Re: [PATCH 5/6] clk: qcom: gcc-sdm660: Account for needed adjustments in probe function

From: Stephen Boyd <sboyd@kernel.org>
To: Konrad Dybcio <konrad.dybcio@somainline.org>,
	phone-devel@vger.kernel.org
Cc: ~postmarketos/upstreaming@lists.sr.ht,
	martin.botka@somainline.org,
	angelogioacchino.delregno@somainline.org,
	marijn.suijten@somainline.org,
	Konrad Dybcio <konrad.dybcio@somainline.org>,
	Andy Gross <agross@kernel.org>,
	Bjorn Andersson <bjorn.andersson@linaro.org>,
	Michael Turquette <mturquette@baylibre.com>,
	Rob Herring <robh+dt@kernel.org>,
	Taniya Das <tdas@codeaurora.org>,
	Craig Tatlor <ctatlor97@gmail.com>,
	linux-arm-msm@vger.kernel.org, linux-clk@vger.kernel.org,
	linux-kernel@vger.kernel.org, devicetree@vger.kernel.org
Subject: Re: [PATCH 5/6] clk: qcom: gcc-sdm660: Account for needed adjustments in probe function
Date: Mon, 22 Feb 2021 16:39:33 -0800	[thread overview]
Message-ID: <161404077336.1254594.15002572465360321874@swboyd.mtv.corp.google.com> (raw)
In-Reply-To: <20210220155618.176559-5-konrad.dybcio@somainline.org>

Quoting Konrad Dybcio (2021-02-20 07:56:16)
> Downstream kernel executes a bunch of commands, such as keeping
> GPU/MMSS interface clocks alive to make sure all subsystems can
> work properly. Add these to make sure they do.
> 
> Fixes: f2a76a2955c0 ("clk: qcom: Add Global Clock controller (GCC) driver for SDM660")
> Signed-off-by: Konrad Dybcio <konrad.dybcio@somainline.org>
> ---
>  drivers/clk/qcom/gcc-sdm660.c | 23 ++++++++++++++++++++++-
>  1 file changed, 22 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/clk/qcom/gcc-sdm660.c b/drivers/clk/qcom/gcc-sdm660.c
> index bc8dfcd6d629..db2185c88b77 100644
> --- a/drivers/clk/qcom/gcc-sdm660.c
> +++ b/drivers/clk/qcom/gcc-sdm660.c
> @@ -11,6 +11,7 @@
>  #include <linux/module.h>
>  #include <linux/of.h>
>  #include <linux/of_device.h>
> +#include <linux/clk.h>
>  #include <linux/clk-provider.h>
>  #include <linux/regmap.h>
>  #include <linux/reset-controller.h>
> @@ -2622,7 +2623,27 @@ static int gcc_sdm660_probe(struct platform_device *pdev)
>         if (ret)
>                 return ret;
>  
> -       return qcom_cc_really_probe(pdev, &gcc_sdm660_desc, regmap);
> +       ret = qcom_cc_really_probe(pdev, &gcc_sdm660_desc, regmap);
> +       if (ret)
> +               return ret;
> +
> +       /* Disable the GPLL0 active input to MMSS and GPU via MISC registers */
> +       regmap_update_bits(regmap, 0x0902c, 0x3, 0x3);
> +       regmap_update_bits(regmap, 0x71028, 0x3, 0x3);
> +
> +       /* This clock is used for all MMSSCC register access */
> +       clk_prepare_enable(gcc_mmss_noc_cfg_ahb_clk.clkr.hw.clk);
> +
> +       /* This clock is used for all GPUCC register access */
> +       clk_prepare_enable(gcc_gpu_cfg_ahb_clk.clkr.hw.clk);
> +
> +       /* Keep bimc gfx clock port on all the time */
> +       clk_prepare_enable(gcc_bimc_gfx_clk.clkr.hw.clk);
> +

Preferably just set these various bits with regmap_update_bits() during
probe. Also, please do it before regsitering the clks, not after.

> +       /* Set the HMSS_GPLL0_SRC for 300MHz to CPU subsystem */
> +       clk_set_rate(hmss_gpll0_clk_src.clkr.hw.clk, 300000000);

Is this not already the case?

> +
> +       return ret;
>  }
>