On our x86 platform, we see a failure case of calling clocksource_cyc2ns(),
which return a negative value. The reason is the time interval was large
(more than 1000 seconds), while its TSC frequency is 2GHz, so the following
fomular overflowed:
((u64) cycles * mult) >> shift
So enlarge the time interval from 10 mins to 40 mins to fix the bug.
Another solution may be adding a "max_interval" in struct clocksource, and
use a default value (like current 10 minutes) when clocksource driver
doesn't set it.
Signed-off-by: Feng Tang <feng.t...@intel.com>
---
kernel/time/clocksource.c | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/kernel/time/clocksource.c b/kernel/time/clocksource.c
index c958338..48fbfcb 100644
--- a/kernel/time/clocksource.c
+++ b/kernel/time/clocksource.c
@@ -663,7 +663,7 @@ void __clocksource_updatefreq_scale(struct clocksource *cs,
u32 scale, u32 freq)
* Calc the maximum number of seconds which we can run before
* wrapping around. For clocksources which have a mask > 32bit
* we need to limit the max sleep time to have a good
- * conversion precision. 10 minutes is still a reasonable
+ * conversion precision. 40 minutes is still a reasonable
* amount. That results in a shift value of 24 for a
* clocksource with mask >= 40bit and f >= 4GHz. That maps to
* ~ 0.06ppm granularity for NTP. We apply the same 12.5%
@@ -674,8 +674,8 @@ void __clocksource_updatefreq_scale(struct clocksource *cs,
u32 scale, u32 freq)
do_div(sec, scale);
if (!sec)
sec = 1;
- else if (sec > 600 && cs->mask > UINT_MAX)
- sec = 600;
+ else if (sec > 2400 && cs->mask > UINT_MAX)
+ sec = 2400;
clocks_calc_mult_shift(&cs->mult, &cs->shift, freq,
NSEC_PER_SEC / scale, sec * scale);
--
1.7.9.5
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
