Re: [Xen-devel] linux 4.9 fails with nested Xen

[Boris Ostrovsky <boris.ostrovsky@xxxxxxxxxx>]

On 04/21/2017 12:06 PM, Boris Ostrovsky wrote:
> On 04/21/2017 09:21 AM, Ian Jackson wrote:
>> osstest service owner writes ("[linux-4.9 test] 107573: regressions - FAIL"):
>>> flight 107573 linux-4.9 real [real]
>>> http://logs.test-lab.xenproject.org/osstest/logs/107573/
>> ...
>>>  test-amd64-amd64-qemuu-nested-intel 13 xen-boot/l1       fail never pass
>>>  test-amd64-amd64-qemuu-nested-amd 13 xen-boot/l1         fail never pass
>> For example,
>>   
>> http://logs.test-lab.xenproject.org/osstest/logs/107573/test-amd64-amd64-qemuu-nested-intel/info.html
>>
>> The L1 seems to stop producing serial output after
>>
>>   [  106.376164] random: crng init done
>>   6.080169] random: crng init done
>>
>> but it does produce a login prompt.
>>
>> This passes in 3.14.  I would like to update osstest to use a recent
>> kernel by default, but this failure is blocking that.
>>
>> Could someone investigate please ?
>>
>> Let me know if you need more info, or for me to set up a repro for you
>> to poke at.
>
> Let me see if I can reproduce it locally. I know 4.11-rc<something>
> works since I've been using it recently.

This is the same failure that was discussed earlier --- lack of commit
da72ff5bfcb02 (partially revert "xen: Remove event channel notification
through Xen PCI platform device") in 4.9 tree.

This actual offending commit (72a9b186292d ("xen: Remove event channel
notification through Xen PCI platform device")) will be removed from 4.9
as soon as we revert it in Linus' tree.


-boris

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
https://lists.xen.org/xen-devel