On Fri, May 06, 2016 at 06:39:15 -0400, John Ferlan wrote: > > > On 05/06/2016 06:29 AM, Peter Krempa wrote: > > On Fri, May 06, 2016 at 12:23:04 +0200, Jiri Denemark wrote: > >> On Fri, May 06, 2016 at 10:02:12 +0200, Peter Krempa wrote: > >>> On Thu, May 05, 2016 at 18:42:13 +0200, Jiri Denemark wrote: > >>>> Signed-off-by: Jiri Denemark <jdenemar@xxxxxxxxxx> > >>> > >>> Please explain the difference between caps_2.6.0-1.aarch64.replies > >>> caps_2.6.0-2.aarch64.replies. At least the difference how they were > >>> obtained. > >> > >> Well, I asked Andrea to provide me some data from aarch64 and ppc64le > >> hosts :-) The difference is in GIC version supported on a two different > >> hosts. > > > > Well I kind of implied that you should put it into the commit message > > for further reference. I don't really care at this point about the > > difference but somebody might care in the future when touching the > > tests. > > > > > > FWIW: For example, is there a process or script (or something) somewhere > that you use in order to generate the files? Yes, "tests/qemucapsprobe /path/to/binary >caps.replies". It's mainly useful for adding data files. It can be used for updating existing ones in case you still have the binary and can just re-probe it. If you want to add staff to existing files manually, qemucapsprobe won't help you. But, Peter is planning to enhance the fake monitor and *.replies data format so that you don't have to care about ids or order (to some extent) in which you add the new replies. Jirka -- libvir-list mailing list libvir-list@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/libvir-list