Re: [PATCH v11 2/2] qtest/cxl: Add aarch64 virt test for CXL

[Jonathan Cameron via <qemu-devel@nongnu.org>]

On Fri, 24 Jun 2022 17:12:25 +0100
Peter Maydell <peter.maydell@linaro.org> wrote:

> On Thu, 16 Jun 2022 at 15:20, Jonathan Cameron
> <Jonathan.Cameron@huawei.com> wrote:
> >
> > Add a single complex case for aarch64 virt machine.
> >
> > Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
> > ---
> >  tests/qtest/cxl-test.c  | 48 +++++++++++++++++++++++++++++++++--------
> >  tests/qtest/meson.build |  1 +
> >  2 files changed, 40 insertions(+), 9 deletions(-)
> >
> > diff --git a/tests/qtest/cxl-test.c b/tests/qtest/cxl-test.c
> > index 2133e973f4..1015d0e7c2 100644
> > --- a/tests/qtest/cxl-test.c
> > +++ b/tests/qtest/cxl-test.c
> > @@ -17,6 +17,11 @@
> >                        "-device pxb-cxl,id=cxl.1,bus=pcie.0,bus_nr=53 " \
> >                        "-M cxl-fmw.0.targets.0=cxl.0,cxl-fmw.0.targets.1=cxl.1,cxl-fmw.0.size=4G "
> >
> > +#define QEMU_VIRT_2PXB_CMD "-machine virt,cxl=on "                      \
> > +                      "-device pxb-cxl,id=cxl.0,bus=pcie.0,bus_nr=52 "  \
> > +                      "-device pxb-cxl,id=cxl.1,bus=pcie.0,bus_nr=53 "  \
> > +                      "-M cxl-fmw.0.targets.0=cxl.0,cxl-fmw.0.targets.1=cxl.1,cxl-fmw.0.size=4G "
> > +  
> 
> If CXL requires booting via UEFI, what does this test case do?
> It doesn't seem to be passing in a BIOS image.

Not a lot beyond checking device creation is valid etc and the machine boots.

There is a bios tables test that checks we pass the right tables to the BIOS image.
I didn't duplicate that for ARM on the basis it's more or less identical, but
perhaps that is worth adding.

To do any useful functional testing will require a mass of complex OS
handling after booting. That testing is definitely something I'd like to
add, but the userspace tooling isn't all in place yet. Final kernel series that's
needed to get to the point where you can use the non volatile memory had
a new version posted yesterday.

Jonathan

> 
> thanks
> -- PMM