lkml.org 
[lkml]   [2022]   [Sep]   [19]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
SubjectRe: [RFC v4 08/10] pwm: dwc: add of/platform support
From
On 14/09/2022 17:47, Uwe Kleine-König wrote:
> On Tue, Aug 16, 2022 at 10:14:52PM +0100, Ben Dooks wrote:
>> The dwc pwm controller can be used in non-PCI systems, so allow
>> either platform or OF based probing.
>>
>> Signed-off-by: Ben Dooks <ben.dooks@sifive.com>
>> ---
>> v4:
>> - split the of code out of the core
>> - moved the compile test code earlier
>> - fixed review comments
>> - used NS_PER_SEC
>> - use devm_clk_get_enabled
>> v3:
>> - changed compatible name
>> ---
>> drivers/pwm/Kconfig | 9 +++++
>> drivers/pwm/Makefile | 1 +
>> drivers/pwm/pwm-dwc-of.c | 78 ++++++++++++++++++++++++++++++++++++++++
>> 3 files changed, 88 insertions(+)
>> create mode 100644 drivers/pwm/pwm-dwc-of.c
>>
>> diff --git a/drivers/pwm/Kconfig b/drivers/pwm/Kconfig
>> index a9f1c554db2b..f1735653365f 100644
>> --- a/drivers/pwm/Kconfig
>> +++ b/drivers/pwm/Kconfig
>> @@ -192,6 +192,15 @@ config PWM_DWC_PCI
>> To compile this driver as a module, choose M here: the module
>> will be called pwm-dwc-pci.
>>
>> +config PWM_DWC_OF
>> + tristate "DesignWare PWM Controller (OF bus)
>> + depends on PWM_DWC && OF
>> + help
>> + PWM driver for Synopsys DWC PWM Controller on an OF bus.
>> +
>> + To compile this driver as a module, choose M here: the module
>> + will be called pwm-dwc-of.
>> +
>> config PWM_EP93XX
>> tristate "Cirrus Logic EP93xx PWM support"
>> depends on ARCH_EP93XX || COMPILE_TEST
>> diff --git a/drivers/pwm/Makefile b/drivers/pwm/Makefile
>> index a70d36623129..d1fd1641f077 100644
>> --- a/drivers/pwm/Makefile
>> +++ b/drivers/pwm/Makefile
>> @@ -15,6 +15,7 @@ obj-$(CONFIG_PWM_CLPS711X) += pwm-clps711x.o
>> obj-$(CONFIG_PWM_CRC) += pwm-crc.o
>> obj-$(CONFIG_PWM_CROS_EC) += pwm-cros-ec.o
>> obj-$(CONFIG_PWM_DWC) += pwm-dwc.o
>> +obj-$(CONFIG_PWM_DWC_OF) += pwm-dwc-of.o
>> obj-$(CONFIG_PWM_DWC_PCI) += pwm-dwc-pci.o
>> obj-$(CONFIG_PWM_EP93XX) += pwm-ep93xx.o
>> obj-$(CONFIG_PWM_FSL_FTM) += pwm-fsl-ftm.o
>> diff --git a/drivers/pwm/pwm-dwc-of.c b/drivers/pwm/pwm-dwc-of.c
>> new file mode 100644
>> index 000000000000..d18fac287325
>> --- /dev/null
>> +++ b/drivers/pwm/pwm-dwc-of.c
>> @@ -0,0 +1,78 @@
>> +// SPDX-License-Identifier: GPL-2.0
>> +/*
>> + * DesignWare PWM Controller driver OF
>> + *
>> + * Copyright (C) 2022 SiFive, Inc.
>> + */
>> +
>> +#include <linux/bitops.h>
>> +#include <linux/export.h>
>> +#include <linux/kernel.h>
>> +#include <linux/module.h>
>> +#include <linux/clk.h>
>> +#include <linux/platform_device.h>
>> +#include <linux/pm_runtime.h>
>> +#include <linux/pwm.h>
>> +#include <linux/io.h>
>> +
>> +#include "pwm-dwc.h"
>> +
>> +static int dwc_pwm_plat_probe(struct platform_device *pdev)
>> +{
>> + struct device *dev = &pdev->dev;
>> + struct dwc_pwm *dwc;
>> + int ret;
>> +
>> + dwc = dwc_pwm_alloc(dev);
>> + if (!dwc)
>> + return -ENOMEM;
>> +
>> + dwc->base = devm_platform_ioremap_resource(pdev, 0);
>> + if (IS_ERR(dwc->base))
>> + return dev_err_probe(dev, PTR_ERR(dwc->base),
>> + "failed to map IO\n");
>
> devm_platform_ioremap_resource() already emits an error message.

ok, fixed


>> +
>> + dwc->clk = devm_clk_get_enabled(dev, "timer");
>> + if (IS_ERR(dwc->clk))
>> + return dev_err_probe(dev, PTR_ERR(dwc->clk),
>> + "failed to get timer clock\n");
>> +
>> + clk_prepare_enable(dwc->clk);
>
> You don't need clk_prepare_enable() as you used devm_clk_get_enabled().
>
> (Otherwise, when keeping clk_prepare_enable() you need to check the
> return value.)

ok, fixed. I didn't notice that when changing to devm_

>
>> + dwc->clk_ns = NSEC_PER_SEC / clk_get_rate(dwc->clk);
>
> I think I already pointed out this is non-optimal.
>
> Later you use clk_ns in __dwc_pwm_configure_timer():
>
> tmp = DIV_ROUND_CLOSEST_ULL(state->duty_cycle, dwc->clk_ns);
>
> So what you really want here is:
>
> tmp = DIV_ROUND_CLOSEST_ULL(state->duty_cycle * clk_get_rate(dwc->clk), NSEC_PER_SEC);
>
> With for example clk_get_rate(dwc->clk) = 201171875 and duty_cycle =
> 4096 you now get clk_ns = 4 (while the exact value is 4.97087..), tmp =
> 1024, while the exact value is 824.
>
> So the idea is to add a clkrate member to the private driver struct, let
> it default to 100000000 for the pci case and use the line I suggested
> above.

ok, will consider keeping the rate in hz and modifiying the pci
version to use 10 * NSEC_PER_SEC for the rate.

I've been trying to avoid changing the pci case, but I think for
anything with a clk pointer we should be doing clk_get_rate in the
calc code

>
>> +
>> + ret = devm_pwmchip_add(dev, &dwc->chip);
>> + if (ret)
>> + return ret;
>> +
>> + return 0;
>
> This is equivalent to
>
> return ret;

Will sort that out.


>> +}
>> +
>> +static int dwc_pwm_plat_remove(struct platform_device *pdev)
>> +{
>> + struct dwc_pwm *dwc = platform_get_drvdata(pdev);
>> +
>> + clk_disable_unprepare(dwc->clk);
>> + return 0;
>> +}
>
> When dropping clk_prepare_enable() the .remove callback can go away,
> too.

thanks, spotted that one a while ago.

>> +
>> +static const struct of_device_id dwc_pwm_dt_ids[] = {
>> + { .compatible = "snps,dw-apb-timers-pwm2" },
>> + { },
>> +};
>> +MODULE_DEVICE_TABLE(of, dwc_pwm_dt_ids);
>> +
>> +static struct platform_driver dwc_pwm_plat_driver = {
>> + .driver = {
>> + .name = "dwc-pwm",
>> + .of_match_table = dwc_pwm_dt_ids,
>> + },
>> + .probe = dwc_pwm_plat_probe,
>> + .remove = dwc_pwm_plat_remove,
>> +};
>> +
>> +module_platform_driver(dwc_pwm_plat_driver);
>> +
>> +MODULE_ALIAS("platform:dwc-pwm-of");
>> +MODULE_AUTHOR("Ben Dooks <ben.dooks@sifive.com>");
>> +MODULE_DESCRIPTION("DesignWare PWM Controller");
>> +MODULE_LICENSE("GPL");
>
> Best regards
> Uwe
>

\
 
 \ /
  Last update: 2022-09-20 00:31    [W:0.068 / U:0.368 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site