Re: [PATCH 2/2] watchdog: rti: tweak min_hw_heartbeat_ms to match initial allowed window

[Guenter Roeck]

On 6/25/20 1:32 AM, Tero Kristo wrote:
> On 24/06/2020 18:24, Jan Kiszka wrote:
>> On 24.06.20 13:45, Tero Kristo wrote:
>>> If the RTI watchdog has been started by someone (like bootloader) when
>>> the driver probes, we must adjust the initial ping timeout to match the
>>> currently running watchdog window to avoid generating watchdog reset.
>>>
>>> Signed-off-by: Tero Kristo <t-kristo@xxxxxx>
>>> ---
>>> Â drivers/watchdog/rti_wdt.c | 25 +++++++++++++++++++++++++
>>> Â 1 file changed, 25 insertions(+)
>>>
>>> diff --git a/drivers/watchdog/rti_wdt.c b/drivers/watchdog/rti_wdt.c
>>> index d456dd72d99a..02ea2b2435f5 100644
>>> --- a/drivers/watchdog/rti_wdt.c
>>> +++ b/drivers/watchdog/rti_wdt.c
>>> @@ -55,11 +55,13 @@ static int heartbeat;
>>> ÂÂ * @base - base io address of WD device
>>> ÂÂ * @freq - source clock frequency of WDT
>>>  * @wdd - hold watchdog device as is in WDT core
>>> + * @min_hw_heartbeat_save - save of the min hw heartbeat value
>>> ÂÂ */
>>> Â struct rti_wdt_device {
>>> ÂÂÂÂÂ void __iomemÂÂÂÂÂÂÂ *base;
>>> ÂÂÂÂÂ unsigned longÂÂÂÂÂÂÂ freq;
>>> ÂÂÂÂÂ struct watchdog_deviceÂÂÂ wdd;
>>> +ÂÂÂ unsigned intÂÂÂÂÂÂÂ min_hw_heartbeat_save;
>>> Â };
>>> Â static int rti_wdt_start(struct watchdog_device *wdd)
>>> @@ -107,6 +109,11 @@ static int rti_wdt_ping(struct watchdog_device *wdd)
>>> ÂÂÂÂÂ /* put watchdog in active state */
>>> ÂÂÂÂÂ writel_relaxed(WDKEY_SEQ1, wdt->base + RTIWDKEY);
>>> +ÂÂÂ if (wdt->min_hw_heartbeat_save) {
>>> +ÂÂÂÂÂÂÂ wdd->min_hw_heartbeat_ms = wdt->min_hw_heartbeat_save;
>>> +ÂÂÂÂÂÂÂ wdt->min_hw_heartbeat_save = 0;
>>> +ÂÂÂ }
>>> +
>>> ÂÂÂÂÂ return 0;
>>> Â }
>>> @@ -201,6 +208,24 @@ static int rti_wdt_probe(struct platform_device *pdev)
>>> ÂÂÂÂÂÂÂÂÂ goto err_iomap;
>>> ÂÂÂÂÂ }
>>> +ÂÂÂ if (readl(wdt->base + RTIDWDCTRL) == WDENABLE_KEY) {
>>> +ÂÂÂÂÂÂÂ u32 time_left;
>>> +ÂÂÂÂÂÂÂ u32 heartbeat;
>>> +
>>> +ÂÂÂÂÂÂÂ set_bit(WDOG_HW_RUNNING, &wdd->status);
>>> +ÂÂÂÂÂÂÂ time_left = rti_wdt_get_timeleft(wdd);
>>> +ÂÂÂÂÂÂÂ heartbeat = readl(wdt->base + RTIDWDPRLD);
>>> +ÂÂÂÂÂÂÂ heartbeat <<= WDT_PRELOAD_SHIFT;
>>> +ÂÂÂÂÂÂÂ heartbeat /= wdt->freq;
>>> +ÂÂÂÂÂÂÂ if (time_left < heartbeat / 2)
>>> +ÂÂÂÂÂÂÂÂÂÂÂ wdd->min_hw_heartbeat_ms = 0;
>>> +ÂÂÂÂÂÂÂ else
>>> +ÂÂÂÂÂÂÂÂÂÂÂ wdd->min_hw_heartbeat_ms =
>>> +ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂ (time_left - heartbeat / 2 + 1) * 1000;
>>> +
>>> +ÂÂÂÂÂÂÂ wdt->min_hw_heartbeat_save = 11 * heartbeat * 1000 / 20;
>>> +ÂÂÂ }
>>> +
>>> ÂÂÂÂÂ ret = watchdog_register_device(wdd);
>>> ÂÂÂÂÂ if (ret) {
>>> ÂÂÂÂÂÂÂÂÂ dev_err(dev, "cannot register watchdog device\n");
>>>
>>
>> This assumes that the bootloader also programmed a 50% window, right? The pending U-Boot patch will do that, but what if that may chance or someone uses a different setup?
> 
> Yes, we assume 50%. I think based on the hw design, 50% is the only sane value to be used, otherwise you just shrink the open window too much and for no apparent reason.
> 

Not sure if that is a valid assumption. Someone who designs a watchdog
with such a narrow ping window might as well also use it. The question
is if you want to rely on that assumption, or check and change it if needed.

Also, I wonder if we should add an API function such as
"set_last_hw_keepalive()" to avoid all that complexity.

Thanks,
Guenter