Subject: Re: [PATCH] drivers/watchdog: Add sys files to modify watchdog
 behavior
To: Guenter Roeck <linux@roeck-us.net>
References: <20170601213521.58141-1-cbostic@linux.vnet.ibm.com>
 <20170601220358.GC8803@roeck-us.net>
Cc: wim@iguana.be, joel@jms.id.au, linux-kernel@vger.kernel.org,
        linux-watchdog@vger.kernel.org, andrew@aj.id.au
From: Christopher Bostic <cbostic@linux.vnet.ibm.com>
Date: Mon, 12 Jun 2017 15:30:19 -0500
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.11; rv:45.0)
 Gecko/20100101 Thunderbird/45.7.0
MIME-Version: 1.0
In-Reply-To: <20170601220358.GC8803@roeck-us.net>
Content-Type: text/plain; charset=windows-1252; format=flowed
Content-Transfer-Encoding: 7bit
Message-Id: <aa5b4d11-9c8e-50bf-677e-95bed56cdddf@linux.vnet.ibm.com>
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 6592
Lines: 186


On 6/1/17 5:03 PM, Guenter Roeck wrote:
> On Thu, Jun 01, 2017 at 04:35:21PM -0500, Christopher Bostic wrote:
>> Add files to enable/disable Aspeed watchdog:
>> * External signal after timeout
>> * Reset system after timeout
>>
> The subject line is heavily misleading.
>
>> Signed-off-by: Christopher Bostic <cbostic@linux.vnet.ibm.com>
>> ---
>>   .../ABI/testing/sysfs-platform-aspeed-wdt          | 40 +++++++++++
>>   drivers/watchdog/aspeed_wdt.c                      | 79 +++++++++++++++++++++-
>>   2 files changed, 118 insertions(+), 1 deletion(-)
>>   create mode 100644 Documentation/ABI/testing/sysfs-platform-aspeed-wdt
>>
>> diff --git a/Documentation/ABI/testing/sysfs-platform-aspeed-wdt b/Documentation/ABI/testing/sysfs-platform-aspeed-wdt
>> new file mode 100644
>> index 0000000..a582176
>> --- /dev/null
>> +++ b/Documentation/ABI/testing/sysfs-platform-aspeed-wdt
>> @@ -0,0 +1,40 @@
>> +What:           /sys/bus/platform/devices/1e785000.wdt/watchdog/watchdog0/
>> +		reset_system
>> +Date:           May 2017
>> +KernelVersion:  4.12
>> +Contact:        Christopher Bostic <cbostic@linux.vnet.ibm.com>
>> +Description:
>> +                Value indicates if a watchdog time out results in a reset of
>> +		system.  'off': no system reset on timeout.  'on' there will
>> +		be a system reset on timeout.
>> +
>> +                Default: off
>> +
>> +                - Enable 'reset system' on watchdog timeout:
>> +                echo on > /sys/bus/platform/devices/1e785000.wdt/watchdog/
>> +			watchdog0/reset_system;
>> +
>> +                - Disable 'reset system' on watchdog timeout:
>> +                echo off > /sys/bus/platform/devices/1e785000.wdt/watchdog/
>> +			watchdog0/reset_system;
>> +
>> +What:           /sys/bus/platform/devices/1e785000.wdt/watchdog/watchdog0/
>> +		ext_signal
>> +Date:           May 2017
>> +KernelVersion:  4.12
>> +Contact:        Christopher Bostic <cbostic@linux.vnet.ibm.com>
>> +Description:
>> +                Value indicates if a watchdog time out results in an active
>> +		high pulse sent out an output pin.  'off': No output pulse
>> +		generated on watchdog time out.  'on': output pulse
>> +		generated on watchdog time out.
>> +
>> +                Default: off
>> +
>> +                - Enable 'external signal' on watchdog timeout:
>> +                echo on > /sys/bus/platform/devices/1e785000.wdt/watchdog/
>> +			watchdog0/ext_signal;
>> +
>> +                - Disable 'external signal' on watchdog timeout:
>> +                echo off > /sys/bus/platform/devices/1e785000.wdt/watchdog/
>> +			watchdog0/ext_signal;
> I'll want to see some discussion on those, both for the values written (on/off
> vs. 0/1) and the action to be taken. Both seem to be system parameters rather
> than something to set from user space. Also, there is no explanation what else
> the watchdog should do on timeout if not to reset the system. If it doesn't do
> anything, why run it in the first place ?
Hi Guenter,

I'll be resubmitting with a change to remove the sysfs files and instead 
add a device tree attribute
that will be specific to the system affected.

