Am 05.12.2017 um 20:31 hat Daniel Henrique Barboza geschrieben: > On 12/05/2017 04:06 PM, Dr. David Alan Gilbert wrote: > > * Daniel Henrique Barboza (danie...@linux.vnet.ibm.com) wrote: > > > This patch implements a test case for the scenario that was failing > > > prior to the patch "migration/ram.c: do not set 'postcopy_running' in > > > POSTCOPY_INCOMING_END", commit acab30b85d. > > > > > > This new test file 201 was derived from the test file 181 authored > > > by Kevin Wolf. > > The test is interesting because it enables postcopy but never triggers > > postcopy entry (using migrate_start_postcopy). Still it's a good path > > to check. > > > > It will fail to enable postcopy on older linux hosts or non-linux hosts > > that don't have userfault though; I'm not sure how your test fails > > with that. > Hmmm I am not sure either. I have tested it with Travis and it passed all > tests, > including a claimed Mac OS environment. Not sure if that counts. > > I'll see if I can get this tested on a Windows env to see how this test > fails.
qemu-iotests is basically Linux-only and we don't even try to keep things portable. If it works on Windows, that's more by accident than intentional. Kevin
