Hibernate with fwts will cause hwclock command to time out

Can you see if the following does the same:

sudo hwclock -r ; sleep 3; sudo pm-hibernate ; sleep 3 ;sudo hwclock -r

Hello Colin,
sudo hwclock -r ; sleep 3; sudo pm-hibernate (for approx. 2min) ; sleep 3 ;sudo hwclock -r
Works fine

Also, if I wake it up before it wake itself up automatically with fwts, this issue won't happen

ubuntu@201401-14542:~$ sudo hwclock -r ; sleep 3;sudo fwts -P s4 --s4-device-check --s4-device-check-delay=45 --s4-sleep-delay=120; sleep 3 ;sudo hwclock -r
Tue 24 Jun 2014 05:17:01 AM EDT -0.234698 seconds
Running 1 tests, results appended to results.log
Test: S4 hibernate/resume test.
(Wake it up manually after 1 minute)
  S4 hibernate/resume test. 4 passed, 1 failed
Tue 24 Jun 2014 05:19:45 AM EDT -0.437938 seconds
ubuntu@201401-14542:~$ sudo hwclock -r ; sleep 3;sudo fwts -P s4 --s4-device-check --s4-device-check-delay=45 --s4-sleep-delay=120; sleep 3 ;sudo hwclock -r
Tue 24 Jun 2014 05:20:33 AM EDT -0.812936 seconds
Running 1 tests, results appended to results.log
Test: S4 hibernate/resume test.
(Let it wake itself up)
  S4 hibernate/resume test. 4 passed, 1 failed
hwclock: select() to /dev/rtc to wait for clock tick timed out: No such file or directory

I can repoduce this with a 3.15 kernel, strace is showing that hwlock is timing out on the RTC:

open("/dev/rtc", O_RDONLY) = 3
stat("/etc/adjtime", 0x7fff75509770) = -1 ENOENT (No such file or directory)
ioctl(3, PHN_SET_REGS or RTC_UIE_ON, 0) = 0
select(4, [3], NULL, NULL, {5, 0}) = 0 (Timeout)

I'm currently thinking that fwts has tickled a RTC bug in the kernel rather than this being a fwts fault.

I've observed that before hibernate the rtc is not firing:

watch -n 1 "cat /proc/interrupts | grep rtc"

however, after the resume from hibernate it is firing off at ~60Hz, so it maybe that the RTC is in a broken state after resume from hibernate for some reason.

And this *only* happens if the machine resumes from hibernate using the RTC wakealarm. So it smells like a kernel bug and not an issue in fwts per se.

Which kernel was being used?

Ah, 3.13.0-29-generic, I'll see if this bug has been around in previous upstream versions now I've got an easy reproducer.

This bug is missing log files that will aid in diagnosing the problem. From a terminal window please run:

apport-collect 1333569

and then change the status of the bug to 'Confirmed'.

If, due to the nature of the issue you have encountered, you are unable to run this command, please add a comment stating that fact and change the bug status to 'Confirmed'.

This change has been made by an automated script, maintained by the Ubuntu Kernel Team.

apport information

apport information

apport information

apport information

apport information

apport information

apport information

apport information

apport information

apport information

apport information

apport information

apport information

apport information

apport information

I can't reproduce this with the latest version of fwts 14.09.00 and the 3.13.0-37 kernel. Can you re-test this for me with the latest fwts and kernel?

Correction, I can see the failure now, I didn't observe the final hwclock was timing out after 10 seconds since I was using utopic userspace and not trusty userspace hwclock.

Just a note:

1. The utopic version of hwclock has a 10 second timeout but does not report a timeout error (which is misleading)
2. The trusty version reports the timeout, but prints the errno from the previous system call, so the error message is misleading
3. When I get hwclock time-out the following kernel message appears:

   hpet1: lost 20 rtc interrupts

4. I've tested this back to 3.12 and we see the same issue there too.

Fix sent upstream for review. https://lkml.org/lkml/2014/9/26/701

This bug was fixed in the package linux - 3.18.0-9.10

---------------
linux (3.18.0-9.10) vivid; urgency=low

  [ Colin Ian King ]

  * SAUCE: drivers/rtc/interface.c: ignore exprired times when enqueing new
    timers
    - LP: #1333569

  [ Joseph Salisbury ]

  * [Config] Enable CONFIG_N_GSM as module.
    - LP: #1404670

  [ K. Y. Srinivasan ]

  * SAUCE: storvsc: force SPC-3 compliance on win8 and win8 r2 hosts
    - LP: #1406867

  [ Leann Ogasawara ]

  * rebase to v3.18.2
  * Release Tracking Bug
    - LP: #1409794

  [ Upstream Kernel Changes ]

  * Drivers: hv: util: make struct hv_do_fcopy match Hyper-V host messages
    - LP: #1401903
  * rebase to v3.18.2
 -- Leann Ogasawara <email address hidden> Mon, 12 Jan 2015 08:53:12 -0800