Hi Martin, On 03/04/2016 09:31 PM, Martin Pitt wrote: > Christian Seiler [2016-03-04 18:17 +0100]: >> 0001-Support-nested-KVM-by-default-by-emulating-a-CPU-wit.patch > > Fixed the fd leak,
Oops, sorry! ;-) (I really should work with python more regularly.) > > http://anonscm.debian.org/cgit/autopkgtest/autopkgtest.git/commit/?id=3a9cb0b9 > > http://anonscm.debian.org/cgit/autopkgtest/autopkgtest.git/commit/?id=a46c7a69 > > http://anonscm.debian.org/cgit/autopkgtest/autopkgtest.git/commit/?id=b6557eb02 Thanks for committing this! :-) > Thank you! I'm running the full testsuite now, do some finishing > touches, and will do a new upload RSN. Ok, then I'm going to write some tests that use this feature now. Thanks a lot for making this possible. Question: I think we should also define a canonical way of detecting baseimage support. I propose to use: [ -e /dev/baseimage ] Because that is very agnostic to details: if for example some cloud runner also wants to support this in the future, but provides access to a file instead of a block device, they can symlink that to /dev/baseimage and this test would still catch that. Using -b would be suboptimal. Do you agree? Regards, Christian
signature.asc
Description: OpenPGP digital signature
