Module Name: src
Committed By: bouyer
Date: Sun Sep 10 15:23:01 UTC 2023
Modified Files:
src/sys/arch/xen/xen: xen_clock.c
Log Message:
If Xen system time went backward, instead of ignoring the event assume
one tick passed and reinitialise the local state.
If we need to call the hardclock timer several time to catch up missed events,
don't do more than one second in a single event; and schedule the next
event faster.
Hopefully fixes occasionnal
panic: cpu1: time has not advanced in 1501 heartbeats
or
panic: cpu0: softints stuck for 16 seconds
seen in daily Xen atf runs.
To generate a diff of this commit:
cvs rdiff -u -r1.17 -r1.18 src/sys/arch/xen/xen/xen_clock.c
Please note that diffs are not public domain; they are subject to the
copyright notices on the relevant files.
Modified files:
Index: src/sys/arch/xen/xen/xen_clock.c
diff -u src/sys/arch/xen/xen/xen_clock.c:1.17 src/sys/arch/xen/xen/xen_clock.c:1.18
--- src/sys/arch/xen/xen/xen_clock.c:1.17 Tue Aug 1 20:11:13 2023
+++ src/sys/arch/xen/xen/xen_clock.c Sun Sep 10 15:23:01 2023
@@ -1,4 +1,4 @@
-/* $NetBSD: xen_clock.c,v 1.17 2023/08/01 20:11:13 riastradh Exp $ */
+/* $NetBSD: xen_clock.c,v 1.18 2023/09/10 15:23:01 bouyer Exp $ */
/*-
* Copyright (c) 2017, 2018 The NetBSD Foundation, Inc.
@@ -36,7 +36,7 @@
#endif
#include <sys/cdefs.h>
-__KERNEL_RCSID(0, "$NetBSD: xen_clock.c,v 1.17 2023/08/01 20:11:13 riastradh Exp $");
+__KERNEL_RCSID(0, "$NetBSD: xen_clock.c,v 1.18 2023/09/10 15:23:01 bouyer Exp $");
#include <sys/param.h>
#include <sys/types.h>
@@ -800,7 +800,6 @@ xen_timer_handler(void *cookie, struct c
#if defined(XENPV)
frame = NULL; /* We use values cached in curcpu() */
#endif
-again:
/*
* Find how many nanoseconds of Xen system time has elapsed
* since the last hardclock tick.
@@ -817,7 +816,11 @@ again:
last - now);
#endif
ci->ci_xen_systime_backwards_hardclock_evcnt.ev_count++;
- now = last;
+ /*
+ * we've lost track of time. Just pretends that one
+ * tick elapsed, and reset our idea of last tick.
+ */
+ ci->ci_xen_hardclock_systime_ns = last = now - ns_per_tick;
}
delta = now - last;
@@ -845,6 +848,9 @@ again:
xen_timecounter.tc_counter_mask);
ci->ci_xen_timecounter_jump_evcnt.ev_count++;
}
+ /* don't try to catch up more than one second at once */
+ if (delta > 1000000000UL)
+ delta = 1000000000UL;
}
while (delta >= ns_per_tick) {
ci->ci_xen_hardclock_systime_ns += ns_per_tick;
@@ -859,13 +865,19 @@ again:
/*
* Re-arm the timer. If it fails, it's probably because the
- * time is in the past, so update our idea of what the Xen
- * system time is and try again.
+ * time is in the past, possibly because we're in the
+ * process of catching up missed hardclock calls.
+ * In this case schedule a tick in the nead future.
*/
next = ci->ci_xen_hardclock_systime_ns + ns_per_tick;
error = HYPERVISOR_set_timer_op(next);
- if (error)
- goto again;
+ if (error) {
+ next = xen_vcputime_systime_ns() + ns_per_tick / 2;
+ error = HYPERVISOR_set_timer_op(next);
+ if (error) {
+ panic("failed to re-arm Xen timer %d", error);
+ }
+ }
/* Success! */
return 0;
