mail archive of the barebox mailing list
 help / color / mirror / Atom feed
From: "Clément Leger" <cleger@kalray.eu>
To: Ahmad Fatoum <a.fatoum@pengutronix.de>
Cc: Barebox List <barebox@lists.infradead.org>
Subject: Re: [PATCH 5/5] watchdog: k1c: Add k1c watchdog support
Date: Mon, 20 Jan 2020 17:08:56 +0100 (CET)	[thread overview]
Message-ID: <1043390847.13051146.1579536536042.JavaMail.zimbra@kalray.eu> (raw)
In-Reply-To: <f34584e3-f756-f756-f75f-39813587f40d@pengutronix.de>

Hi Ahmad,

----- On 20 Jan, 2020, at 16:10, Ahmad Fatoum a.fatoum@pengutronix.de wrote:

> On 1/15/20 11:26 AM, Clement Leger wrote:
>> Add a watchdog for k1c architecture based on core watchdog.
>> 
>> Signed-off-by: Clement Leger <cleger@kalray.eu>
>> ---
>>  arch/k1c/configs/generic_defconfig |   3 +
>>  drivers/watchdog/Kconfig           |   7 +++
>>  drivers/watchdog/Makefile          |   1 +
>>  drivers/watchdog/k1c_wdt.c         | 126 +++++++++++++++++++++++++++++++++++++
>>  4 files changed, 137 insertions(+)
>>  create mode 100644 drivers/watchdog/k1c_wdt.c
>> 
>> diff --git a/arch/k1c/configs/generic_defconfig
>> b/arch/k1c/configs/generic_defconfig
>> index 0821030f6..6c8d676df 100644
>> --- a/arch/k1c/configs/generic_defconfig
>> +++ b/arch/k1c/configs/generic_defconfig
>> @@ -5,6 +5,9 @@ CONFIG_CLOCKSOURCE_K1C=y
>>  CONFIG_CMD_CMP=y
>>  CONFIG_CMD_OF_DUMP=y
>>  CONFIG_CMD_POWEROFF=y
>> +CONFIG_CMD_WD=y
>>  CONFIG_CONSOLE_RATP=y
>>  CONFIG_DRIVER_SERIAL_NS16550=y
>>  CONFIG_PINCTRL_SINGLE=y
>> +CONFIG_WATCHDOG=y
>> +CONFIG_WATCHDOG_K1C=y
>> diff --git a/drivers/watchdog/Kconfig b/drivers/watchdog/Kconfig
>> index 45dd41a2a..17a850de6 100644
>> --- a/drivers/watchdog/Kconfig
>> +++ b/drivers/watchdog/Kconfig
>> @@ -70,6 +70,13 @@ config WATCHDOG_ORION
>>  	help
>>  	  Add support for watchdog on the Marvall Armada XP
>>  
>> +config WATCHDOG_K1C
>> +	bool "K1C Core watchdog"
>> +	select RESET_CONTROLLER
> 
> This seems only used for struct reset_control *rst, which you
> don't use anywhere.

Indeed.

> 
>> +	depends on K1C
>> +	help
>> +	  Add support for the K1C core watchdog.
>> +
>>  config WATCHDOG_BCM2835
>>  	bool "Watchdog for BCM283x SoCs"
>>  	depends on ARCH_BCM283X
>> diff --git a/drivers/watchdog/Makefile b/drivers/watchdog/Makefile
>> index 63efc2a87..bbf5e8767 100644
>> --- a/drivers/watchdog/Makefile
>> +++ b/drivers/watchdog/Makefile
>> @@ -8,6 +8,7 @@ obj-$(CONFIG_WATCHDOG_DW) += dw_wdt.o
>>  obj-$(CONFIG_WATCHDOG_JZ4740) += jz4740.o
>>  obj-$(CONFIG_WATCHDOG_IMX_RESET_SOURCE) += imxwd.o
>>  obj-$(CONFIG_WATCHDOG_IMX) += imxwd.o
>> +obj-$(CONFIG_WATCHDOG_K1C) += k1c_wdt.o
>>  obj-$(CONFIG_WATCHDOG_ORION) += orion_wdt.o
>>  obj-$(CONFIG_ARCH_BCM283X) += bcm2835_wdt.o
>>  obj-$(CONFIG_RAVE_SP_WATCHDOG) += rave-sp-wdt.o
>> diff --git a/drivers/watchdog/k1c_wdt.c b/drivers/watchdog/k1c_wdt.c
>> new file mode 100644
>> index 000000000..23b6cbf33
>> --- /dev/null
>> +++ b/drivers/watchdog/k1c_wdt.c
>> @@ -0,0 +1,126 @@
>> +// SPDX-License-Identifier: GPL-2.0-or-later
>> +/*
>> + * Copyright (C) 2018 Kalray Inc.
>> + */
>> +
>> +#include <common.h>
>> +#include <init.h>
>> +#include <io.h>
>> +#include <of.h>
>> +#include <restart.h>
>> +#include <watchdog.h>
>> +
>> +#include <linux/clk.h>
>> +#include <linux/err.h>
>> +
>> +#include <asm/sfr.h>
>> +
>> +#define to_dw_wdt(wdd)	container_of(wdd, struct dw_wdt, wdd)
> 
> used no where

