Content-Type: text/plain; charset="utf-8"
MIME-Version: 1.0
Content-Transfer-Encoding: quoted-printable
In-Reply-To: <20211209163720.106185-3-nikita@trvn.ru>
References: <20211209163720.106185-1-nikita@trvn.ru> <20211209163720.106185-3-nikita@trvn.ru>
Subject: Re: [PATCH 2/4] clk: qcom: clk-rcg2: Make sure to not write d=0 to the NMD register
From:   Stephen Boyd <sboyd@kernel.org>
Cc:     bjorn.andersson@linaro.org, agross@kernel.org, tdas@codeaurora.org,
        joonwoop@codeaurora.org, svarbanov@mm-sol.com,
        linux-arm-msm@vger.kernel.org, linux-clk@vger.kernel.org,
        linux-gpio@vger.kernel.org, linux-kernel@vger.kernel.org,
        ~postmarketos/upstreaming@lists.sr.ht,
        Nikita Travkin <nikita@trvn.ru>
To:     Nikita Travkin <nikita@trvn.ru>, linus.walleij@linaro.org,
        mturquette@baylibre.com
Date:   Fri, 07 Jan 2022 16:56:02 -0800
User-Agent: alot/0.9.1
Message-Id: <20220108005604.05575C36AEB@smtp.kernel.org>
Precedence: bulk

Quoting Nikita Travkin (2021-12-09 08:37:18)
> Sometimes calculation of d value may result in 0 because of the
> rounding after integer division. This causes the following error:
>=20
> [  113.969689] camss_gp1_clk_src: rcg didn't update its configuration.
> [  113.969754] WARNING: CPU: 3 PID: 35 at drivers/clk/qcom/clk-rcg2.c:122=
 update_config+0xc8/0xdc
>=20
> Make sure that D value is never zero.
>=20
> Fixes: 7f891faf596e ("clk: qcom: clk-rcg2: Add support for duty-cycle for=
 RCG")
> Signed-off-by: Nikita Travkin <nikita@trvn.ru>
> ---
>  drivers/clk/qcom/clk-rcg2.c | 4 ++++
>  1 file changed, 4 insertions(+)
>=20
> diff --git a/drivers/clk/qcom/clk-rcg2.c b/drivers/clk/qcom/clk-rcg2.c
> index 6964cf914b60..fdfd43e2a01b 100644
> --- a/drivers/clk/qcom/clk-rcg2.c
> +++ b/drivers/clk/qcom/clk-rcg2.c
> @@ -424,6 +424,10 @@ static int clk_rcg2_set_duty_cycle(struct clk_hw *hw=
, struct clk_duty *duty)
>         if (d > mask)
>                 d =3D mask;
> =20
> +       /* Hardware can't handle d=3D0, make sure it's at least 1 */
> +       if (!d)
> +               d =3D 1;

Maybe clamp() would be better

	/*
 	 * Check bit widths of 2d. If D is too big reduce duty cycle and
	 * ensure it is non-zero.
	 */
	clamp(d, 1, mask);

> +
>         if ((d / 2) > (n - m))
>                 d =3D (n - m) * 2;
>         else if ((d / 2) < (m / 2))