Skip to contentSkip to conversation

timekeeping: always make sure wall_to_monotonic isn't positive

PATCH In review Wang YanQing 2 participants 2 messages Jun 23, 2015 – Jul 1, 2015 Timekeeping, clocksource core, NTP, alarmtimer

lore
Share & tools
lore
#0 lore raw

PATCH v3 1/1 timekeeping: always make sure wall_to_monotonic isn't positive

I meet two issues on an IMX6 development board without enable RTC device(so timekeeping_init will initialize the boot time and monotonic to 0).

Issue 1:exportfs -a generate:
"exportfs: /opt/nfs/arm does not support NFS export"
Issue 2:cat /proc/stat:

       "btime 4294967236"

Exact reproduction of the same issues on x86 after run below code:
" int main(void)

	{
	    struct timeval val;
	    int ret;

	    val.tv_sec = 0;
	    val.tv_usec = 0;
	    ret = settimeofday(&val, NULL);
	    return 0;
	}

"
Two issues are different symptoms of same problem:
The reason is positive wall_to_monotonic push boot time back to the time before Epoch, getboottime will return negative value.

In symptom 1:
negative boot time cause get_expiry overflow time_t when input expire time is 2147483647, then cache_flush always clear entries just added in ip_map_parse.
In symptom 2:
show_stat use "unsigned long" to print
negative value return by getboottime.

This patch fix the problem.

Impact:
You can't set the CLOCK_REALTIME time prior to (1970 + system uptime).

---
Changes
v1-v2:
1: fix subject, use "isn't positive" instead of "is negative". 2: rewrite changelog.
3: simplify code as suggested by John Stultz.

v2-v3:
1: reword changelog, no change of source code.

 kernel/time/timekeeping.c | 13 ++++++++++---
 1 file changed, 10 insertions(+), 3 deletions(-)
kernel/time/timekeeping.c +10 -3
index bca3667..4cdb771 100644
@@ -911,6 +911,7 @@ int do_settimeofday64(const struct timespec64 *ts)
 	struct timekeeper *tk = &tk_core.timekeeper;
 	struct timespec64 ts_delta, xt;
 	unsigned long flags;
+	int ret = 0;
 
 	if (!timespec64_valid_strict(ts))
 		return -EINVAL;
@@ -924,10 +925,15 @@ int do_settimeofday64(const struct timespec64 *ts)
 	ts_delta.tv_sec = ts->tv_sec - xt.tv_sec;
 	ts_delta.tv_nsec = ts->tv_nsec - xt.tv_nsec;
 
+	if (timespec64_compare(&tk->wall_to_monotonic, &ts_delta) > 0) {
+		ret = -EINVAL;
+		goto out;
+	}
+
 	tk_set_wall_to_mono(tk, timespec64_sub(tk->wall_to_monotonic, ts_delta));
 
 	tk_set_xtime(tk, ts);
-
+out:
 	timekeeping_update(tk, TK_CLEAR_NTP | TK_MIRROR | TK_CLOCK_WAS_SET);
 
 	write_seqcount_end(&tk_core.seq);
@@ -936,7 +942,7 @@ int do_settimeofday64(const struct timespec64 *ts)
 	/* signal hrtimers about time change */
 	clock_was_set();
 
-	return 0;
+	return ret;
 }
 EXPORT_SYMBOL(do_settimeofday64);
 
@@ -965,7 +971,8 @@ int timekeeping_inject_offset(struct timespec *ts)
 
 	/* Make sure the proposed value is valid */
 	tmp = timespec64_add(tk_xtime(tk),  ts64);
-	if (!timespec64_valid_strict(&tmp)) {
+	if (timespec64_compare(&tk->wall_to_monotonic, &ts64) > 0 ||
+	    !timespec64_valid_strict(&tmp)) {
 		ret = -EINVAL;
 		goto error;
 	}
Signature · 1 line
1.8.5.6.2.g3d8a54e.dirty
#1 lore raw
Quoted 34 lines · Wang YanQing

I meet two issues on an IMX6 development board without enable RTC device(so timekeeping_init will initialize the boot time and monotonic to 0).

Issue 1:exportfs -a generate:
"exportfs: /opt/nfs/arm does not support NFS export"
Issue 2:cat /proc/stat:

       "btime 4294967236"

Exact reproduction of the same issues on x86 after run below code:
" int main(void)

        {
            struct timeval val;
            int ret;

            val.tv_sec = 0;
            val.tv_usec = 0;
            ret = settimeofday(&val, NULL);
            return 0;
        }

"
Two issues are different symptoms of same problem:
The reason is positive wall_to_monotonic push boot time back to the time before Epoch, getboottime will return negative value.

In symptom 1:
negative boot time cause get_expiry overflow time_t when input expire time is 2147483647, then cache_flush always clear entries just added in ip_map_parse.
In symptom 2:
show_stat use "unsigned long" to print
negative value return by getboottime.

Quoted 6 lines

This patch fix the problem.

Impact:
You can't set the CLOCK_REALTIME time prior to (1970 + system uptime).

Ok, the patch looks alright. The commit message is still not super readable, but I went ahead and re-worded much of it. I've got it queued for 4.3, and will be doing some testing. When I sent my stack out to Thomas/Ingo for -tip, please review the rewording to make sure you agree with the changes.

thanks
-john