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;
Message-ID: <0dc457cbd13ea76a3aa3c70b2a31a537.sboyd@kernel.org>
Content-Type: text/plain; charset="utf-8"
MIME-Version: 1.0
Content-Transfer-Encoding: quoted-printable
In-Reply-To: <20230419133013.2563-2-quic_tdas@quicinc.com>
References: <20230419133013.2563-1-quic_tdas@quicinc.com> <20230419133013.2563-2-quic_tdas@quicinc.com>
Subject: Re: [PATCH 1/4] clk: qcom: branch: Extend the invert logic for branch2 clocks
From:   Stephen Boyd <sboyd@kernel.org>
Cc:     quic_skakitap@quicinc.com, Imran Shaik <quic_imrashai@quicinc.com>,
        linux-arm-msm@vger.kernel.org, linux-clk@vger.kernel.org,
        devicetree@vger.kernel.org, linux-kernel@vger.kernel.org,
        Taniya Das <quic_tdas@quicinc.com>, quic_rohiagar@quicinc.com,
        netdev@vger.kernel.org
To:     Andy Gross <agross@kernel.org>,
        Bjorn Andersson <andersson@kernel.org>,
        Krzysztof Kozlowski <krzysztof.kozlowski+dt@linaro.org>,
        Michael Turquette <mturquette@baylibre.com>,
        Richard Cochran <richardcochran@gmail.com>,
        Rob Herring <robh+dt@kernel.org>,
        Taniya Das <quic_tdas@quicinc.com>
Date:   Wed, 19 Apr 2023 14:37:10 -0700
User-Agent: alot/0.10
Precedence: bulk

Quoting Taniya Das (2023-04-19 06:30:10)
> From: Imran Shaik <quic_imrashai@quicinc.com>
>=20
> Add support to handle the invert logic for branch2 clocks.
> Invert branch halt would indicate the clock ON when CLK_OFF
> bit is '1' and OFF when CLK_OFF bit is '0'.
>=20
> Signed-off-by: Imran Shaik <quic_imrashai@quicinc.com>
> Signed-off-by: Taniya Das <quic_tdas@quicinc.com>
> ---
>  drivers/clk/qcom/clk-branch.c | 8 ++++++++
>  1 file changed, 8 insertions(+)
>=20
> diff --git a/drivers/clk/qcom/clk-branch.c b/drivers/clk/qcom/clk-branch.c
> index f869fc6aaed6..4b24d45be771 100644
> --- a/drivers/clk/qcom/clk-branch.c
> +++ b/drivers/clk/qcom/clk-branch.c
> @@ -48,6 +48,7 @@ static bool clk_branch2_check_halt(const struct clk_bra=
nch *br, bool enabling)
>  {
>         u32 val;
>         u32 mask;
> +       bool invert =3D (br->halt_check =3D=3D BRANCH_HALT_ENABLE);
> =20
>         mask =3D BRANCH_NOC_FSM_STATUS_MASK << BRANCH_NOC_FSM_STATUS_SHIF=
T;
>         mask |=3D BRANCH_CLK_OFF;
> @@ -56,9 +57,16 @@ static bool clk_branch2_check_halt(const struct clk_br=
anch *br, bool enabling)
> =20
>         if (enabling) {
>                 val &=3D mask;
> +
> +               if (invert)
> +                       return (val & BRANCH_CLK_OFF) =3D=3D BRANCH_CLK_O=
FF;
> +
>                 return (val & BRANCH_CLK_OFF) =3D=3D 0 ||
>                         val =3D=3D BRANCH_NOC_FSM_STATUS_ON;

Do these clks have a NOC_FSM_STATUS bit? I think it would be better to
make a local variable for the val we're looking for, and then test for
that. We may need a mask as well, but the idea is to not duplicate the
test and return from multiple places.

>         } else {
> +               if (invert)
> +                       return (val & BRANCH_CLK_OFF) =3D=3D 0;
> +
>                 return val & BRANCH_CLK_OFF;
>         }

While at it, I'd get rid of this else and de-indent the code because if
we're 'enabling' we'll return from the function regardless.