| Message ID | 20210512084724.14634-1-shruthi.sanil@intel.com |
|---|---|
| Headers | show |
| Series | Intel Keem Bay WDT bug fixes | expand |
> -----Original Message----- > From: Guenter Roeck <groeck7@gmail.com> On Behalf Of Guenter Roeck > Sent: Wednesday, May 12, 2021 7:32 PM > To: Sanil, Shruthi <shruthi.sanil@intel.com> > Cc: wim@linux-watchdog.org; linux-watchdog@vger.kernel.org; linux- > kernel@vger.kernel.org; andriy.shevchenko@linux.intel.com; > kris.pan@linux.intel.com; mgross@linux.intel.com; Thokala, Srikanth > <srikanth.thokala@intel.com>; Raja Subramanian, Lakshmi Bai > <lakshmi.bai.raja.subramanian@intel.com>; Sangannavar, Mallikarjunappa > <mallikarjunappa.sangannavar@intel.com> > Subject: Re: [PATCH 07/10] watchdog: keembay: Update the check in > keembay_wdt_resume() > > On Wed, May 12, 2021 at 02:17:21PM +0530, shruthi.sanil@intel.com wrote: > > From: Shruthi Sanil <shruthi.sanil@intel.com> > > > > Corrected the typo in the function keembay_wdt_resume, we need to > > enable the WDT if it is disabled/not active. > > > > Fixes: fa0f8d51e90d ("watchdog: Add watchdog driver for Intel Keembay > > Soc") > > Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> > > Tested-by: Kris Pan <kris.pan@intel.com> > > Signed-off-by: Shruthi Sanil <shruthi.sanil@intel.com> > > --- > > drivers/watchdog/keembay_wdt.c | 2 +- > > 1 file changed, 1 insertion(+), 1 deletion(-) > > > > diff --git a/drivers/watchdog/keembay_wdt.c > > b/drivers/watchdog/keembay_wdt.c index dd192b8dff55..10896415f8c7 > > 100644 > > --- a/drivers/watchdog/keembay_wdt.c > > +++ b/drivers/watchdog/keembay_wdt.c > > @@ -260,7 +260,7 @@ static int __maybe_unused > > keembay_wdt_resume(struct device *dev) { > > struct keembay_wdt *wdt = dev_get_drvdata(dev); > > > > - if (watchdog_active(&wdt->wdd)) > > + if (!watchdog_active(&wdt->wdd)) > > Have you tested this ? "watchdog_active" refers to the watchdog core state. > Your code now keeps the watchdog stopped after resume if it was running > before, and starts it if it wasn't. Please run through a suspend/resume cycle > with watchdog disabled and see what happens. > > Guenter I had understood it wrongly. I was assuming that watchdog_active refers to the actual state of the HW. Hence I made that change. But since watchdog_active refers to the state of the core, the change that I have done is incorrect. Hence I'll drop this patch. Thanks Shruthi > > > return keembay_wdt_start(&wdt->wdd); > > > > return 0; > > -- > > 2.17.1 > >
From: Shruthi Sanil <shruthi.sanil@intel.com> The series of patches include the below bug fixes in the Intel Keem Bay watchdog timer driver: Patch 1/10: - Update WDT pre-timeout during the initialization The pretimeout register has a default reset value. Hence when a smaller WDT timeout is set which would be lesser than the default pretimeout, the system behaves abnormally, starts triggering the pretimeout interrupt even when the WDT is not enabled, most of the times leading to system crash. Hence an update in the pre-timeout is also required for the default timeout that is being configured. Patch 2/10: - Upadate WDT pretimeout for every update in timeout The pre-timeout value to be programmed to the register has to be calculated and updated for every change in the timeout value. Else the threshold time wouldn't be calculated to its corresponding timeout. Patch 3/10: - Update pretimeout to 0 in the TH ISR The pretimeout has to be updated to 0 during the ISR of the ThresHold interrupt. Else the TH interrupt would be triggerred for every tick until the timeout. Patch 4/10: - Clear either the TO or TH interrupt bit During the interrupt service routine of the TimeOut interrupt and the ThresHold interrupt, the respective interrupt clear bit have to be cleared and not both. Patch 5/10: - Remove timeout update in the WDT start function Removed set timeout from the start WDT function. There is a function defined to set the timeout. Hence no need to set the timeout again in start function as the timeout would have been already updated before calling the start/enable. Patch 6/10: - Removed timeout update in the TO ISR In the TO ISR removed updating the Timeout value because its not serving any purpose as the timer would have already expired and the system would be rebooting. Patch 7/10: - Update the check in keembay_wdt_resume() Corrected the typo in the function keembay_wdt_resume, we need to enable the WDT if it is disabled/not active. Patch 8/10: - MACRO for WDT enable and disable values Introduced MACRO's for WDT enable and disable values for better readability Patch 9/10: - WDT SMC handler MACRO name update Updated the WDT SMC handler MACRO name to make it clear that its a ARM SMC handler that helps in clearing the WDT interrupt bit. Patch 10/10: - Typo corrections and other blank operations Corrected typos, aligned the tabs and added new lines wherever required for better readability Shruthi Sanil (10): watchdog: keembay: Update WDT pre-timeout during the initialization watchdog: keembay: Upadate WDT pretimeout for every update in timeout watchdog: keembay: Update pretimeout to zero in the TH ISR watchdog: keembay: Clear either the TO or TH interrupt bit watchdog: keembay: Remove timeout update in the WDT start function watchdog: keembay: Removed timeout update in the TO ISR watchdog: keembay: Update the check in keembay_wdt_resume() watchdog: keembay: MACRO for WDT enable and disable values watchdog: keembay: WDT SMC handler MACRO name update watchdog: keembay: Typo corrections and other blank operations drivers/watchdog/keembay_wdt.c | 36 ++++++++++++++++++++++------------ 1 file changed, 23 insertions(+), 13 deletions(-) base-commit: 88b06399c9c766c283e070b022b5ceafa4f63f19