Linux clock framework development
 help / color / mirror / Atom feed
From: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
To: Konrad Dybcio <konrad.dybcio@linaro.org>,
	Andy Gross <agross@kernel.org>,
	Bjorn Andersson <andersson@kernel.org>,
	Stephen Boyd <sboyd@kernel.org>,
	Michael Turquette <mturquette@baylibre.com>,
	Rob Herring <robh+dt@kernel.org>,
	Krzysztof Kozlowski <krzysztof.kozlowski+dt@linaro.org>,
	Taniya Das <quic_tdas@quicinc.com>
Cc: linux-arm-msm@vger.kernel.org, linux-clk@vger.kernel.org,
	devicetree@vger.kernel.org,
	Konrad Dybcio <konrad.dybcio@somainline.org>
Subject: Re: [PATCH 2/5] clk: qcom: add msm8996 Core Bus Framework (CBF) support
Date: Thu, 12 Jan 2023 21:26:29 +0200	[thread overview]
Message-ID: <b395f413-34ec-b327-7a2d-6c56b0c5f861@linaro.org> (raw)
In-Reply-To: <1b25e736-1894-5144-2c15-f45fc70d8c74@linaro.org>

On 12/01/2023 16:58, Konrad Dybcio wrote:
> 
> 
> On 11.01.2023 20:57, Dmitry Baryshkov wrote:
>> Add CBF clock driver as a part of MSM8996 CPU clocks. Significantly
>> based on AngeloGioacchino del Regno's work at [1].
>>
>> The CBF is an interconnect between two CPU clusters, setting it up
>> properly is required for booting the MSM8996 with all four cores
>> enabled.
>>
>> [1] https://github.com/sonyxperiadev/kernel/blob/aosp/LE.UM.2.3.2.r1.4/drivers/clk/qcom/clk-cpu-8996.c
>>
>> Co-developed-by: Konrad Dybcio <konrad.dybcio@somainline.org>
>> Signed-off-by: Konrad Dybcio <konrad.dybcio@somainline.org>
>> Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
>> ---
>>   drivers/clk/qcom/Makefile       |   2 +-
>>   drivers/clk/qcom/clk-cbf-8996.c | 318 ++++++++++++++++++++++++++++++++
>>   2 files changed, 319 insertions(+), 1 deletion(-)
>>   create mode 100644 drivers/clk/qcom/clk-cbf-8996.c
>>
>> diff --git a/drivers/clk/qcom/Makefile b/drivers/clk/qcom/Makefile
>> index ca2f586edb3e..d6ecd2bf7040 100644
>> --- a/drivers/clk/qcom/Makefile
>> +++ b/drivers/clk/qcom/Makefile
>> @@ -52,7 +52,7 @@ obj-$(CONFIG_MSM_MMCC_8998) += mmcc-msm8998.o
>>   obj-$(CONFIG_QCOM_A53PLL) += a53-pll.o
>>   obj-$(CONFIG_QCOM_A7PLL) += a7-pll.o
>>   obj-$(CONFIG_QCOM_CLK_APCS_MSM8916) += apcs-msm8916.o
>> -obj-$(CONFIG_QCOM_CLK_APCC_MSM8996) += apcs-msm8996.o clk-cpu-8996.o
>> +obj-$(CONFIG_QCOM_CLK_APCC_MSM8996) += apcs-msm8996.o clk-cpu-8996.o clk-cbf-8996.o
>>   obj-$(CONFIG_QCOM_CLK_APCS_SDX55) += apcs-sdx55.o
>>   obj-$(CONFIG_QCOM_CLK_RPM) += clk-rpm.o
>>   obj-$(CONFIG_QCOM_CLK_RPMH) += clk-rpmh.o
>> diff --git a/drivers/clk/qcom/clk-cbf-8996.c b/drivers/clk/qcom/clk-cbf-8996.c
>> new file mode 100644
>> index 000000000000..bdd4f8b48a7e
>> --- /dev/null
>> +++ b/drivers/clk/qcom/clk-cbf-8996.c
>> @@ -0,0 +1,318 @@
>> +// SPDX-License-Identifier: GPL-2.0
>> +/*
>> + * Copyright (c) 2022, Linaro Ltd.
> Happy new year!
> 
>> + */
>> +#include <linux/bitfield.h>
>> +#include <linux/clk.h>
>> +#include <linux/clk-provider.h>
>> +#include <linux/of.h>
>> +#include <linux/module.h>
>> +#include <linux/platform_device.h>
>> +#include <linux/regmap.h>
>> +
>> +#include "clk-alpha-pll.h"
>> +#include "clk-regmap.h"
>> +
>> +/* Need to match the order of clocks in DT binding */
>> +enum {
>> +	DT_XO,
> No, it should be XO_A, otherwise the platform will never sleep.
> 
> +	DT_APCS_AUX,
>> +};
>> +
>> +enum {
>> +	CBF_XO_INDEX,
>> +	CBF_PLL_INDEX,
>> +	CBF_DIV_INDEX,
>> +	CBF_APCS_AUX_INDEX,
>> +};
>> +
>> +#define DIV_THRESHOLD		600000000
>> +
>> +#define CBF_MUX_OFFSET		0x18
>> +#define CBF_MUX_PARENT_MASK		GENMASK(1, 0)
>> +#define CBF_MUX_AUTO_CLK_SEL_ALWAYS_ON_MASK GENMASK(5, 4)
>> +#define CBF_MUX_AUTO_CLK_SEL_ALWAYS_ON_GPLL0_SEL \
>> +	FIELD_PREP(CBF_MUX_AUTO_CLK_SEL_ALWAYS_ON_MASK, 0x03)
>> +#define CBF_MUX_AUTO_CLK_SEL_BIT	BIT(6)
>> +
>> +#define CBF_PLL_OFFSET 0xf000
>> +
>> +static const u8 cbf_pll_regs[PLL_OFF_MAX_REGS] = {
>> +	[PLL_OFF_L_VAL] = 0x08,
>> +	[PLL_OFF_ALPHA_VAL] = 0x10,
>> +	[PLL_OFF_USER_CTL] = 0x18,
>> +	[PLL_OFF_CONFIG_CTL] = 0x20,
>> +	[PLL_OFF_CONFIG_CTL_U] = 0x24,
>> +	[PLL_OFF_TEST_CTL] = 0x30,
>> +	[PLL_OFF_TEST_CTL_U] = 0x34,
>> +	[PLL_OFF_STATUS] = 0x28,
>> +};
>> +
>> +static const struct alpha_pll_config cbfpll_config = {
>> +	.l = 72,
>> +	.config_ctl_val = 0x200d4828,
>> +	.config_ctl_hi_val = 0x006,
>> +	.test_ctl_val = 0x1c000000,
>> +	.test_ctl_hi_val = 0x00004000,
>> +	.pre_div_mask = BIT(12),
>> +	.post_div_mask = 0x3 << 8,
>> +	.post_div_val = 0x1 << 8,
>> +	.main_output_mask = BIT(0),
>> +	.early_output_mask = BIT(3),
>> +};
>> +
>> +static struct clk_alpha_pll cbf_pll = {
>> +	.offset = CBF_PLL_OFFSET,
>> +	.regs = cbf_pll_regs,
>> +	.flags = SUPPORTS_DYNAMIC_UPDATE | SUPPORTS_FSM_MODE,
>> +	.clkr.hw.init = &(struct clk_init_data){
>> +		.name = "cbf_pll",
>> +		.parent_data = (const struct clk_parent_data[]) {
>> +			{ .index = DT_XO, },
>> +		},
>> +		.num_parents = 1,
>> +		.ops = &clk_alpha_pll_hwfsm_ops,
>> +	},
>> +};
>> +
>> +static struct clk_fixed_factor cbf_pll_postdiv = {
>> +	.mult = 1,
>> +	.div = 2,
>> +	.hw.init = &(struct clk_init_data){
>> +		.name = "cbf_pll_postdiv",
>> +		.parent_data = &(const struct clk_parent_data){
> parent_hws?

