| Message ID | 20201026080919.28413-3-zhang.lyra@gmail.com (mailing list archive) |
|---|---|
| State | Superseded |
| Headers | show |
| Series | A few fixes to sprd watchdog driver | expand |
On 10/26/20 1:09 AM, Chunyan Zhang wrote: > From: Lingling Xu <ling_ling.xu@unisoc.com> > > Because cpu_relax() takes different time on different SoCs, for some rare > cases, it would take more than 1000 cycles for waitting load operation waiting > finished. The result of many times testing verified that changing the > timeout value to 2000 can solve the issue. > This is just a kludge that doesn't address the underlying problem. As the wait loop states, "Waiting the load value operation done, it needs two or three RTC clock cycles". This means the loop should wait for a maximum number of clock cycles, and not run as hot loop. If we assume that clk_get_rate() returns the clock frequency, that frequency can be used to determine how long this needs to be retried. It might also make sense - depending on how long this actually takes - to use usleep_range() instead of cpu_relax() to avoid the hot loop. Guenter > Fixes: 477603467009 ("watchdog: Add Spreadtrum watchdog driver") > Signed-off-by: Lingling Xu <ling_ling.xu@unisoc.com> > Signed-off-by: Chunyan Zhang <chunyan.zhang@unisoc.com> > --- > drivers/watchdog/sprd_wdt.c | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/drivers/watchdog/sprd_wdt.c b/drivers/watchdog/sprd_wdt.c > index f3c90b4afead..4f2a8c6d6485 100644 > --- a/drivers/watchdog/sprd_wdt.c > +++ b/drivers/watchdog/sprd_wdt.c > @@ -53,7 +53,7 @@ > > #define SPRD_WDT_CNT_HIGH_SHIFT 16 > #define SPRD_WDT_LOW_VALUE_MASK GENMASK(15, 0) > -#define SPRD_WDT_LOAD_TIMEOUT 1000 > +#define SPRD_WDT_LOAD_TIMEOUT 2000 > > struct sprd_wdt { > void __iomem *base; >
On Mon, 26 Oct 2020 at 22:36, Guenter Roeck <linux@roeck-us.net> wrote: > > On 10/26/20 1:09 AM, Chunyan Zhang wrote: > > From: Lingling Xu <ling_ling.xu@unisoc.com> > > > > Because cpu_relax() takes different time on different SoCs, for some rare > > cases, it would take more than 1000 cycles for waitting load operation > > waiting Ok. > > > finished. The result of many times testing verified that changing the > > timeout value to 2000 can solve the issue. > > > > This is just a kludge that doesn't address the underlying problem. > As the wait loop states, "Waiting the load value operation done, > it needs two or three RTC clock cycles". This means the loop > should wait for a maximum number of clock cycles, and not run > as hot loop. If we assume that clk_get_rate() returns the clock > frequency, that frequency can be used to determine how long this > needs to be retried. It might also make sense - depending on how > long this actually takes - to use usleep_range() instead of > cpu_relax() to avoid the hot loop. Agree, using usleep_range() instead makes more sense, I will look into that. Thanks for your review. Chunyan > > Guenter > > > Fixes: 477603467009 ("watchdog: Add Spreadtrum watchdog driver") > > Signed-off-by: Lingling Xu <ling_ling.xu@unisoc.com> > > Signed-off-by: Chunyan Zhang <chunyan.zhang@unisoc.com> > > --- > > drivers/watchdog/sprd_wdt.c | 2 +- > > 1 file changed, 1 insertion(+), 1 deletion(-) > > > > diff --git a/drivers/watchdog/sprd_wdt.c b/drivers/watchdog/sprd_wdt.c > > index f3c90b4afead..4f2a8c6d6485 100644 > > --- a/drivers/watchdog/sprd_wdt.c > > +++ b/drivers/watchdog/sprd_wdt.c > > @@ -53,7 +53,7 @@ > > > > #define SPRD_WDT_CNT_HIGH_SHIFT 16 > > #define SPRD_WDT_LOW_VALUE_MASK GENMASK(15, 0) > > -#define SPRD_WDT_LOAD_TIMEOUT 1000 > > +#define SPRD_WDT_LOAD_TIMEOUT 2000 > > > > struct sprd_wdt { > > void __iomem *base; > > >
diff --git a/drivers/watchdog/sprd_wdt.c b/drivers/watchdog/sprd_wdt.c index f3c90b4afead..4f2a8c6d6485 100644 --- a/drivers/watchdog/sprd_wdt.c +++ b/drivers/watchdog/sprd_wdt.c @@ -53,7 +53,7 @@ #define SPRD_WDT_CNT_HIGH_SHIFT 16 #define SPRD_WDT_LOW_VALUE_MASK GENMASK(15, 0) -#define SPRD_WDT_LOAD_TIMEOUT 1000 +#define SPRD_WDT_LOAD_TIMEOUT 2000 struct sprd_wdt { void __iomem *base;