Re: [PULL 31/33] tests/acpi: add test case for VIOT

[Peter Maydell]

On Thu, 16 Dec 2021 at 09:58, Jean-Philippe Brucker
<jean-philippe@linaro.org> wrote:
>
> On Wed, Dec 15, 2021 at 04:59:10PM -0800, Richard Henderson wrote:
> > On 12/15/21 2:40 AM, Peter Maydell wrote:
> > > From: Jean-Philippe Brucker <jean-philippe@linaro.org>
> > >
> > > Add two test cases for VIOT, one on the q35 machine and the other on
> > > virt. To test complex topologies the q35 test has two PCIe buses that
> > > bypass the IOMMU (and are therefore not described by VIOT), and two
> > > buses that are translated by virtio-iommu.
> > >
> > > Reviewed-by: Eric Auger <eric.auger@redhat.com>
> > > Reviewed-by: Igor Mammedov <imammedo@redhat.com>
> > > Signed-off-by: Jean-Philippe Brucker <jean-philippe@linaro.org>
> > > Message-id: 20211210170415.583179-7-jean-philippe@linaro.org
> > > Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
> > > ---
> > >   tests/qtest/bios-tables-test.c | 38 ++++++++++++++++++++++++++++++++++
> > >   1 file changed, 38 insertions(+)
> >
> > I should have been more careful while applying.  The aarch64 host failure
> > for this is not transient as I first assumed:
> >
> > PASS 5 qtest-aarch64/bios-tables-test /aarch64/acpi/virt/oem-fields
> > qemu-system-aarch64: kvm_init_vcpu: kvm_arch_init_vcpu failed (0): Invalid 
> > argument
> > Broken pipe
> > ERROR qtest-aarch64/bios-tables-test - too few tests run (expected 6, got 5)
> > make: *** [Makefile.mtest:312: run-test-37] Error 1
>
> I'm guessing adding "tcg_only = true", like all other virt machine tests
> in this file, should work around this, but I don't really understand the
> problem because I can't reproduce it on my aarch64 host (I'm running
> "configure --target-list=aarch64-softmmu" followed by "make -j10
> check-qtest V=1" in a loop)

What host are you testing on? The test case explicitly asks
for "-cpu cortex-a57", so it is only going to work with KVM
on hosts with a Cortex-A57.

Richard: given I'm off work for Christmas now, if Jean-Philippe's
suggested fix fixes this are you OK with just applying it directly
to un-break the CI ?

thanks
-- PMM