hello,
failure of a test in selftest:timers
selftests: raw_skew
========================================
WARNING: ADJ_OFFSET in progress, this will cause inaccurate results
Estimating clock drift: 17.910(est) 16.820(act) [FAILED]
Bail out!
Pass 0 Fail 0 Xfail 0 Xpass 0 Skip 0 Error 0
1..0
not ok 1..7 selftests: raw_skew [FAIL]
To reproduce :
make -C tools/testing/selftests/timers/ run_tests
Linux debian 4.17.0-rc1+ #3 SMP Sat Apr 21 03:41:44 IST 2018 x86_64 GNU/Linux
GNU Make 4.2.1
Binutils 2.30
Util-linux 2.31.1
Mount 2.31.1
Linux C Library 2.27
Dynamic linker (ldd) 2.27
readlink: missing operand
Try 'readlink --help' for more information.
Procps 3.3.14
Kbd 2.0.4
Console-tools 2.0.4
Sh-utils 8.28
Udev 238
--
software engineer
rajagiri school of engineering and technology
On Sat, 21 Apr 2018, Jeffrin Thalakkottoor wrote:
Cc'ing John and Miroslav
> hello,
>
> failure of a test in selftest:timers
>
> selftests: raw_skew
> ========================================
> WARNING: ADJ_OFFSET in progress, this will cause inaccurate results
> Estimating clock drift: 17.910(est) 16.820(act) [FAILED]
> Bail out!
> Pass 0 Fail 0 Xfail 0 Xpass 0 Skip 0 Error 0
> 1..0
> not ok 1..7 selftests: raw_skew [FAIL]
>
>
> To reproduce :
>
> make -C tools/testing/selftests/timers/ run_tests
>
>
>
>
> Linux debian 4.17.0-rc1+ #3 SMP Sat Apr 21 03:41:44 IST 2018 x86_64 GNU/Linux
>
> GNU Make 4.2.1
> Binutils 2.30
> Util-linux 2.31.1
> Mount 2.31.1
> Linux C Library 2.27
> Dynamic linker (ldd) 2.27
> readlink: missing operand
> Try 'readlink --help' for more information.
> Procps 3.3.14
> Kbd 2.0.4
> Console-tools 2.0.4
> Sh-utils 8.28
> Udev 238
>
> --
> software engineer
> rajagiri school of engineering and technology
>
On Sat, Apr 21, 2018 at 7:47 AM, Jeffrin Thalakkottoor
<[email protected]> wrote:
> hello,
>
> failure of a test in selftest:timers
>
> selftests: raw_skew
> ========================================
> WARNING: ADJ_OFFSET in progress, this will cause inaccurate results
> Estimating clock drift: 17.910(est) 16.820(act) [FAILED]
So you're ignoring the warning above. If an ADJ_OFFSET operation is
going on, ntpd (or something similar) is changing the clock, which
will kill the estimated drift calculation.
-john
On Thu, Apr 26, 2018 at 11:28:29PM +0200, Thomas Gleixner wrote:
> On Sat, 21 Apr 2018, Jeffrin Thalakkottoor wrote:
> > selftests: raw_skew
> > ========================================
> > WARNING: ADJ_OFFSET in progress, this will cause inaccurate results
> > Estimating clock drift: 17.910(est) 16.820(act) [FAILED]
Was ntpd, systemd-timesyncd, or some other NTP/PTP daemon running
shortly before or during the test?
The warning indicates that the clock was slewed by adjtime() or
adjtimex(), which makes the measurement of the frequency less accurate
and the test may fail.
Maybe this test and other tests that measure the frequency of the
system clock should be modified to SKIP when adjtimex() returns a
non-zero offset (or the frequency changes during the test)?
--
Miroslav Lichvar
I kind of ignored. i was not sure what that meant actually.
On Fri, Apr 27, 2018 at 3:03 AM, John Stultz <[email protected]> wrote:
> On Sat, Apr 21, 2018 at 7:47 AM, Jeffrin Thalakkottoor
> <[email protected]> wrote:
>> hello,
>>
>> failure of a test in selftest:timers
>>
>> selftests: raw_skew
>> ========================================
>> WARNING: ADJ_OFFSET in progress, this will cause inaccurate results
>> Estimating clock drift: 17.910(est) 16.820(act) [FAILED]
>
> So you're ignoring the warning above. If an ADJ_OFFSET operation is
> going on, ntpd (or something similar) is changing the clock, which
> will kill the estimated drift calculation.
>
> -john
--
software engineer
rajagiri school of engineering and technology
i think may be systemd-timesyncd was running.
anyway currently the status is as follows...
$systemctl status systemd-timesyncd.service
● systemd-timesyncd.service - Network Time Synchronization
Loaded: loaded (/lib/systemd/system/systemd-timesyncd.service;
enabled; vendor preset: enabled)
Active: active (running) since Fri 2018-04-27 19:26:16 IST; 1h 11min ago
Docs: man:systemd-timesyncd.service(8)
Main PID: 514 (systemd-timesyn)
Status: "Synchronized to time server 123.108.200.124:123
(0.debian.pool.ntp.org)."
Tasks: 2 (limit: 4382)
Memory: 1.8M
CGroup: /system.slice/systemd-timesyncd.service
└─514 /lib/systemd/systemd-timesyncd
$
The test in the latest run show "PASS"
see below...
selftests: raw_skew
========================================
Estimating clock drift: 26.836(est) 26.838(act) [OK]
Pass 0 Fail 0 Xfail 0 Xpass 0 Skip 0 Error 0
1..0
ok 1..7 selftests: raw_skew [PASS]
On Fri, Apr 27, 2018 at 1:11 PM, Miroslav Lichvar <[email protected]> wrote:
> On Thu, Apr 26, 2018 at 11:28:29PM +0200, Thomas Gleixner wrote:
>> On Sat, 21 Apr 2018, Jeffrin Thalakkottoor wrote:
>> > selftests: raw_skew
>> > ========================================
>> > WARNING: ADJ_OFFSET in progress, this will cause inaccurate results
>> > Estimating clock drift: 17.910(est) 16.820(act) [FAILED]
>
> Was ntpd, systemd-timesyncd, or some other NTP/PTP daemon running
> shortly before or during the test?
>
> The warning indicates that the clock was slewed by adjtime() or
> adjtimex(), which makes the measurement of the frequency less accurate
> and the test may fail.
>
> Maybe this test and other tests that measure the frequency of the
> system clock should be modified to SKIP when adjtimex() returns a
> non-zero offset (or the frequency changes during the test)?
>
> --
> Miroslav Lichvar
--
software engineer
rajagiri school of engineering and technology