Best Regards! Anson Huang
> -----Original Message----- > From: Viresh Kumar [mailto:viresh.ku...@linaro.org] > Sent: 2018年11月21日 13:21 > To: Anson Huang <anson.hu...@nxp.com> > Cc: rui.zh...@intel.com; edubez...@gmail.com; daniel.lezc...@linaro.org; > linux...@vger.kernel.org; linux-kernel@vger.kernel.org; > l.st...@pengutronix.de; dl-linux-imx <linux-...@nxp.com> > Subject: Re: [PATCH V4] thermal: imx: fix for dependency on cpu-freq > > On 21-11-18, 05:08, Anson Huang wrote: > > The thermal driver is a standalone driver for monitoring SoC > > temperature by enabling thermal sensor, so it can be enabled even when > > CONFIG_CPU_FREQ is NOT set. So remove the dependency with > CPU_THERMAL. > > > > Introduce dummy function of legacy cooling register/unregister to make > > thermal driver probe successfully when CONFIG_CPU_FREQ is NOT set. > > > > Signed-off-by: Anson Huang <anson.hu...@nxp.com> > > --- > > changes since V3: > > rename the label of "cpufreq_put" with "legacy_cleanup". > > drivers/thermal/Kconfig | 2 +- > > drivers/thermal/imx_thermal.c | 44 > > +++++++++++++++++++++++++++++++------------ > > 2 files changed, 33 insertions(+), 13 deletions(-) > > > > diff --git a/drivers/thermal/Kconfig b/drivers/thermal/Kconfig index > > 5422523..93bd3bb 100644 > > --- a/drivers/thermal/Kconfig > > +++ b/drivers/thermal/Kconfig > > @@ -212,7 +212,7 @@ config HISI_THERMAL > > > > config IMX_THERMAL > > tristate "Temperature sensor driver for Freescale i.MX SoCs" > > - depends on (ARCH_MXC && CPU_THERMAL) || COMPILE_TEST > > + depends on ARCH_MXC || COMPILE_TEST > > depends on NVMEM || !NVMEM > > depends on MFD_SYSCON > > depends on OF > > diff --git a/drivers/thermal/imx_thermal.c > > b/drivers/thermal/imx_thermal.c index 1566154..328ee05 100644 > > --- a/drivers/thermal/imx_thermal.c > > +++ b/drivers/thermal/imx_thermal.c > > @@ -648,15 +648,24 @@ static const struct of_device_id > > of_imx_thermal_match[] = { }; MODULE_DEVICE_TABLE(of, > > of_imx_thermal_match); > > > > +#ifdef CONFIG_CPU_FREQ > > /* > > * Create cooling device in case no #cooling-cells property is available in > > * CPU node > > */ > > static int imx_thermal_register_legacy_cooling(struct > > imx_thermal_data *data) { > > - struct device_node *np = of_get_cpu_node(data->policy->cpu, NULL); > > + struct device_node *np; > > int ret; > > > > + data->policy = cpufreq_cpu_get(0); > > + if (!data->policy) { > > + pr_debug("%s: CPUFreq policy not found\n", __func__); > > + return -EPROBE_DEFER; > > + } > > + > > + np = of_get_cpu_node(data->policy->cpu, NULL); > > + > > if (!np || !of_find_property(np, "#cooling-cells", NULL)) { > > data->cdev = cpufreq_cooling_register(data->policy); > > if (IS_ERR(data->cdev)) { > > @@ -669,6 +678,22 @@ static int > imx_thermal_register_legacy_cooling(struct imx_thermal_data *data) > > return 0; > > } > > > > +static void imx_thermal_unregister_legacy_cooling(struct > > +imx_thermal_data *data) { > > + cpufreq_cooling_unregister(data->cdev); > > + cpufreq_cpu_put(data->policy); > > +} > > +#else > > +static inline int imx_thermal_register_legacy_cooling(struct > > +imx_thermal_data *data) { > > + return 0; > > +} > > + > > +static inline void imx_thermal_unregister_legacy_cooling(struct > > +imx_thermal_data *data) { } #endif > > + > > static int imx_thermal_probe(struct platform_device *pdev) { > > struct imx_thermal_data *data; > > @@ -743,13 +768,9 @@ static int imx_thermal_probe(struct > platform_device *pdev) > > regmap_write(map, data->socdata->sensor_ctrl + REG_SET, > > data->socdata->power_down_mask); > > > > - data->policy = cpufreq_cpu_get(0); > > - if (!data->policy) { > > - pr_debug("%s: CPUFreq policy not found\n", __func__); > > - return -EPROBE_DEFER; > > - } > > - > > ret = imx_thermal_register_legacy_cooling(data); > > + if (ret == -EPROBE_DEFER) > > + return ret; > > if (ret) { > > Sorry for not noticing earlier, but wouldn't it be better to move the > EPROBE_DEFER check inside of this if block ? Otherwise we will have two > conditional blocks in the success (normal) case. Since there is similar case of DEFER PROBE for the case of imx_init_from_nvmem_cells check, should I create another patch of same fix for it in V5 patch set? Anson. > > Something like this: > > if (ret) { > if (ret == -EPROBE_DEFER) > return ret; > > dev_err(..); > ... > } > > > dev_err(&pdev->dev, > > "failed to register cpufreq cooling device: %d\n", > > ret); @@ > -762,7 > > +783,7 @@ static int imx_thermal_probe(struct platform_device *pdev) > > if (ret != -EPROBE_DEFER) > > dev_err(&pdev->dev, > > "failed to get thermal clk: %d\n", ret); > > - goto cpufreq_put; > > + goto legacy_cleanup; > > } > > > > /* > > @@ -775,7 +796,7 @@ static int imx_thermal_probe(struct platform_device > *pdev) > > ret = clk_prepare_enable(data->thermal_clk); > > if (ret) { > > dev_err(&pdev->dev, "failed to enable thermal clk: %d\n", ret); > > - goto cpufreq_put; > > + goto legacy_cleanup; > > } > > > > data->tz = thermal_zone_device_register("imx_thermal_zone", > > @@ -829,9 +850,8 @@ static int imx_thermal_probe(struct platform_device > *pdev) > > thermal_zone_device_unregister(data->tz); > > clk_disable: > > clk_disable_unprepare(data->thermal_clk); > > -cpufreq_put: > > - cpufreq_cooling_unregister(data->cdev); > > - cpufreq_cpu_put(data->policy); > > +legacy_cleanup: > > + imx_thermal_unregister_legacy_cooling(data); > > > > return ret; > > } > > -- > > 2.7.4 > > -- > viresh