linux-watchdog.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Guenter Roeck <linux@roeck-us.net>
To: Simon Arlott <simon@fire.lp0.eu>, linux-watchdog@vger.kernel.org
Cc: "devicetree@vger.kernel.org" <devicetree@vger.kernel.org>,
	Ralf Baechle <ralf@linux-mips.org>,
	Thomas Gleixner <tglx@linutronix.de>,
	Jason Cooper <jason@lakedaemon.net>,
	Marc Zyngier <marc.zyngier@arm.com>,
	Kevin Cernekee <cernekee@gmail.com>,
	Florian Fainelli <f.fainelli@gmail.com>,
	Wim Van Sebroeck <wim@iguana.be>,
	Maxime Bizon <mbizon@freebox.fr>,
	Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
	linux-mips@linux-mips.org, Rob Herring <robh+dt@kernel.org>,
	Pawel Moll <pawel.moll@arm.com>,
	Mark Rutland <mark.rutland@arm.com>,
	Ian Campbell <ijc+devicetree@hellion.org.uk>,
	Kumar Gala <galak@codeaurora.org>,
	Jonas Gorski <jogo@openwrt.org>
Subject: Re: [PATCH (v2) 7/10] watchdog: bcm63xx_wdt: Add get_timeleft function
Date: Tue, 24 Nov 2015 18:51:00 -0800	[thread overview]
Message-ID: <56552214.2050808@roeck-us.net> (raw)
In-Reply-To: <a1461a17c94353f38316d2c6ae04d6b77c91bfd4@8b5064a13e22126c1b9329f0dc35b8915774b7c3.invalid>

On 11/24/2015 02:15 PM, Simon Arlott wrote:
> Return the remaining time from the hardware control register.
>
> Warn when the device is registered if the hardware watchdog is currently
> running and report the remaining time left.

This is really two logical changes, isn't it ?

Nice trick to figure out if the watchdog is running.

What is the impact ? Will this result in interrupts ?
If so, would it make sense to _not_ reset the system after a timeout
in this case, but to keep pinging the watchdog while the watchdog device
is not open ?

