On Thu, 2022-04-07 at 12:20 +0900, Chanwoo Choi wrote: > Hi, > > It seems that this patch depends on patches[1] of devfreq-testing > branch. > [1] > https://urldefense.com/v3/__https://git.kernel.org/pub/scm/linux/kernel/git/chanwoo/linux.git/log/?h=devfreq-testing__;!!CTRNKA9wMg0ARbw!zpzMoFhloy2fpTawf8kkY5nwKerht5CKCnKTwwmSHYBN2vN5WfVMOcGn5TDb8SPTQTSaHA$ > > > I'll send the updated patches[1]. And I add the my comment below. > Yes, I will add your patches as my dependency. > > On 3/7/22 9:25 PM, Tim Chang wrote: > > This adds a devfreq driver for the Cache Coherent Interconnect > > (CCI) of > > the Mediatek MT8183. > > > > On mt8183 the cci is supplied by the same regulator as the little > > cores. > > The driver is notified when the regulator voltage changes (driven > > by > > cpufreq) and adjusts the cci frequency to the maximum possible > > value. > > > > Add need_voltage_tracking variable to platforma data. if true, it > > indicates soc is required to realize the voltage tracking between > > voltage of sram and voltage of cci by software approach. otherwise, > > the > > voltage tracking is realized by hardware appraoch. > > > > Add the notifier to cci so that it could react after svs driver > > changes > > opp table of cci. > > > > Signed-off-by: Jia-Wei Chang < > > jia-wei.chang@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx> > > --- > > drivers/devfreq/Kconfig | 11 +- > > drivers/devfreq/Makefile | 2 +- > > drivers/devfreq/mtk-cci-devfreq.c | 471 > > ++++++++++++++++++++++++++++++ > > 3 files changed, 482 insertions(+), 2 deletions(-) > > create mode 100644 drivers/devfreq/mtk-cci-devfreq.c > > > > diff --git a/drivers/devfreq/Kconfig b/drivers/devfreq/Kconfig > > index 1ec36ae93f31..a6be3c6b5691 100644 > > --- a/drivers/devfreq/Kconfig > > +++ b/drivers/devfreq/Kconfig > > @@ -110,7 +110,7 @@ config ARM_IMX8M_DDRC_DEVFREQ > > > > config ARM_MT8183_CCI_DEVFREQ > > tristate "MT8183 CCI DEVFREQ Driver" > > - depends on ARM_MEDIATEK_CPUFREQ > > + depends on OF && ARM_MEDIATEK_CPUFREQ > > help > > ARM_MT8183_CCI_DEVFREQ is not existed on mainline kernel. > You need to rework this patch on the latest kernel (either devfreq- > next > or linux-next.git). Sure, I will rebase my patch onto linux-next. > > > This adds a devfreq driver for Cache Coherent > > Interconnect > > of Mediatek MT8183, which is shared the same regulator > > @@ -130,6 +130,15 @@ config ARM_TEGRA_DEVFREQ > > It reads ACTMON counters of memory controllers and adjusts > > the > > operating frequencies and voltages with OPP support. > > > > +config ARM_MEDIATEK_CCI_DEVFREQ > > + tristate "MEDIATEK CCI DEVFREQ Driver" > > + depends on OF && ARM_MEDIATEK_CPUFREQ > > + help > > + This adds a devfreq driver for Mediatek Cache Coherent > > Interconnect > > + which is shared the same regulator with cpu cluster. It can > > track > > + buck voltage and update a proper CCI frequency. Use > > notification > > + to get regulator status. > > + > > config ARM_RK3399_DMC_DEVFREQ > > tristate "ARM RK3399 DMC DEVFREQ Driver" > > depends on (ARCH_ROCKCHIP && HAVE_ARM_SMCCC) || \ > > diff --git a/drivers/devfreq/Makefile b/drivers/devfreq/Makefile > > index 991ef7740759..0493516a16f2 100644 > > --- a/drivers/devfreq/Makefile > > +++ b/drivers/devfreq/Makefile > > @@ -11,7 +11,7 @@ obj-$(CONFIG_DEVFREQ_GOV_PASSIVE) += > > governor_passive.o > > obj-$(CONFIG_ARM_EXYNOS_BUS_DEVFREQ) += exynos-bus.o > > obj-$(CONFIG_ARM_IMX_BUS_DEVFREQ) += imx-bus.o > > obj-$(CONFIG_ARM_IMX8M_DDRC_DEVFREQ) += imx8m-ddrc.o > > -obj-$(CONFIG_ARM_MT8183_CCI_DEVFREQ) += mt8183-cci-devfreq.o > > +obj-$(CONFIG_ARM_MEDIATEK_CCI_DEVFREQ) += mtk-cci-devfreq.o > > As I commented, mt8183-cci-devfreq.c is not existed on latest kernel. Yes, it will be fixed after rebasing onto linux-next. > > > obj-$(CONFIG_ARM_RK3399_DMC_DEVFREQ) += rk3399_dmc.o > > obj-$(CONFIG_ARM_TEGRA_DEVFREQ) += tegra30-devfreq.o > > > > diff --git a/drivers/devfreq/mtk-cci-devfreq.c > > b/drivers/devfreq/mtk-cci-devfreq.c > > new file mode 100644 > > index 000000000000..986f34689f5c > > --- /dev/null > > +++ b/drivers/devfreq/mtk-cci-devfreq.c > > @@ -0,0 +1,471 @@ > > +// SPDX-License-Identifier: GPL-2.0-only > > +/* > > + * Copyright (C) 2022 MediaTek Inc. > > + */ > > + > > +#include <linux/clk.h> > > +#include <linux/devfreq.h> > > +#include <linux/minmax.h> > > +#include <linux/module.h> > > +#include <linux/of.h> > > +#include <linux/of_device.h> > > +#include <linux/platform_device.h> > > +#include <linux/pm_opp.h> > > +#include <linux/regulator/consumer.h> > > + > > +struct mtk_ccifreq_platform_data; > > Instead of unneeded declaration, > just define 'strcut mmk_ccifreq_platform_data' at here. Yes, I will remove this forward declaration. > > > + > > +struct mtk_ccifreq_drv { > > + struct device *cci_dev; > > + struct devfreq *devfreq; > > + struct regulator *proc_reg; > > + struct regulator *sram_reg; > > + struct clk *cci_clk; > > + struct clk *inter_clk; > > + int inter_voltage; > > + int old_voltage; > > + unsigned long old_freq; > > + struct mutex lock; /* avoid notify and policy race condition > > */ > > + struct notifier_block opp_nb; > > + const struct mtk_ccifreq_platform_data *soc_data; > > +}; > > + > > +struct mtk_ccifreq_platform_data { > > + int min_volt_shift; > > + int max_volt_shift; > > + int proc_max_volt; > > + int sram_min_volt; > > + int sram_max_volt; > > + bool need_voltage_tracking; > > +}; > > As I commented, move it before struct mtk_ccifreq_drv. Yes, I will update it in the next version. > > > + > > +static int mtk_ccifreq_voltage_tracking(struct mtk_ccifreq_drv > > *drv, int new_voltage) > > +{ > > + const struct mtk_ccifreq_platform_data *soc_data = drv- > > >soc_data; > > + struct regulator *proc_reg = drv->proc_reg; > > + struct regulator *sram_reg = drv->sram_reg; > > + int old_voltage, old_vsram, new_vsram, vsram, voltage, ret; > > + > > + old_voltage = regulator_get_voltage(proc_reg); > > + if (old_voltage < 0) { > > + pr_err("%s: invalid vproc value: %d\n", __func__, > > old_voltage); > > Use dev_err instead of pr_err and don't use '__func__'. Yes, I will fix it in the next version. > > > + return old_voltage; > > + } > > + > > + old_vsram = regulator_get_voltage(sram_reg); > > + if (old_vsram < 0) { > > + pr_err("%s: invalid vsram value: %d\n", __func__, > > old_vsram); > > ditto. Yes, I will fix it in the next version. > > > + return old_vsram; > > + } > > + > > + new_vsram = clamp(new_voltage + soc_data->min_volt_shift, > > + soc_data->sram_min_volt, soc_data- > > >sram_max_volt); > > + > > + do { > > + if (old_voltage <= new_voltage) { > > + vsram = clamp(old_voltage + soc_data- > > >max_volt_shift, > > + soc_data->sram_min_volt, > > new_vsram); > > + ret = regulator_set_voltage(sram_reg, vsram, > > + soc_data- > > >sram_max_volt); > > + if (ret) > > + return ret; > > + > > + if (vsram == soc_data->sram_max_volt || > > + new_vsram == soc_data->sram_min_volt) > > + voltage = new_voltage; > > + else > > + voltage = vsram - soc_data- > > >min_volt_shift; > > + > > + ret = regulator_set_voltage(proc_reg, voltage, > > + soc_data- > > >proc_max_volt); > > + if (ret) { > > + regulator_set_voltage(sram_reg, > > old_vsram, > > + soc_data- > > >sram_max_volt); > > + return ret; > > + } > > + } else if (old_voltage > new_voltage) { > > + voltage = max(new_voltage, > > + old_vsram - soc_data- > > >max_volt_shift); > > + ret = regulator_set_voltage(proc_reg, voltage, > > + soc_data- > > >proc_max_volt); > > + if (ret) > > + return ret; > > + > > + if (voltage == new_voltage) > > + vsram = new_vsram; > > + else > > + vsram = max(new_vsram, > > + voltage + soc_data- > > >min_volt_shift); > > + > > + ret = regulator_set_voltage(sram_reg, vsram, > > + soc_data- > > >sram_max_volt); > > + if (ret) { > > + regulator_set_voltage(proc_reg, > > old_voltage, > > + soc_data- > > >proc_max_volt); > > + return ret; > > + } > > + } > > + > > + old_voltage = voltage; > > + old_vsram = vsram; > > + } while (voltage != new_voltage || vsram != new_vsram); > > + > > + return 0; > > +} > > + > > +static int mtk_ccifreq_set_voltage(struct mtk_ccifreq_drv *drv, > > int voltage) > > +{ > > + int ret; > > + > > + if (drv->soc_data->need_voltage_tracking) > > + ret = mtk_ccifreq_voltage_tracking(drv, voltage); > > + else > > + ret = regulator_set_voltage(drv->proc_reg, voltage, > > + drv->soc_data- > > >proc_max_volt); > > + > > + if (!ret) > > + drv->old_voltage = voltage; > > + > > + return ret; > > +} > > + > > +static int mtk_ccifreq_target(struct device *dev, unsigned long > > *freq, > > + u32 flags) > > +{ > > I recommend that you handle the error case by using 'goto' statement > instead of handling them in each error case. You can refer the > other devfreq device driver for handling error case with goto. Sure, I will refactor it with 'goto' statement in the next version. > > > + struct mtk_ccifreq_drv *drv = dev_get_drvdata(dev); > > + struct clk *cci_pll = clk_get_parent(drv->cci_clk); > > + struct dev_pm_opp *opp; > > + unsigned long opp_rate; > > + int voltage, old_voltage, inter_voltage, target_voltage, ret; > > + > > + if (!drv) > > + return -EINVAL; > > + > > + if (drv->old_freq == *freq) > > + return 0; > > + > > + inter_voltage = drv->inter_voltage; > > + > > + opp_rate = *freq; > > + opp = devfreq_recommended_opp(dev, &opp_rate, 1); > > + if (IS_ERR(opp)) { > > + pr_err("cci: failed to find opp for freq: %ld\n", > > opp_rate); > > ditto. Ok, I will use 'goto' to handle error. > > > + return PTR_ERR(opp); > > + } > > + voltage = dev_pm_opp_get_voltage(opp); > > + dev_pm_opp_put(opp); > > + > > + old_voltage = drv->old_voltage; > > + if (old_voltage == 0) > > + old_voltage = regulator_get_voltage(drv->proc_reg); > > + if (old_voltage < 0) { > > + pr_err("cci: invalid vproc value: %d\n", old_voltage); > > ditto. Use dev_err instead of pr_err. Ok, I will use dev_err instread. > > > + return old_voltage; > > + } > > + > > + mutex_lock(&drv->lock); > > + > > + /* scale up: set voltage first then freq. */ > > + target_voltage = max(inter_voltage, voltage); > > + if (old_voltage <= target_voltage) { > > + ret = mtk_ccifreq_set_voltage(drv, target_voltage); > > + if (ret) { > > + pr_err("cci: failed to scale up voltage\n"); > > ditto. Ok, I will use dev_err instread. > > > + mtk_ccifreq_set_voltage(drv, old_voltage); > > + mutex_unlock(&drv->lock); > > + return ret; > > + } > > + } > > + > > + /* switch the cci clock to intermediate clock source. */ > > + ret = clk_set_parent(drv->cci_clk, drv->inter_clk); > > + if (ret) { > > + pr_err("cci: failed to re-parent cci clock\n"); > > ditto. Ok, I will use dev_err instread. > > > + mtk_ccifreq_set_voltage(drv, old_voltage); > > + WARN_ON(1); > > I think that it is not necessary. Please remove it. Yes, I will remove it. > > > + mutex_unlock(&drv->lock); > > + return ret; > > + } > > + > > + /* set the original clock to target rate. */ > > + ret = clk_set_rate(cci_pll, *freq); > > + if (ret) { > > + pr_err("cci: failed to set cci pll rate: %d\n", ret); > > ditto. Ok, I will use dev_err instread. > > > + clk_set_parent(drv->cci_clk, cci_pll); > > + mtk_ccifreq_set_voltage(drv, old_voltage); > > + mutex_unlock(&drv->lock); > > + return ret; > > + } > > + > > + /* switch the cci clock back to the original clock source. */ > > + ret = clk_set_parent(drv->cci_clk, cci_pll); > > + if (ret) { > > + pr_err("cci: failed to re-parent cci clock\n"); > > ditto. Ok, I will use dev_err instread. > > > + mtk_ccifreq_set_voltage(drv, inter_voltage); > > + WARN_ON(1); > > ditto. Ok, I will remove it. > > > + mutex_unlock(&drv->lock); > > + return ret; > > + } > > + > > + /* > > + * If the new voltage is lower than the intermediate voltage or > > the > > + * original voltage, scale down to the new voltage. > > + */ > > + if (voltage < inter_voltage || voltage < old_voltage) { > > + ret = mtk_ccifreq_set_voltage(drv, voltage); > > + if (ret) { > > + pr_err("cci: failed to scale down voltage\n"); > > ditto. Ok, I will use dev_err instread. > > > + WARN_ON(1); > > ditto. Ok, I will remove it. > > > + mutex_unlock(&drv->lock); > > + return ret; > > + } > > + } > > + > > + drv->old_freq = *freq; > > + mutex_unlock(&drv->lock); > > + > > + return 0; > > +} > > + > > +static int mtk_ccifreq_opp_notifier(struct notifier_block *nb, > > + unsigned long event, void *data) > > +{ > > + struct dev_pm_opp *opp = data; > > + struct mtk_ccifreq_drv *drv; > > + unsigned long freq, volt; > > + > > + drv = container_of(nb, struct mtk_ccifreq_drv, opp_nb); > > + > > + if (event == OPP_EVENT_ADJUST_VOLTAGE) { > > + freq = dev_pm_opp_get_freq(opp); > > + > > + mutex_lock(&drv->lock); > > + /* current opp item is changed */ > > + if (freq == drv->old_freq) { > > + volt = dev_pm_opp_get_voltage(opp); > > + mtk_ccifreq_set_voltage(drv, volt); > > + } > > + mutex_unlock(&drv->lock); > > + } > > + > > + return 0; > > +} > > + > > +static struct devfreq_dev_profile mtk_ccifreq_profile = { > > + .target = mtk_ccifreq_target, > > +}; > > + > > +static int mtk_ccifreq_probe(struct platform_device *pdev) > > +{ > > + struct device *cci_dev = &pdev->dev; > > + struct mtk_ccifreq_drv *drv; > > + struct devfreq_passive_data *passive_data; > > + struct dev_pm_opp *opp; > > + unsigned long rate, opp_volt; > > + int ret; > > + > > + drv = devm_kzalloc(cci_dev, sizeof(*drv), GFP_KERNEL); > > + if (!drv) > > + return -ENOMEM; > > + > > + drv->cci_dev = cci_dev; > > + drv->soc_data = (const struct mtk_ccifreq_platform_data *) > > + of_device_get_match_data(&pdev->dev); > > + mutex_init(&drv->lock); > > + platform_set_drvdata(pdev, drv); > > + > > + drv->cci_clk = devm_clk_get(cci_dev, "cci"); > > + if (IS_ERR(drv->cci_clk)) { > > + ret = PTR_ERR(drv->cci_clk); > > + return dev_err_probe(cci_dev, ret, > > + "cci: failed to get cci clk: > > %d\n", > > + ret); > > + } > > + > > + drv->inter_clk = devm_clk_get(cci_dev, "intermediate"); > > + if (IS_ERR(drv->inter_clk)) { > > + ret = PTR_ERR(drv->inter_clk); > > + return dev_err_probe(cci_dev, ret, > > + "cci: failed to get intermediate > > clk: %d\n", > > + ret); > > + } > > + > > + if (drv->soc_data->need_voltage_tracking) { > > + drv->sram_reg = regulator_get_optional(cci_dev, > > "sram"); > > + if (IS_ERR_OR_NULL(drv->sram_reg)) { > > + ret = PTR_ERR(drv->sram_reg); > > + return dev_err_probe(cci_dev, ret, > > + "cci: failed to get sram > > regulator: %d\n", > > + ret); > > + } > > + > > + ret = regulator_enable(drv->sram_reg); > > + if (ret) { > > + dev_warn(cci_dev, > > + "cci: failed to enable sram > > regulator\n"); > > If you want to break the probe sequence, better to use 'dev_err' > instead of 'dev_warn' if there is no any reason to use dev_warn. > Ok, I will use dev_err. > > > + return ret; > > + } > > + } > > + > > + drv->proc_reg = devm_regulator_get_optional(cci_dev, "proc"); > > + if (IS_ERR(drv->proc_reg)) { > > + ret = PTR_ERR(drv->proc_reg); > > + dev_err_probe(cci_dev, ret, > > + "cci: failed to get proc regulator: > > %d\n", > > + ret); > > + goto out_disable_sram_reg; > > + } > > + > > + ret = regulator_enable(drv->proc_reg); > > + if (ret) { > > + dev_warn(cci_dev, "cci: failed to enable proc > > regulator\n"); > > ditto. Use dev_err. Ok, I will replace all dev_warn with dev_err in the patch. > > > + goto out_disable_sram_reg; > > + } > > + > > + ret = clk_prepare_enable(drv->cci_clk); > > + if (ret) > > + goto out_disable_proc_reg; > > + > > + ret = clk_prepare_enable(drv->inter_clk); > > + if (ret) > > + goto out_disable_cci_clk; > > + > > + ret = dev_pm_opp_of_add_table(cci_dev); > > + if (ret) { > > + dev_warn(cci_dev, "cci: failed to add opp table: %d\n", > > ret); > > + goto out_disable_inter_clk; > > + } > > + > > + rate = clk_get_rate(drv->inter_clk); > > + opp = dev_pm_opp_find_freq_ceil(cci_dev, &rate); > > + if (IS_ERR(opp)) { > > + ret = PTR_ERR(opp); > > + dev_err(cci_dev, "cci: failed to get intermediate opp: > > %d\n", > > + ret); > > + goto out_remove_opp_table; > > + } > > + drv->inter_voltage = dev_pm_opp_get_voltage(opp); > > + dev_pm_opp_put(opp); > > + > > + rate = U32_MAX; > > + opp = dev_pm_opp_find_freq_floor(drv->cci_dev, &rate); > > + if (IS_ERR(opp)) { > > + pr_err("failed to get opp\n"); > > + ret = PTR_ERR(opp); > > + goto out_remove_opp_table; > > + } > > + > > + opp_volt = dev_pm_opp_get_voltage(opp); > > + dev_pm_opp_put(opp); > > + ret = mtk_ccifreq_set_voltage(drv, opp_volt); > > + if (ret) { > > + pr_err("failed to scale to highest voltage %lu in > > proc_reg\n", > > + opp_volt); > > Use dev_err. Ok, I will use dev_err instread. > > > + goto out_remove_opp_table; > > + } > > + > > + passive_data = devm_kzalloc(cci_dev, sizeof(struct > > devfreq_passive_data), GFP_KERNEL); > > + if (!passive_data) { > > + ret = -ENOMEM; > > + goto out_remove_opp_table; > > + } > > + > > + passive_data->parent_type = CPUFREQ_PARENT_DEV; > > It seems that depend on patchset[1]. I'll send the updated patch > within one week. > [1] > https://urldefense.com/v3/__https://git.kernel.org/pub/scm/linux/kernel/git/chanwoo/linux.git/log/?h=devfreq-testing__;!!CTRNKA9wMg0ARbw!zpzMoFhloy2fpTawf8kkY5nwKerht5CKCnKTwwmSHYBN2vN5WfVMOcGn5TDb8SPTQTSaHA$ > Should I send my patch after yours? Or the order makes no difference? > > > + drv->devfreq = devm_devfreq_add_device(cci_dev, > > + &mtk_ccifreq_profile, > > + DEVFREQ_GOV_PASSIVE, > > + passive_data); > > + if (IS_ERR(drv->devfreq)) { > > + ret = -EPROBE_DEFER; > > + dev_err(cci_dev, "cci: failed to add devfreq device: > > %d\n", > > + PTR_ERR(drv->devfreq)); > > + goto out_remove_opp_table; > > + } > > + > > + drv->opp_nb.notifier_call = mtk_ccifreq_opp_notifier; > > + ret = dev_pm_opp_register_notifier(cci_dev, &drv->opp_nb); > > + if (ret) { > > + dev_warn(cci_dev, "cci: failed to register opp > > notifier: %d\n", > > + ret); > > + goto out_remove_devfreq_device; > > + } > > + > > + return 0; > > + > > +out_remove_devfreq_device: > > + devm_devfreq_remove_device(cci_dev, drv->devfreq); > > + > > +out_remove_opp_table: > > + dev_pm_opp_of_remove_table(cci_dev); > > + > > +out_disable_inter_clk: > > + clk_disable_unprepare(drv->inter_clk); > > + > > +out_disable_cci_clk: > > + clk_disable_unprepare(drv->cci_clk); > > + > > +out_disable_proc_reg: > > + regulator_disable(drv->proc_reg); > > + > > +out_disable_sram_reg: > > + if (drv->soc_data->need_voltage_tracking) > > + regulator_disable(drv->sram_reg); > > + > > + return ret; > > +} > > + > > +static int mtk_ccifreq_remove(struct platform_device *pdev) > > +{ > > + struct device *cci_dev = &pdev->dev; > > + struct mtk_ccifreq_drv *drv; > > + > > + drv = platform_get_drvdata(pdev); > > + > > + dev_pm_opp_unregister_notifier(cci_dev, &drv->opp_nb); > > + dev_pm_opp_of_remove_table(cci_dev); > > + regulator_disable(drv->proc_reg); > > + if (!IS_ERR(drv->sram_reg)) > > + regulator_disable(drv->sram_reg); > > + > > + return 0; > > +} > > + > > +static const struct mtk_ccifreq_platform_data mtk_platform_data = > > { > > + .min_volt_shift = 0, > > + .max_volt_shift = 0, > > + .proc_max_volt = 1150000, > > + .sram_min_volt = 0, > > + .sram_max_volt = 0, > > + .need_voltage_tracking = false, > > +}; > > + > > +static const struct of_device_id mtk_ccifreq_machines[] = { > > + { .compatible = "mediatek,mt8183-cci", .data = > > &mtk_platform_data }, > > + { }, > > +}; > > +MODULE_DEVICE_TABLE(of, mtk_ccifreq_machines); > > + > > +static struct platform_driver mtk_ccifreq_platdrv = { > > + .probe = mtk_ccifreq_probe, > > + .remove = mtk_ccifreq_remove, > > + .driver = { > > + .name = "mtk-ccifreq", > > + .of_match_table = of_match_ptr(mtk_ccifreq_machines), > > + }, > > +}; > > + > > +static int __init mtk_ccifreq_platdrv_init(void) > > +{ > > + return platform_driver_register(&mtk_ccifreq_platdrv); > > +} > > +module_init(mtk_ccifreq_platdrv_init) > > + > > +static void __exit mtk_ccifreq_platdrv_exit(void) > > +{ > > + platform_driver_unregister(&mtk_ccifreq_platdrv); > > +} > > +module_exit(mtk_ccifreq_platdrv_exit) > > + > > +MODULE_DESCRIPTION("Mediatek CCI devfreq driver"); > > +MODULE_AUTHOR("Jia-Wei Chang <jia-wei.chang@xxxxxxxxxxxx>"); > > +MODULE_LICENSE("GPL v2"); > > > >
