On some arm64 machines, the agtimer(4) ticks slowly enough that the
tc_delta() doesn't overflow across brief suspends. While working on
arm64 suspend/resume, kettenis@ has been seeing warnings like this
during resume:
tc_setclock: cannot rewind uptime to 307.253324249
The warning is misleading and should be removed. The code is behaving
as intended, but in a way I didn't anticipate when I added the warning
a few years ago.
It might be useful print a warning in inittodr(9) during resume if the
RTC time predates the system UTC suspend timestamp, but that's a
distinct concern.
ok?
Index: kern_tc.c
===================================================================
RCS file: /cvs/src/sys/kern/kern_tc.c,v
retrieving revision 1.78
diff -u -p -r1.78 kern_tc.c
--- kern_tc.c 18 Sep 2022 20:47:09 -0000 1.78
+++ kern_tc.c 8 Nov 2022 11:53:01 -0000
@@ -552,7 +552,6 @@ void
tc_setclock(const struct timespec *ts)
{
struct bintime new_naptime, old_naptime, uptime, utc;
- struct timespec tmp;
static int first = 1;
#ifndef SMALL_KERNEL
struct bintime elapsed;
@@ -582,12 +581,6 @@ tc_setclock(const struct timespec *ts)
new_naptime = timehands->th_naptime;
mtx_leave(&windup_mtx);
-
- if (bintimecmp(&old_naptime, &new_naptime, ==)) {
- BINTIME_TO_TIMESPEC(&uptime, &tmp);
- printf("%s: cannot rewind uptime to %lld.%09ld\n",
- __func__, (long long)tmp.tv_sec, tmp.tv_nsec);
- }
#ifndef SMALL_KERNEL
/* convert the bintime to ticks */