Thanks,
Guenter

>
> Signed-off-by: Simon Arlott <simon@fire.lp0.eu>
> ---
> Changed "if (timeleft > 0)" to "if (hw->running)" when checking if a
> warning should be printed, in case the time left is truncated down to
> 0 seconds.
>
>   drivers/watchdog/bcm63xx_wdt.c | 37 +++++++++++++++++++++++++++++++++++++
>   1 file changed, 37 insertions(+)
>
> diff --git a/drivers/watchdog/bcm63xx_wdt.c b/drivers/watchdog/bcm63xx_wdt.c
> index 3c7667a..9d099e0 100644
> --- a/drivers/watchdog/bcm63xx_wdt.c
> +++ b/drivers/watchdog/bcm63xx_wdt.c
> @@ -14,6 +14,7 @@
>   #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
>
>   #include <linux/clk.h>
> +#include <linux/delay.h>
>   #include <linux/errno.h>
>   #include <linux/io.h>
>   #include <linux/kernel.h>
> @@ -75,6 +76,19 @@ static int bcm63xx_wdt_stop(struct watchdog_device *wdd)
>   	return 0;
>   }
>
> +static unsigned int bcm63xx_wdt_get_timeleft(struct watchdog_device *wdd)
> +{
> +	struct bcm63xx_wdt_hw *hw = watchdog_get_drvdata(wdd);
> +	unsigned long flags;
> +	u32 val;
> +
> +	raw_spin_lock_irqsave(&hw->lock, flags);
> +	val = __raw_readl(hw->regs + WDT_CTL_REG);
> +	val /= hw->clock_hz;
> +	raw_spin_unlock_irqrestore(&hw->lock, flags);
> +	return val;
> +}
> +
>   static int bcm63xx_wdt_set_timeout(struct watchdog_device *wdd,
>   	unsigned int timeout)
>   {
> @@ -130,6 +144,7 @@ static struct watchdog_ops bcm63xx_wdt_ops = {
>   	.owner = THIS_MODULE,
>   	.start = bcm63xx_wdt_start,
>   	.stop = bcm63xx_wdt_stop,
> +	.get_timeleft = bcm63xx_wdt_get_timeleft,
>   	.set_timeout = bcm63xx_wdt_set_timeout,
>   };
>
> @@ -144,6 +159,8 @@ static int bcm63xx_wdt_probe(struct platform_device *pdev)
>   	struct bcm63xx_wdt_hw *hw;
>   	struct watchdog_device *wdd;
>   	struct resource *r;
> +	u32 timeleft1, timeleft2;
> +	unsigned int timeleft;
>   	int ret;
>
>   	hw = devm_kzalloc(&pdev->dev, sizeof(*hw), GFP_KERNEL);
> @@ -197,6 +214,23 @@ static int bcm63xx_wdt_probe(struct platform_device *pdev)
>   	watchdog_init_timeout(wdd, 0, &pdev->dev);
>   	watchdog_set_nowayout(wdd, nowayout);
>
> +	/* Compare two reads of the time left value, 2 clock ticks apart */
> +	rmb();
> +	timeleft1 = __raw_readl(hw->regs + WDT_CTL_REG);
> +	udelay(DIV_ROUND_UP(1000000, hw->clock_hz / 2));
> +	/* Ensure the register is read twice */
> +	rmb();
> +	timeleft2 = __raw_readl(hw->regs + WDT_CTL_REG);
> +
> +	/* If the time left is changing, the watchdog is running */
> +	if (timeleft1 != timeleft2) {
> +		hw->running = true;
> +		timeleft = bcm63xx_wdt_get_timeleft(wdd);
> +	} else {
> +		hw->running = false;
> +		timeleft = 0;
> +	}
> +
>   	ret = bcm63xx_timer_register(TIMER_WDT_ID, bcm63xx_wdt_isr, wdd);
>   	if (ret < 0) {
>   		dev_err(&pdev->dev, "failed to register wdt timer isr\n");
> @@ -214,6 +248,8 @@ static int bcm63xx_wdt_probe(struct platform_device *pdev)
>   		dev_name(wdd->dev), hw->regs,
>   		wdd->timeout, wdd->max_timeout);
>
> +	if (hw->running)
> +		dev_alert(wdd->dev, "running, reboot in %us\n", timeleft);
>   	return 0;
>
>   unregister_timer:
> @@ -255,6 +291,7 @@ module_platform_driver(bcm63xx_wdt_driver);
>
>   MODULE_AUTHOR("Miguel Gaio <miguel.gaio@efixo.com>");
>   MODULE_AUTHOR("Florian Fainelli <florian@openwrt.org>");
> +MODULE_AUTHOR("Simon Arlott");
>   MODULE_DESCRIPTION("Driver for the Broadcom BCM63xx SoC watchdog");
>   MODULE_LICENSE("GPL");
>   MODULE_ALIAS("platform:bcm63xx-wdt");
>


  parent reply	other threads:[~2015-11-25  2:51 UTC|newest]