Ack

> 
>> +			.hw = &cbf_pll.clkr.hw
>> +		},
>> +		.num_parents = 1,
>> +		.ops = &clk_fixed_factor_ops,
>> +		.flags = CLK_SET_RATE_PARENT,
>> +	},
>> +};
>> +
>> +static const struct clk_parent_data cbf_mux_parent_data[] = {
>> +	{ .index = DT_XO },
>> +	{ .hw = &cbf_pll.clkr.hw },
>> +	{ .hw = &cbf_pll_postdiv.hw },
>> +	{ .index = DT_APCS_AUX },
>> +};
>> +
>> +struct clk_cbf_8996_mux {
>> +	u32 reg;
>> +	struct notifier_block nb;
>> +	struct clk_regmap clkr;
>> +};
>> +
>> +static struct clk_cbf_8996_mux *to_clk_cbf_8996_mux(struct clk_regmap *clkr)
>> +{
>> +	return container_of(clkr, struct clk_cbf_8996_mux, clkr);
>> +}
>> +
>> +static int cbf_clk_notifier_cb(struct notifier_block *nb, unsigned long event,
>> +			       void *data);
> Please align the indentation with the opening bracket.
> 
>> +
>> +static u8 clk_cbf_8996_mux_get_parent(struct clk_hw *hw)
>> +{
>> +	struct clk_regmap *clkr = to_clk_regmap(hw);
>> +	struct clk_cbf_8996_mux *mux = to_clk_cbf_8996_mux(clkr);
>> +	u32 val;
>> +
>> +	regmap_read(clkr->regmap, mux->reg, &val);
>> +
>> +	return FIELD_GET(CBF_MUX_PARENT_MASK, val);
>> +}
>> +
>> +static int clk_cbf_8996_mux_set_parent(struct clk_hw *hw, u8 index)
>> +{
>> +	struct clk_regmap *clkr = to_clk_regmap(hw);
>> +	struct clk_cbf_8996_mux *mux = to_clk_cbf_8996_mux(clkr);
>> +	u32 val;
>> +
>> +	val = FIELD_PREP(CBF_MUX_PARENT_MASK, index);
>> +
>> +	return regmap_update_bits(clkr->regmap, mux->reg, CBF_MUX_PARENT_MASK, val);
>> +}
>> +
>> +static int clk_cbf_8996_mux_determine_rate(struct clk_hw *hw,
>> +					   struct clk_rate_request *req)
>> +{
>> +	struct clk_hw *parent;
>> +
>> +	if (req->rate < (DIV_THRESHOLD / 2))
>> +		return -EINVAL;
> Shouldn't you parent it to APCS AUX here?

No, frequencies less than 300 MHz are not supported at all, they are not 
listed in the tables. I think returning an error is more appropriate 
here than setting the clock to the fixed 300 MHz.

