Re: [PATCH v2 3/6] hw/acpi: Generic Port Affinity Structure support

[Richard Henderson]

On 6/5/24 11:21, Jonathan Cameron via wrote:
> On Fri, 24 May 2024 11:05:04 +0100
> Jonathan Cameron <Jonathan.Cameron@huawei.com> wrote:
>
> > These are very similar to the recently added Generic Initiators
> > but instead of representing an initiator of memory traffic they
> > represent an edge point beyond which may lie either targets or
> > initiators.  Here we add these ports such that they may
> > be targets of hmat_lb records to describe the latency and
> > bandwidth from host side initiators to the port.  A descoverable
> > mechanism such as UEFI CDAT read from CXL devices and switches
> > is used to discover the remainder fo the path and the OS can build
> > up full latency and bandwidth numbers as need for work and data
> > placement decisions.
> >
> > Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
>
> To join up the streams.  The tests added in this series failed
> CI on s390 bios-tables-test.
>
> ad6d572b-f39e-43ff-b11b-74fbe8ae3148@linaro.org/T/#m0f6531d67ba28663bd35b359e32ddfea42db2dea">https://lore.kernel.org/qemu-devel/ad6d572b-f39e-43ff-b11b-74fbe8ae3148@linaro.org/T/#m0f6531d67ba28663bd35b359e32ddfea42db2dea
>
> has my current theory on why and Richard is grabbing the SRAT table
> which will hopefully have this as the smoking gun.
>
> Comes back to my normal question to management.  Can I have an s390
> for tests?  Where are those up to date big endian test boxes for
> every developer to have on their desks?

In this particular case, it's easy to reproduce within an emulated s390x chroot.

Annoyingly, while the gcc compile farm has a big-endian ppc64 host, the OS install is 
obsolete centos 7 so it's no longer possible to build there.  Nor is there an s390x.  So 
there are no usable big-endian hosts there.

We have an s390x for qemu CI use, obviously, but I can't give out access to 
that.


r~