On Thu, Aug 01, 2024 at 03:52:58PM GMT, Peter Krempa wrote:
> +++ b/tests/qemuxmlconfdata/riscv64-virt-acpi.xml
> @@ -0,0 +1,15 @@
> +<domain type='qemu'>
> + <name>guest</name>
> + <uuid>1ccfd97d-5eb4-478a-bbe6-88d254c16db7</uuid>
> + <memory>4194304</memory>
> + <vcpu>4</vcpu>
> + <os>
> + <type arch='riscv64' machine='virt'>hvm</type>
> + </os>
> + <features>
> + <acpi/>
> + </features>
> + <devices>
> + <emulator>/usr/bin/qemu-system-riscv64</emulator>
> + </devices>
> +</domain>
This (and s390x-ccw-acpi) could still do with explicitly dropping the
memballoon IMO, but it's fine either way.
> + /* tests for ACPI support handling:
> + * - existing positive test cases enabling ACPI for aarch64/x86_64/loongarch:
> + * - firmware-manual-efi-acpi-q35
> + * - firmware-manual-efi-acpi-aarch64
> + * - firmware-auto-efi-loongarch64
> + *
> + * - negative case for aarch64 with 'borzoi' machine not supporting ACPI
> + *
> + * - s390x has hack to strip ACPI to preserve migration of old configs,
> + * but should produce error when ABI_UPDATE is requested
> + */
> + DO_TEST_CAPS_ARCH_LATEST_PARSE_ERROR("aarch64-noacpi-acpi", "aarch64");
> + DO_TEST_CAPS_ARCH_LATEST("riscv64-virt-acpi", "riscv64");
> + DO_TEST_CAPS_ARCH_LATEST("s390x-ccw-acpi", "s390x");
> + DO_TEST_CAPS_ARCH_LATEST_ABI_UPDATE_PARSE_ERROR("s390x-ccw-acpi", "s390x");
Oh, you managed to increase coverage while at the same time reducing
the number of additional files! And the comment is quite detailed and
useful. Very nice indeed :)
Reviewed-by: Andrea Bolognani <abologna@xxxxxxxxxx>
--
Andrea Bolognani / Red Hat / Virtualization