> 
>> +
>> +	if (req->rate < DIV_THRESHOLD)
>> +		parent = clk_hw_get_parent_by_index(hw, CBF_DIV_INDEX);
>> +	else
>> +		parent = clk_hw_get_parent_by_index(hw, CBF_PLL_INDEX);
>> +
>> +	if (!parent)
>> +		return -EINVAL;
>> +
>> +	req->best_parent_rate = clk_hw_round_rate(parent, req->rate);
>> +	req->best_parent_hw = parent;
>> +
>> +	return 0;
>> +}
>> +
>> +static const struct clk_ops clk_cbf_8996_mux_ops = {
>> +	.set_parent = clk_cbf_8996_mux_set_parent,
>> +	.get_parent = clk_cbf_8996_mux_get_parent,
>> +	.determine_rate = clk_cbf_8996_mux_determine_rate,
>> +};
>> +
>> +static struct clk_cbf_8996_mux cbf_mux = {
>> +	.reg = CBF_MUX_OFFSET,
>> +	.nb.notifier_call = cbf_clk_notifier_cb,
>> +	.clkr.hw.init = &(struct clk_init_data) {
>> +		.name = "cbf_mux",
>> +		.parent_data = cbf_mux_parent_data,
>> +		.num_parents = ARRAY_SIZE(cbf_mux_parent_data),
>> +		.ops = &clk_cbf_8996_mux_ops,
>> +		/* CPU clock is critical and should never be gated */
>> +		.flags = CLK_SET_RATE_PARENT | CLK_IS_CRITICAL,
>> +	},
>> +};
>> +
>> +static int cbf_clk_notifier_cb(struct notifier_block *nb, unsigned long event,
>> +			       void *data)
>> +{
>> +	struct clk_notifier_data *cnd = data;
>> +	int ret;
>> +
>> +	switch (event) {
>> +	case PRE_RATE_CHANGE:
>> +		/*
>> +		 * Avoid overvolting. clk_core_set_rate_nolock() walks from top
>> +		 * to bottom, so it will change the rate of the PLL before
>> +		 * chaging the parent of PMUX. This can result in pmux getting
>> +		 * clocked twice the expected rate.
>> +		 *
>> +		 * Manually switch to PLL/2 here.
>> +		 */
>> +		if (cnd->old_rate > DIV_THRESHOLD &&
>> +		    cnd->new_rate < DIV_THRESHOLD)
>> +			clk_cbf_8996_mux_set_parent(&cbf_mux.clkr.hw, CBF_DIV_INDEX);
>> +		break;
>> +	case ABORT_RATE_CHANGE:
>> +		/* Revert manual change */
>> +		if (cnd->new_rate < DIV_THRESHOLD &&
>> +		    cnd->old_rate > DIV_THRESHOLD)
>> +			clk_cbf_8996_mux_set_parent(&cbf_mux.clkr.hw, CBF_PLL_INDEX);
>> +		break;
>> +	default:
>> +		ret = 0;
>> +		break;
>> +	}
>> +
>> +	return notifier_from_errno(ret);
>> +};
>> +
>> +static struct clk_hw *cbf_msm8996_hw_clks[] = {
>> +	&cbf_pll_postdiv.hw,
>> +};
>> +
>> +static struct clk_regmap *cbf_msm8996_clks[] = {
>> +	&cbf_pll.clkr,
>> +	&cbf_mux.clkr,
>> +};
>> +
>> +static const struct regmap_config cbf_msm8996_regmap_config = {
>> +	.reg_bits		= 32,
>> +	.reg_stride		= 4,
>> +	.val_bits		= 32,
>> +	.max_register		= 0x10000,
>> +	.fast_io		= true,
>> +	.val_format_endian	= REGMAP_ENDIAN_LITTLE,
>> +};
>> +
>> +static int qcom_msm8996_cbf_probe(struct platform_device *pdev)
>> +{
>> +	void __iomem *base;
>> +	struct regmap *regmap;
>> +	struct device *dev = &pdev->dev;
>> +	int i, ret;
>> +