Yes

> 
>> +
>> +struct k1c_wdt {
>> +	uint64_t clk_rate;
>> +	struct watchdog wdd;
>> +	struct restart_handler restart;
>> +	struct reset_control *rst;
>> +};
>> +
>> +static void k1c_watchdog_disable(void)
>> +{
>> +	k1c_sfr_clear_bit(K1C_SFR_TCR, K1C_SFR_TCR_WUI_SHIFT);
>> +	k1c_sfr_clear_bit(K1C_SFR_TCR, K1C_SFR_TCR_WCE_SHIFT);
>> +}
>> +
>> +static int k1c_wdt_set_timeout(struct watchdog *wdd, unsigned int timeout)
>> +{
>> +	struct k1c_wdt *wdt = container_of(wdd, struct k1c_wdt, wdd);
>> +	uint64_t cycle_timeout = wdt->clk_rate * timeout;
>> +
>> +	/* Disable watchdog */
>> +	if (timeout == 0) {
>> +		k1c_watchdog_disable();
>> +		return 0;
>> +	}
>> +
>> +	k1c_sfr_set(K1C_SFR_WDV, cycle_timeout);
>> +	k1c_sfr_set(K1C_SFR_WDR, 0);
>> +
>> +	/* Start watchdog counting */
>> +	k1c_sfr_set_bit(K1C_SFR_TCR, K1C_SFR_TCR_WUI_SHIFT);
>> +	k1c_sfr_set_bit(K1C_SFR_TCR, K1C_SFR_TCR_WCE_SHIFT);
>> +
>> +	return 0;
>> +}
>> +
>> +static void __noreturn k1c_wdt_restart_handle(struct restart_handler *rst)
>> +{
>> +	struct k1c_wdt *k1c_wdt;
>> +
>> +	k1c_wdt = container_of(rst, struct k1c_wdt, restart);
>> +
>> +	k1c_wdt->wdd.set_timeout(&k1c_wdt->wdd, 1);
> 
> Do you do the same under Linux as well? This means you can't differentiate
> between watchdog and normal resets.

Agreed, on linux we did not implemented it like that and you are right, the
reset cause will be the same.
I think I should remove this restart handler part (See my comment below). 

> 
>> +
>> +	mdelay(1000);
>> +
>> +	hang();
>> +}
>> +
>> +static int count;
>> +
>> +static int k1c_wdt_drv_probe(struct device_d *dev)
>> +{
>> +	struct watchdog *wdd;
>> +	struct clk *clk;
>> +	struct k1c_wdt *k1c_wdt;
>> +	int ret;
>> +
>> +	if (count != 0) {
>> +		dev_warn(dev, "Tried to register core watchdog twice\n");
>> +		return -EINVAL;
>> +	}
>> +	count++;
>> +
>> +	k1c_wdt = xzalloc(sizeof(*k1c_wdt));
>> +	clk = clk_get(dev, NULL);
>> +	if (IS_ERR(clk))
>> +		return PTR_ERR(clk);
>> +
>> +	k1c_wdt->clk_rate = clk_get_rate(clk);
>> +	clk_put(clk);
>> +
>> +	wdd = &k1c_wdt->wdd;
>> +	wdd->name = "k1c_wdt";
>> +	wdd->hwdev = dev;
>> +	wdd->set_timeout = k1c_wdt_set_timeout;
>> +
>> +	/* Be sure that interrupt is disabled */
>> +	k1c_sfr_clear_bit(K1C_SFR_TCR, K1C_SFR_TCR_WIE_SHIFT);
>> +
>> +	k1c_watchdog_disable();
>> +
>> +	ret = watchdog_register(wdd);
>> +	if (ret)
>> +		return -EINVAL;
>> +
>> +	k1c_wdt->restart.name = "k1c_wdt";
>> +	k1c_wdt->restart.priority = 50;
> 
> Maybe add a comment on why you set a lower than default priority?

