increment:                                         =
       */
> +static long                    time_constant_inc =3D 4;
> +

I'd probably use a `#define TIME_CONSTANT_INC 4` for this.

>  /* maximum error (usecs):                                              *=
/
>  static long                    time_maxerror =3D NTP_PHASE_LIMIT;
>
> @@ -734,10 +737,10 @@ static inline void process_adjtimex_modes(const str=
uct __kernel_timex *txc,
>
>         if (txc->modes & ADJ_TIMECONST) {
>                 time_constant =3D txc->constant;
> -               if (!(time_status & STA_NANO))
> -                       time_constant +=3D 4;
> -               time_constant =3D min(time_constant, (long)MAXTC);
> -               time_constant =3D max(time_constant, 0l);
> +               if (!(time_status & STA_NANO) &&
> +                   unlikely(LONG_MAX - time_constant_inc >=3D time_const=
ant))
> +                       time_constant +=3D time_constant_inc;
> +               time_constant =3D clamp_t(long, time_constant, 0, MAXTC);
>         }

Overall, this looks fine. Though the time_status conditional is now a
little unwieldy.

I wonder if some sort of a helper like:
      time_constant =3D safe_add(time_constant, TIME_CONSTANT_INC, LONG_MAX=
);

Might make this a little easier to read?

thanks
-john