>> +	base = devm_platform_ioremap_resource(pdev, 0);
>> +	if (IS_ERR(base))
>> +		return PTR_ERR(base);
>> +
>> +	regmap = devm_regmap_init_mmio(dev, base, &cbf_msm8996_regmap_config);
>> +	if (IS_ERR(regmap))
>> +		return PTR_ERR(regmap);
>> +
>> +	/* Select GPLL0 for 300MHz for the CBF clock */
>> +	regmap_write(regmap, CBF_MUX_OFFSET, 0x3);
>> +
>> +	/* Ensure write goes through before PLLs are reconfigured */
>> +	udelay(5);
>> +
>> +	/* Set the auto clock sel always-on source to GPLL0/2 (300MHz) */
>> +	regmap_update_bits(regmap, CBF_MUX_OFFSET,
>> +			   CBF_MUX_AUTO_CLK_SEL_ALWAYS_ON_MASK,
>> +			   CBF_MUX_AUTO_CLK_SEL_ALWAYS_ON_GPLL0_SEL);
>> +
>> +	clk_alpha_pll_configure(&cbf_pll, regmap, &cbfpll_config);
>> +
>> +	/* Wait for PLL(s) to lock */
>> +        udelay(50);
> Weird indentation?
> 
>> +
>> +	/* Enable auto clock selection for CBF */
>> +	regmap_update_bits(regmap, CBF_MUX_OFFSET,
>> +			   CBF_MUX_AUTO_CLK_SEL_BIT,
>> +			   CBF_MUX_AUTO_CLK_SEL_BIT);
>> +
>> +	/* Ensure write goes through before muxes are switched */
>> +	udelay(5);
>> +
>> +	/* Switch CBF to use the primary PLL */
>> +	regmap_update_bits(regmap, CBF_MUX_OFFSET,
>> +			   CBF_MUX_PARENT_MASK, 0x1);
> I think you can unwrap this.
> 
>> +
>> +	for (i = 0; i < ARRAY_SIZE(cbf_msm8996_hw_clks); i++) {
>> +		ret = devm_clk_hw_register(dev, cbf_msm8996_hw_clks[i]);
>> +		if (ret)
>> +			return ret;
>> +	}
>> +
>> +	for (i = 0; i < ARRAY_SIZE(cbf_msm8996_clks); i++) {
>> +		ret = devm_clk_register_regmap(dev, cbf_msm8996_clks[i]);
>> +		if (ret)
>> +			return ret;
>> +	}
>> +
>> +	ret = devm_clk_notifier_register(dev, cbf_mux.clkr.hw.clk, &cbf_mux.nb);
>> +	if (ret)
>> +		return ret;
>> +
>> +	return devm_of_clk_add_hw_provider(dev, of_clk_hw_simple_get, &cbf_mux.clkr.hw);
>> +}
>> +
>> +static const struct of_device_id qcom_msm8996_cbf_match_table[] = {
>> +	{ .compatible = "qcom,msm8996-cbf" },
>> +	{ /* sentinel */ },
>> +};
>> +MODULE_DEVICE_TABLE(of, qcom_msm8996_cbf_match_table);
>> +
>> +static struct platform_driver qcom_msm8996_cbf_driver = {
>> +	.probe = qcom_msm8996_cbf_probe,
>> +	.driver = {
>> +		.name = "qcom-msm8996-cbf",
>> +		.of_match_table = qcom_msm8996_cbf_match_table,
>> +	},
>> +};
>> +
>> +/* Register early enough to fix the clock to be used for other cores */
>> +static int __init qcom_msm8996_cbf_init(void)
>> +{
>> +	return platform_driver_register(&qcom_msm8996_cbf_driver);
>> +}
>> +postcore_initcall(qcom_msm8996_cbf_init);
>> +
>> +static void __exit qcom_msm8996_cbf_exit(void)
>> +{
>> +	platform_driver_unregister(&qcom_msm8996_cbf_driver);
>> +}
>> +module_exit(qcom_msm8996_cbf_exit);
>> +
>> +MODULE_DESCRIPTION("QCOM MSM8996 CPU Bus Fabric Clock Driver");
>> +MODULE_LICENSE("GPL v2");
> "GPL"

