Specs don't say anything about UIP being cleared within 10ms. They
only say that UIP won't occur for another 244uS. If a long NMI occurs
while UIP is still updating it might not be possible to get valid
data in 10ms.

This has been observed in the wild that around s2idle some calls can
take up to 480ms before UIP is clear.

Adjust callers from outside an interrupt context to wait for up to a
1s instead of 10ms.

Cc: stable@vger.kernel.org # 6.1.y: commit d2a632a8a117 ("rtc: mc146818-lib: reduce RTC_UIP polling period")
Fixes: ec5895c0f2d8 ("rtc: mc146818-lib: extract mc146818_avoid_UIP")
Reported-by: xmb8dsv4@gmail.com
Closes: https://bugzilla.kernel.org/show_bug.cgi?id=217626
Signed-off-by: Mario Limonciello <mario.limonciello@amd.com>
---
v1->v2:
 * Add tags
---
 arch/x86/kernel/rtc.c          | 2 +-
 drivers/base/power/trace.c     | 2 +-
 drivers/rtc/rtc-cmos.c         | 2 +-
 drivers/rtc/rtc-mc146818-lib.c | 2 +-
 4 files changed, 4 insertions(+), 4 deletions(-)

diff --git a/arch/x86/kernel/rtc.c b/arch/x86/kernel/rtc.c
index 961ebc7f1872..2e7066980f3e 100644
--- a/arch/x86/kernel/rtc.c
+++ b/arch/x86/kernel/rtc.c
@@ -67,7 +67,7 @@ void mach_get_cmos_time(struct timespec64 *now)
 		return;
 	}
 
-	if (mc146818_get_time(&tm, 10)) {
+	if (mc146818_get_time(&tm, 1000)) {
 		pr_err("Unable to read current time from RTC\n");
 		now->tv_sec = now->tv_nsec = 0;
 		return;
diff --git a/drivers/base/power/trace.c b/drivers/base/power/trace.c
index c2e925357474..cd6e559648b2 100644
--- a/drivers/base/power/trace.c
+++ b/drivers/base/power/trace.c
@@ -120,7 +120,7 @@ static unsigned int read_magic_time(void)
 	struct rtc_time time;
 	unsigned int val;
 
-	if (mc146818_get_time(&time, 10) < 0) {
+	if (mc146818_get_time(&time, 1000) < 0) {
 		pr_err("Unable to read current time from RTC\n");
 		return 0;
 	}
diff --git a/drivers/rtc/rtc-cmos.c b/drivers/rtc/rtc-cmos.c
index d278b085821e..6bc1e0279cd9 100644
--- a/drivers/rtc/rtc-cmos.c
+++ b/drivers/rtc/rtc-cmos.c
@@ -231,7 +231,7 @@ static int cmos_read_time(struct device *dev, struct rtc_time *t)
 	if (!pm_trace_rtc_valid())
 		return -EIO;
 
-	ret = mc146818_get_time(t, 10);
+	ret = mc146818_get_time(t, 1000);
 	if (ret < 0) {
 		dev_err_ratelimited(dev, "unable to read current time\n");
 		return ret;
diff --git a/drivers/rtc/rtc-mc146818-lib.c b/drivers/rtc/rtc-mc146818-lib.c
index 5abb925da4fa..9191e5f50fae 100644
--- a/drivers/rtc/rtc-mc146818-lib.c
+++ b/drivers/rtc/rtc-mc146818-lib.c
@@ -91,7 +91,7 @@ EXPORT_SYMBOL_GPL(mc146818_avoid_UIP);
  */
 bool mc146818_does_rtc_work(void)
 {
-	return mc146818_avoid_UIP(NULL, 10, NULL);
+	return mc146818_avoid_UIP(NULL, 1000, NULL);
 }
 EXPORT_SYMBOL_GPL(mc146818_does_rtc_work);
 
-- 
2.34.1

W dniu 20.11.2023 o 15:15, Mario Limonciello pisze:
> Specs don't say anything about UIP being cleared within 10ms. They
> only say that UIP won't occur for another 244uS. If a long NMI occurs
> while UIP is still updating it might not be possible to get valid
> data in 10ms.
>
> This has been observed in the wild that around s2idle some calls can
> take up to 480ms before UIP is clear.
>
> Adjust callers from outside an interrupt context to wait for up to a
> 1s instead of 10ms.
>
> Cc: stable@vger.kernel.org # 6.1.y: commit d2a632a8a117 ("rtc: mc146818-lib: reduce RTC_UIP polling period")
Same as for patch 3/4.
> Fixes: ec5895c0f2d8 ("rtc: mc146818-lib: extract mc146818_avoid_UIP")
> Reported-by: xmb8dsv4@gmail.com

Maybe nitpick, but consider providing the full name of Carsten Hatger, not just
an email address.

Otherwise,

Reviewed-by: Mateusz Jończyk <mat.jonczyk@o2.pl>
Acked-by: Mateusz Jończyk <mat.jonczyk@o2.pl>

> Closes: https://bugzilla.kernel.org/show_bug.cgi?id=217626
> Signed-off-by: Mario Limonciello <mario.limonciello@amd.com>
> ---
> v1->v2:
>  * Add tags
> ---
>  arch/x86/kernel/rtc.c          | 2 +-
>  drivers/base/power/trace.c     | 2 +-
>  drivers/rtc/rtc-cmos.c         | 2 +-
>  drivers/rtc/rtc-mc146818-lib.c | 2 +-
>  4 files changed, 4 insertions(+), 4 deletions(-)
>  

Greetings,

Mateusz