Thanks
Chris


> Also, specifying absolute path names seems wrong. There is no real guarantee
> that the device is always watchdog0, that the path always includes
> "1e785000.wdt", or that the path name will always start with
> /sys/bus/platform/devices.
>
> Thanks,
> Guenter
>
>> diff --git a/drivers/watchdog/aspeed_wdt.c b/drivers/watchdog/aspeed_wdt.c
>> index 1c65258..66b27e8 100644
>> --- a/drivers/watchdog/aspeed_wdt.c
>> +++ b/drivers/watchdog/aspeed_wdt.c
>> @@ -136,6 +136,83 @@ static int aspeed_wdt_restart(struct watchdog_device *wdd,
>>   	.identity	= KBUILD_MODNAME,
>>   };
>>   
>> +static ssize_t aspeed_wdt_cntl_get_state(struct device *dev, char *buf,
>> +					uint32_t mask)
>> +{
>> +	struct platform_device *pdev = to_platform_device(dev);
>> +	struct aspeed_wdt *wdt = platform_get_drvdata(pdev);
>> +	int ret;
>> +
>> +	if (readl(wdt->base + WDT_CTRL) & mask)
>> +		ret = sprintf(buf, "on\n");
>> +	else
>> +		ret = sprintf(buf, "off\n");
>> +
>> +	return ret;
>> +}
>> +
>> +static ssize_t aspeed_wdt_cntl_set_state(struct device *dev, const char *buf,
>> +					size_t count, uint32_t mask)
>> +{
>> +	struct platform_device *pdev = to_platform_device(dev);
>> +	struct aspeed_wdt *wdt = platform_get_drvdata(pdev);
>> +
>> +	if (!strncmp(buf, "on", strlen("on"))) {
>> +		wdt->ctrl |= mask;
>> +		writel(wdt->ctrl, wdt->base + WDT_CTRL);
>> +	} else if (!strncmp(buf, "off", strlen("off"))) {
>> +		wdt->ctrl &= ~mask;
>> +		writel(wdt->ctrl, wdt->base + WDT_CTRL);
>> +	} else {
>> +		dev_warn(dev, "Unknown reset system mode command: [%s]\n", buf);
>> +		return -EINVAL;
>> +	}
>> +
>> +	return count;
>> +}
>> +
>> +static ssize_t aspeed_wdt_reset_system_store(struct device *dev,
>> +					struct device_attribute *attr,
>> +					const char *buf, size_t count)
>> +{
>> +	return aspeed_wdt_cntl_set_state(dev, buf, count,
>> +					WDT_CTRL_RESET_SYSTEM);
>> +}
>> +
>> +static ssize_t aspeed_wdt_reset_system_show(struct device *dev,
>> +					struct device_attribute *attr,
>> +					char *buf)
>> +{
>> +	return aspeed_wdt_cntl_get_state(dev, buf, WDT_CTRL_RESET_SYSTEM);
>> +}
>> +
>> +static DEVICE_ATTR(reset_system, 0644,
>> +		aspeed_wdt_reset_system_show, aspeed_wdt_reset_system_store);
>> +
>> +static ssize_t aspeed_wdt_ext_signal_store(struct device *dev,
>> +					struct device_attribute *attr,
>> +					const char *buf, size_t count)
>> +{
>> +	return aspeed_wdt_cntl_set_state(dev, buf, count, WDT_CTRL_WDT_EXT);
>> +}
>> +
>> +static ssize_t aspeed_wdt_ext_signal_show(struct device *dev,
>> +					struct device_attribute *attr,
>> +					char *buf)
>> +{
>> +	return aspeed_wdt_cntl_get_state(dev, buf, WDT_CTRL_WDT_EXT);
>> +}
>> +
>> +static DEVICE_ATTR(ext_signal, 0644,
>> +		aspeed_wdt_ext_signal_show, aspeed_wdt_ext_signal_store);
>> +
>> +static struct attribute *aspeed_wdt_attrs[] = {
>> +	&dev_attr_reset_system.attr,
>> +	&dev_attr_ext_signal.attr,
>> +	NULL
>> +};
>> +ATTRIBUTE_GROUPS(aspeed_wdt);
>> +
>>   static int aspeed_wdt_probe(struct platform_device *pdev)
>>   {
>>   	struct aspeed_wdt *wdt;
>> @@ -160,7 +237,7 @@ static int aspeed_wdt_probe(struct platform_device *pdev)
>>   	wdt->wdd.ops = &aspeed_wdt_ops;
>>   	wdt->wdd.max_hw_heartbeat_ms = WDT_MAX_TIMEOUT_MS;
>>   	wdt->wdd.parent = &pdev->dev;
>> -
>> +	wdt->wdd.groups = aspeed_wdt_groups;
>>   	wdt->wdd.timeout = WDT_DEFAULT_TIMEOUT;
>>   	watchdog_init_timeout(&wdt->wdd, 0, &pdev->dev);
>>   
>> -- 
>> 1.8.2.2
>>