Ack

> 
> The rest looks good..
> 
> Konrad

-- 
With best wishes
Dmitry


  reply	other threads:[~2023-01-12 19:32 UTC|newest]

Thread overview: 17+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-01-11 19:57 [PATCH 0/5] clk: qcom: msm8996: add support for the CBF clock Dmitry Baryshkov
2023-01-11 19:57 ` [PATCH 1/5] dt-bindings: clock: qcom,msm8996-cbf: Describe the MSM8996 CBF clock controller Dmitry Baryshkov
2023-01-12  8:40   ` Krzysztof Kozlowski
2023-01-12  9:47     ` Dmitry Baryshkov
2023-01-12 11:46       ` Krzysztof Kozlowski
2023-01-11 19:57 ` [PATCH 2/5] clk: qcom: add msm8996 Core Bus Framework (CBF) support Dmitry Baryshkov
2023-01-12 14:58   ` Konrad Dybcio
2023-01-12 19:26     ` Dmitry Baryshkov [this message]
2023-01-11 19:57 ` [PATCH 3/5] clk: qcom: cbf-msm8996: scale CBF clock according to the CPUfreq Dmitry Baryshkov
2023-01-12 15:00   ` Konrad Dybcio
2023-01-12 15:29     ` Dmitry Baryshkov
2023-01-12 15:37       ` Konrad Dybcio
2023-01-11 19:57 ` [PATCH 4/5] arm64: dts: qcom: msm8996: add CBF device entry Dmitry Baryshkov
2023-01-12 15:05   ` Konrad Dybcio
2023-01-12 15:38     ` Dmitry Baryshkov
2023-01-12 15:39       ` Konrad Dybcio
2023-01-11 19:57 ` [PATCH 5/5] arm64: dts: qcom: msm8996: scale CBF clock according to the CPUfreq Dmitry Baryshkov

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=b395f413-34ec-b327-7a2d-6c56b0c5f861@linaro.org \
    --to=dmitry.baryshkov@linaro.org \
    --cc=agross@kernel.org \
    --cc=andersson@kernel.org \
    --cc=devicetree@vger.kernel.org \
    --cc=konrad.dybcio@linaro.org \
    --cc=konrad.dybcio@somainline.org \
    --cc=krzysztof.kozlowski+dt@linaro.org \
    --cc=linux-arm-msm@vger.kernel.org \
    --cc=linux-clk@vger.kernel.org \
    --cc=mturquette@baylibre.com \
    --cc=quic_tdas@quicinc.com \
    --cc=robh+dt@kernel.org \
    --cc=sboyd@kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox