Message-ID: <20160216230241.GC2189@piout.net>
Date:	Wed, 17 Feb 2016 00:02:41 +0100
From:	Alexandre Belloni <alexandre.belloni@...e-electrons.com>
To:	Joshua Henderson <joshua.henderson@...rochip.com>
Cc:	linux-kernel@...r.kernel.org, linux-mips@...ux-mips.org,
	Alessandro Zummo <a.zummo@...ertech.it>,
	rtc-linux@...glegroups.com
Subject: Re: [PATCH 2/2] rtc: rtc-pic32: Add PIC32 real time clock driver

Hi,

On 01/02/2016 at 15:43:22 -0700, Joshua Henderson wrote :
> This drivers adds support for the PIC32 real time clock and calendar
> peripheral:
>      - reading and setting time
>      - alarms when connected to an IRQ
> 
> Signed-off-by: Joshua Henderson <joshua.henderson@...rochip.com>
> ---
>  drivers/rtc/Kconfig     |   10 ++
>  drivers/rtc/Makefile    |    1 +
>  drivers/rtc/rtc-pic32.c |  450 +++++++++++++++++++++++++++++++++++++++++++++++

Nitpick: there are multiple alignment and blank lines issues that you
can spot using checkpatch --strict.

> diff --git a/drivers/rtc/Makefile b/drivers/rtc/Makefile
> index 62d61b2..20c9be5 100644
> --- a/drivers/rtc/Makefile
> +++ b/drivers/rtc/Makefile
> @@ -162,3 +162,4 @@ obj-$(CONFIG_RTC_DRV_WM8350)	+= rtc-wm8350.o
>  obj-$(CONFIG_RTC_DRV_X1205)	+= rtc-x1205.o
>  obj-$(CONFIG_RTC_DRV_XGENE)	+= rtc-xgene.o
>  obj-$(CONFIG_RTC_DRV_ZYNQMP)	+= rtc-zynqmp.o
> +obj-$(CONFIG_RTC_DRV_PIC32)	+= rtc-pic32.o

This list has to be ordered alphabetically.

> diff --git a/drivers/rtc/rtc-pic32.c b/drivers/rtc/rtc-pic32.c
> new file mode 100644
> index 0000000..7c46ccb
> --- /dev/null
> +++ b/drivers/rtc/rtc-pic32.c
> @@ -0,0 +1,450 @@
> +/*
> + * PIC32 RTC driver
> + *
> + * Joshua Henderson <joshua.henderson@...rochip.com>
> + * Copyright (C) 2016 Microchip Technology Inc.  All rights reserved.
> + *
> + * This program is free software; you can distribute it and/or modify it
> + * under the terms of the GNU General Public License (Version 2) as
> + * published by the Free Software Foundation.
> + *

This specify GPL v2 but you use MODULE_LICENSE("GPL")

> + * This program is distributed in the hope it will be useful, but WITHOUT
> + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
> + * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License
> + * for more details.
> + */
> +#include <linux/init.h>
> +#include <linux/module.h>
> +#include <linux/of.h>
> +#include <linux/platform_device.h>
> +#include <linux/io.h>
> +#include <linux/slab.h>
> +#include <linux/clk.h>
> +#include <linux/rtc.h>
> +#include <linux/bcd.h>
> +
> +#include <asm/mach-pic32/pic32.h>
> +
> +#define DRIVER_NAME "pic32-rtc"
> +
> +#define PIC32_RTCCON		(0x00)

You probably don't need those parenthesis.

> +#define PIC32_RTCCON_ON		BIT(15)
> +#define PIC32_RTCCON_SIDL	BIT(13)
> +#define PIC32_RTCCON_RTCCLKSEL	(3 << 9)
> +#define PIC32_RTCCON_RTCCLKON	BIT(6)
> +#define PIC32_RTCCON_RTCWREN	BIT(3)
> +#define PIC32_RTCCON_RTCSYNC	BIT(2)
> +#define PIC32_RTCCON_HALFSEC	BIT(1)
> +#define PIC32_RTCCON_RTCOE	BIT(0)
> +
> +#define PIC32_RTCALRM		(0x10)

ditto

> +#define PIC32_RTCALRM_ALRMEN	BIT(15)
> +#define PIC32_RTCALRM_CHIME	BIT(14)
> +#define PIC32_RTCALRM_PIV	BIT(13)
> +#define PIC32_RTCALRM_ALARMSYNC	BIT(12)
> +#define PIC32_RTCALRM_AMASK	(0x0F << 8)
> +#define PIC32_RTCALRM_ARPT	(0xFF << 0)

Nit: For those masks, you can use genmask() or directly 0x0F00 and 0xFF

> +struct pic32_rtc_dev {
> +	struct rtc_device	*rtc;
> +	void __iomem		*reg_base;
> +	struct clk		*clk;
> +	spinlock_t		flags_lock;

flags_lock is unnecessary, it is used only in one function that is
called only at probe time

> +	spinlock_t		alarm_lock;
> +	int			alarm_irq;
> +	bool			alarm_clk_enabled;
> +};
> +

[...]

> +static int pic32_rtc_gettime(struct device *dev, struct rtc_time *rtc_tm)
> +{
> +	struct pic32_rtc_dev *pdata = dev_get_drvdata(dev);
> +	void __iomem *base = pdata->reg_base;
> +	unsigned int have_retried = 0;
> +
> +	clk_enable(pdata->clk);
> +retry_get_time:
> +	rtc_tm->tm_hour = readb(base + PIC32_RTCHOUR);
> +	rtc_tm->tm_min = readb(base + PIC32_RTCMIN);
> +	rtc_tm->tm_mon  = readb(base + PIC32_RTCMON);
> +	rtc_tm->tm_mday = readb(base + PIC32_RTCDAY);
> +	rtc_tm->tm_year = readb(base + PIC32_RTCYEAR);
> +	rtc_tm->tm_sec  = readb(base + PIC32_RTCSEC);
> +
> +	/*
> +	 * The only way to work out whether the system was mid-update
> +	 * when we read it is to check the second counter, and if it
> +	 * is zero, then we re-try the entire read.
> +	 */
> +
> +	if (rtc_tm->tm_sec == 0 && !have_retried) {
> +		have_retried = 1;
> +		goto retry_get_time;
> +	}

Please change that goto in a do while loop.

> +
> +	rtc_tm->tm_sec = bcd2bin(rtc_tm->tm_sec);
> +	rtc_tm->tm_min = bcd2bin(rtc_tm->tm_min);
> +	rtc_tm->tm_hour = bcd2bin(rtc_tm->tm_hour);
> +	rtc_tm->tm_mday = bcd2bin(rtc_tm->tm_mday);
> +	rtc_tm->tm_mon = bcd2bin(rtc_tm->tm_mon);
> +	rtc_tm->tm_year = bcd2bin(rtc_tm->tm_year);
> +
> +	rtc_tm->tm_year += 100;
> +
> +	dev_dbg(dev, "read time %04d.%02d.%02d %02d:%02d:%02d\n",
> +		1900 + rtc_tm->tm_year, rtc_tm->tm_mon, rtc_tm->tm_mday,
> +		rtc_tm->tm_hour, rtc_tm->tm_min, rtc_tm->tm_sec);
> +
> +	rtc_tm->tm_mon -= 1;

Maybe you should do that substraction with the assignment the only
drawback is that it changes the debug output.

> +
> +	clk_disable(pdata->clk);
> +	return rtc_valid_tm(rtc_tm);
> +}
> +

[...]

> +static int pic32_rtc_getalarm(struct device *dev, struct rtc_wkalrm *alrm)
> +{
> +	struct pic32_rtc_dev *pdata = dev_get_drvdata(dev);
> +	struct rtc_time *alm_tm = &alrm->time;
> +	void __iomem *base = pdata->reg_base;
> +	unsigned int alm_en;
> +
> +	clk_enable(pdata->clk);
> +	alm_tm->tm_sec  = readb(base + PIC32_ALRMSEC);
> +	alm_tm->tm_min  = readb(base + PIC32_ALRMMIN);
> +	alm_tm->tm_hour = readb(base + PIC32_ALRMHOUR);
> +	alm_tm->tm_mon  = readb(base + PIC32_ALRMMON);
> +	alm_tm->tm_mday = readb(base + PIC32_ALRMDAY);
> +	alm_tm->tm_year = readb(base + PIC32_ALRMYEAR);
> +
> +	alm_en = readb(base + PIC32_RTCALRM);
> +
> +	alrm->enabled = (alm_en & PIC32_RTCALRM_ALRMEN) ? 1 : 0;
> +
> +	dev_dbg(dev, "getalarm: %d, %04d.%02d.%02d %02d:%02d:%02d\n",
> +		alm_en,
> +		1900 + alm_tm->tm_year, alm_tm->tm_mon, alm_tm->tm_mday,
> +		alm_tm->tm_hour, alm_tm->tm_min, alm_tm->tm_sec);
> +
> +	alm_tm->tm_sec = bcd2bin(alm_tm->tm_sec);
> +	alm_tm->tm_min = bcd2bin(alm_tm->tm_min);
> +	alm_tm->tm_hour = bcd2bin(alm_tm->tm_hour);
> +	alm_tm->tm_mday = bcd2bin(alm_tm->tm_mday);
> +	alm_tm->tm_mon = bcd2bin(alm_tm->tm_mon);
> +	alm_tm->tm_mon -= 1;
You could merge both lines.

> +	alm_tm->tm_year = bcd2bin(alm_tm->tm_year);
> +
> +	clk_disable(pdata->clk);
> +	return 0;
> +}
> +
> +static int pic32_rtc_setalarm(struct device *dev, struct rtc_wkalrm *alrm)
> +{
> +	struct pic32_rtc_dev *pdata = dev_get_drvdata(dev);
> +	struct rtc_time *tm = &alrm->time;
> +	void __iomem *base = pdata->reg_base;
> +
> +	clk_enable(pdata->clk);
> +	dev_dbg(dev, "setalarm: %d, %04d.%02d.%02d %02d:%02d:%02d\n",
> +		alrm->enabled,
> +		1900 + tm->tm_year, tm->tm_mon + 1, tm->tm_mday,
> +		tm->tm_hour, tm->tm_min, tm->tm_sec);
> +
> +	writel(0x00, base + PIC32_ALRMTIME);
> +	writel(0x00, base + PIC32_ALRMDATE);
> +
> +	if (tm->tm_sec < 60 && tm->tm_sec >= 0)
> +		writeb(bin2bcd(tm->tm_sec), base + PIC32_ALRMSEC);
> +
> +	if (tm->tm_min < 60 && tm->tm_min >= 0)
> +		writeb(bin2bcd(tm->tm_min), base + PIC32_ALRMMIN);
> +
> +	if (tm->tm_hour < 24 && tm->tm_hour >= 0)
> +		writeb(bin2bcd(tm->tm_hour), base + PIC32_ALRMHOUR);
> +

Those three tests are unnecessary because rtc_valid_tm() is called
before ops->set_alarm.

> +	pic32_rtc_setaie(dev, alrm->enabled);
> +
> +	clk_disable(pdata->clk);
> +	return 0;
> +}
> +

[...]


> +	pic32_rtc_gettime(&pdev->dev, &rtc_tm);
> +
> +	if (rtc_valid_tm(&rtc_tm)) {
> +		rtc_tm.tm_year	= 100;
> +		rtc_tm.tm_mon	= 0;
> +		rtc_tm.tm_mday	= 1;
> +		rtc_tm.tm_hour	= 0;
> +		rtc_tm.tm_min	= 0;
> +		rtc_tm.tm_sec	= 0;
> +
> +		pic32_rtc_settime(NULL, &rtc_tm);
> +
> +		dev_warn(&pdev->dev,
> +			"warning: invalid RTC value so initializing it\n");

This is a bad idea. If you do that, userspace has no way of knowing
whether the date is valid or not.

[...]

> +static const struct of_device_id pic32_rtc_dt_ids[] = {
> +	{ .compatible = "microchip,pic32mzda-rtc" },
> +	{ /* sentinel */ }
> +};
> +MODULE_DEVICE_TABLE(of, pic32_rtc_dt_ids);

This compatible string needs to be documented. Was that patch 1/2?

> +
> +static struct platform_driver pic32_rtc_driver = {
> +	.probe		= pic32_rtc_probe,
> +	.remove		= pic32_rtc_remove,
> +	.driver		= {
> +		.name	= DRIVER_NAME,
> +		.owner	= THIS_MODULE,
> +		.of_match_table	= of_match_ptr(pic32_rtc_dt_ids),
> +	},
> +};
> +module_platform_driver(pic32_rtc_driver);
> +
> +MODULE_DESCRIPTION("Microchip PIC32 RTC Driver");
> +MODULE_AUTHOR("Joshua Henderson <joshua.henderson@...rochip.com>");
> +MODULE_LICENSE("GPL");
> +MODULE_ALIAS("platform:" DRIVER_NAME);

I think this driver is DT only, you can get rid of MODULE_ALIAS and
DRIVER_NAME


-- 
Alexandre Belloni, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com

Powered by blists - more mailing lists