Thread overview: 53+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-11-21 19:02 [PATCH 1/4] clocksource: Add brcm,bcm6345-timer device tree binding Simon Arlott
2015-11-21 19:03 ` [PATCH 2/4] MIPS: bmips: Add bcm6345-l2-timer interrupt controller Simon Arlott
2015-11-21 19:04 ` [PATCH 3/4] watchdog: Add brcm,bcm6345-wdt device tree binding Simon Arlott
2015-11-22 22:13   ` Rob Herring
2015-11-21 19:05 ` [PATCH 4/4] MIPS: bmips: Convert bcm63xx_wdt to use WATCHDOG_CORE Simon Arlott
2015-11-21 21:32   ` Guenter Roeck
2015-11-21 21:44     ` Simon Arlott
2015-11-22  2:32       ` Guenter Roeck
2015-11-22 14:02         ` [PATCH 4/10] (Was: [PATCH 4/4]) " Simon Arlott
2015-11-22 14:05           ` [PATCH 4/10] watchdog: bcm63xx_wdt: Handle hardware interrupt and remove software timer Simon Arlott
2015-11-24 18:21             ` Guenter Roeck
2015-11-25 20:14               ` Jonas Gorski
2015-11-25 20:28                 ` Simon Arlott
2015-11-25 22:33             ` [PATCH (v2) " Simon Arlott
2015-11-22 14:06           ` [PATCH 5/10] watchdog: bcm63xx_wdt: Use WATCHDOG_CORE Simon Arlott
2015-11-25  2:44             ` Guenter Roeck
2015-11-25 13:02               ` Simon Arlott
2015-11-25 14:10                 ` Guenter Roeck
2015-11-25 19:43                   ` Simon Arlott
2015-11-25 22:40               ` [PATCH (v3) 5/11] " Simon Arlott
2015-11-22 14:07           ` [PATCH 6/10] watchdog: bcm63xx_wdt: Obtain watchdog clock HZ from "periph" clk Simon Arlott
2015-11-23 15:02             ` Jonas Gorski
2015-11-23 18:19               ` Florian Fainelli
2015-11-23 19:00                 ` Simon Arlott
2015-11-24 22:12             ` [PATCH (v2) " Simon Arlott
2015-11-24 22:42               ` Florian Fainelli
2015-11-25 22:47                 ` [PATCH (v3) 6/11] " Simon Arlott
2015-11-22 14:09           ` [PATCH 7/10] watchdog: bcm63xx_wdt: Add get_timeleft function Simon Arlott
2015-11-24 22:15             ` [PATCH (v2) " Simon Arlott
2015-11-24 22:43               ` Florian Fainelli
2015-11-25  2:51               ` Guenter Roeck [this message]
2015-11-25  8:17                 ` Simon Arlott
2015-11-25 22:50                   ` [PATCH (v3) 7/11] " Simon Arlott
2015-11-25 22:54                     ` [PATCH (v4) " Simon Arlott
2015-11-25 22:57                       ` [PATCH (v4) 8/11] watchdog: bcm63xx_wdt: Warn if the watchdog is currently running Simon Arlott
2015-11-22 14:11           ` [PATCH 8/10] watchdog: bcm63xx_wdt: Remove dependency on mach-bcm63xx functions/defines Simon Arlott
2015-11-22 14:12           ` [PATCH 9/10] watchdog: bcm63xx_wdt: Use bcm63xx_timer interrupt directly Simon Arlott
2015-11-25 23:03             ` [PATCH (v2) 10/11] " Simon Arlott
2015-11-22 14:14           ` [PATCH 10/10] watchdog: bcm63xx_wdt: Use brcm,bcm6345-wdt device tree binding Simon Arlott
2015-11-25 23:09             ` [PATCH (v2) 11/11] " Simon Arlott
2015-11-22 22:12 ` [PATCH 1/4] clocksource: Add brcm,bcm6345-timer " Rob Herring
2015-11-23 15:33 ` Jonas Gorski
2015-11-23 18:55   ` [PATCH (v2) 1/10] " Simon Arlott
2015-11-23 18:57     ` [PATCH (v2) 2/10] MIPS: bmips: Add bcm6345-l2-timer interrupt controller Simon Arlott
2015-11-24 22:10       ` [PATCH (v3) " Simon Arlott
2015-11-24 22:36         ` Florian Fainelli
2015-11-26 22:32           ` [PATCH (v4) 2/11] " Simon Arlott
2015-11-27  8:37             ` Thomas Gleixner
2015-11-28 12:26               ` [PATCH (v5) 3/11] " Simon Arlott
2015-12-01  0:22                 ` Guenter Roeck
     [not found]                   ` <256BE4660FB03085.33070B6A-E0AC-4643-92B8-4FD874210CD9@mail.outlook.com>
2016-05-09 13:06                     ` Guenter Roeck
2016-05-11  6:40                       ` [PATCH 2/4] " Álvaro Fernández Rojas
2015-11-25  3:05     ` [PATCH (v2) 1/10] clocksource: Add brcm,bcm6345-timer device tree binding Rob Herring

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=56552214.2050808@roeck-us.net \
    --to=linux@roeck-us.net \
    --cc=cernekee@gmail.com \
    --cc=devicetree@vger.kernel.org \
    --cc=f.fainelli@gmail.com \
    --cc=galak@codeaurora.org \
    --cc=ijc+devicetree@hellion.org.uk \
    --cc=jason@lakedaemon.net \
    --cc=jogo@openwrt.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mips@linux-mips.org \
    --cc=linux-watchdog@vger.kernel.org \
    --cc=marc.zyngier@arm.com \
    --cc=mark.rutland@arm.com \
    --cc=mbizon@freebox.fr \
    --cc=pawel.moll@arm.com \
    --cc=ralf@linux-mips.org \
    --cc=robh+dt@kernel.org \
    --cc=simon@fire.lp0.eu \
    --cc=tglx@linutronix.de \
    --cc=wim@iguana.be \
    /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;
as well as URLs for NNTP newsgroup(s).