Boris Ostrovsky writes ("Re: Nested virt broken in Linux 4.9 (was Re: [OSSTEST PATCH
2/3] ap-common: Switch to Linux 4.9 by default [and 1 more messages])"):
On 05/30/2017 10:28 AM, Ian Jackson wrote:
osstest service owner writes ("[osstest test] 109837: regressions - FAIL"):
Tests which did not succeed and are blocking,
including tests which could not be run:
test-amd64-amd64-qemuu-nested-amd 13 xen-boot/l1 fail REGR. vs. 109601
test-amd64-amd64-qemuu-nested-intel 13 xen-boot/l1 fail REGR. vs. 109601
The L1 console log is here:
...
The test is using 4.9.21 kernel and it looks like the patch that fixed
earlier regression shows up in 4.9 tree at 4.9.28 (commit
5d7ab8339a9a9e745c672279437657654268be81).
Thanks for investigating. osstest tries to to track 4.9.y.
However, it is blocked because of persistent failures like this one:
osstest service owner writes ("[linux-4.9 test] 109836: regressions - trouble:
broken/fail/pass"):
flight 109836 linux-4.9 real [real]
http://logs.test-lab.xenproject.org/osstest/logs/109836/
Regressions :-(
Tests which did not succeed and are blocking,
including tests which could not be run:
test-armhf-armhf-xl-credit2 6 xen-boot fail REGR. vs. 107358
Does anyone have any idea why this test should fail consistently ?
The corresponding test with the default credit1 scheduler fails too.
But it works on the other branches (which are using linux 3.18):
http://logs.test-lab.xenproject.org/osstest/results/history/test-armhf-armhf-xl/ALL
http://logs.test-lab.xenproject.org/osstest/results/history/test-armhf-armhf-xl-credit2/ALL
These are, admittedly, on the unreliable arndales, but:
http://logs.test-lab.xenproject.org/osstest/results/host/arndale-bluewater.html
So I think this is a problem with 4.9.