On Thu, Oct 01, 2020 at 06:52:05AM +0200, Thomas Huth wrote: > On 29/09/2020 18.43, Sean Christopherson wrote: > > Skip the Hyper-V tests on i386, they explicitly run with kvm64 and crash > > immediately when run in i386, i.e. waste 90 seconds waiting for the > > timeout to fire. > > > > Signed-off-by: Sean Christopherson <sean.j.christopherson@xxxxxxxxx> > > --- > > x86/unittests.cfg | 3 +++ > > 1 file changed, 3 insertions(+) > > > > diff --git a/x86/unittests.cfg b/x86/unittests.cfg > > index 3a79151..0651778 100644 > > --- a/x86/unittests.cfg > > +++ b/x86/unittests.cfg > > @@ -314,18 +314,21 @@ arch = x86_64 > > file = hyperv_synic.flat > > smp = 2 > > extra_params = -cpu kvm64,hv_vpindex,hv_synic -device hyperv-testdev > > +arch = x86_64 > > groups = hyperv > > > > [hyperv_connections] > > file = hyperv_connections.flat > > smp = 2 > > extra_params = -cpu kvm64,hv_vpindex,hv_synic -device hyperv-testdev > > +arch = x86_64 > > groups = hyperv > > > > [hyperv_stimer] > > file = hyperv_stimer.flat > > smp = 2 > > extra_params = -cpu kvm64,hv_vpindex,hv_time,hv_synic,hv_stimer -device hyperv-testdev > > +arch = x86_64 > > groups = hyperv > > Looks reasonable, but for some funny reason, this test seems to work on > Travis in the 32-bit builds: > > https://travis-ci.com/github/huth/kvm-unit-tests/jobs/392615222#L699 > > Any idea why it is still working there? Hmm, looks like these crash on my end for 64-bit as well, so maybe they are 32-bit compatible and it's my setup that's busted.