Indeed.
We actually have another mean of resetting so this is not the preferred
way to restart. I'm not even sure if I should keep the restart handler.
Maybe I should left it out and use the reset register.

Thanks for your review

Clément

> 
>> +	k1c_wdt->restart.restart = k1c_wdt_restart_handle;
>> +
>> +	ret = restart_handler_register(&k1c_wdt->restart);
>> +	if (ret)
>> +		dev_warn(dev, "cannot register restart handler\n");
>> +
>> +	return 0;
>> +
>> +}
>> +
>> +static struct of_device_id k1c_wdt_of_match[] = {
>> +	{ .compatible = "kalray,k1c-core-watchdog", },
>> +	{ /* sentinel */ }
>> +};
>> +
>> +static struct driver_d k1c_wdt_driver = {
>> +	.name		= "k1c-wdt",
>> +	.probe		= k1c_wdt_drv_probe,
>> +	.of_compatible	= DRV_OF_COMPAT(k1c_wdt_of_match),
>> +};
>> +device_platform_driver(k1c_wdt_driver);
>> 
> 
> 
> --
> Pengutronix e.K.                           |                             |
> Steuerwalder Str. 21                       | https://www.pengutronix.de/ |
> 31137 Hildesheim, Germany                  | Phone: +49-5121-206917-0    |
> Amtsgericht Hildesheim, HRA 2686           | Fax:   +49-5121-206917-5555 |
> 
> _______________________________________________
> barebox mailing list
> barebox@lists.infradead.org
> http://lists.infradead.org/mailman/listinfo/barebox

_______________________________________________
barebox mailing list
barebox@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/barebox

  reply	other threads:[~2020-01-20 16:09 UTC|newest]

Thread overview: 22+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-01-15 10:26 [PATCH 0/5] Add support for Kalray k1c core Clement Leger
2020-01-15 10:26 ` [PATCH 1/5] k1c: Initial Kalray Coolidge (k1c) architecture support Clement Leger
2020-01-16  9:26   ` Sascha Hauer
2020-01-16 10:49     ` Clément Leger
2020-01-16 10:55       ` Sascha Hauer
2020-01-16 12:24         ` Clément Leger
2020-01-16 13:11           ` Sascha Hauer
2020-01-16 13:26             ` Clément Leger
2020-01-16 13:34               ` Sascha Hauer
2020-01-15 10:26 ` [PATCH 2/5] k1c: Add processor definitions Clement Leger
2020-01-15 10:26 ` [PATCH 3/5] k1c: Add support for device tree Clement Leger
2020-01-15 10:26 ` [PATCH 4/5] clocksource: k1c: Add k1c clocksource support Clement Leger
2020-01-15 10:26 ` [PATCH 5/5] watchdog: k1c: Add k1c watchdog support Clement Leger
2020-01-20 15:10   ` Ahmad Fatoum
2020-01-20 16:08     ` Clément Leger [this message]
2020-01-16  8:25 ` [PATCH 0/5] Add support for Kalray k1c core Sascha Hauer
2020-01-16  8:53   ` Clément Leger
2020-01-16  9:22     ` Sascha Hauer
2020-01-16  9:37       ` Clément Leger
2020-01-16  9:46         ` Sascha Hauer
2020-01-16  9:49 ` Roland Hieber
2020-01-16  9:53   ` Clément Leger

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=1043390847.13051146.1579536536042.JavaMail.zimbra@kalray.eu \
    --to=cleger@kalray.eu \
    --cc=a.fatoum@pengutronix.de \
    --cc=barebox@lists.infradead.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