RTC? test time-out (linux 4.9.0-12.13 ADT test failure with linux 4.9.0-12.13)

[Andy Whitcroft]

Testing failed on:
    amd64: https://objectstorage.prodstack4-5.canonical.com/v1/AUTH_77e2ada1e7a84929a74ba3b87153c0ac/autopkgtest-zesty/zesty/amd64/l/linux/20170122_155800_770b2@/log.gz

[Brad Figg (brad-figg)]

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

apport-collect 1658635

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.

[Andy Whitcroft (apw)]

Looks like the RTC test is timing us out again (maybe):

11:59:18 DEBUG| [stdout] selftests: threadtest [PASS]
11:59:18 ERROR| [stderr]
11:59:18 ERROR| [stderr] RTC Driver Test Example.
11:59:18 ERROR| [stderr]
autopkgtest [15:57:37]: ERROR: timed out on command "su -s /bin/bash ubuntu -c [...]"

[Colin Ian King (colin-king)]

I suspect this is because the test is doing a blocking read on /dev/rtc* and the test will block until 1 second wakeups get triggered. In this case, it seems like the RTC_UIE_ON ioctl did not trigger the wakeup or the wakeup didn't fire on the VM and the ADT test timeout out after 4 hrs.

[Colin Ian King (colin-king)]

@apw, how are you invoking these tests on prodstack? I can't reproduce this issue across the VMs I'm using.

[Andy Whitcroft (apw)]

@cking mostly using magic, i can trigger new ones etc if that helps.

[Colin Ian King (colin-king)]

OK, can you run a few more tests and we can see how variable this is, e.g, does it trigger every time or not.

[Seth Forshee (sforshee)]

@cking: I think this is the same problem I looked into before, see bug #1649718. However I completely dropped the ball on SRU-ing the patches.

tl;dr, the problem is with the host kernel for the VM losing irq acks from the guest and then failing to trigger any subsequent interrupts. So it's a problem in xenial and not zesty.

I will get SRU patches sent to the list for the other bug. You'll probably just want to confirm that this one is the same problem then dup it to that bug.

[Colin Ian King (colin-king)]

@Seth I'll give it a spin. Thanks!

[Colin Ian King (colin-king)]

Let's see how the SRU fixes work on this one and re-test. I'm having issues trying to reproduce this still, my gut feeling is that we get this SRU'd and see if it fixes the issue rather than me waste more cycles trying to verify if the fix does the trick when it seems clear that the fix in bug #1649718 addresses the issue.

[Colin Ian King (colin-king)]

I don't see any issue now with 4.10.0-8 (i386 + amd64), so marking this as fixed released.