On Fri, Jul 02, 2021 at 06:07:38PM +0800, Yanan Wang wrote:
Since a machine type does not support topology parameter of dies,
it's probably more reasonable to reject any explicit specification
to avoid possible confuse, including "dies=0" and "dies=1" although
they won't affect the calculation of non-PC machines.
Also a comment of struct SMPConfiguration is fixed.
Signed-off-by: Yanan Wang <wangyanan55@huawei.com>
---
hw/core/machine.c | 2 +-
qapi/machine.json | 2 +-
2 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/hw/core/machine.c b/hw/core/machine.c
index 58882835be..55785feee2 100644
--- a/hw/core/machine.c
+++ b/hw/core/machine.c
@@ -747,7 +747,7 @@ static void smp_parse(MachineState *ms, SMPConfiguration
*config, Error **errp)
unsigned threads = config->has_threads ? config->threads : 0;
unsigned maxcpus = config->has_maxcpus ? config->maxcpus : 0;
- if (config->has_dies && config->dies != 0 && config->dies != 1) {
+ if (config->has_dies) {
error_setg(errp, "dies not supported by this machine's CPU topology");
}
This will cause a functional regression. Libvirt always specifies
dies=1 if QEMU supports it. I don't see a need to reject it,
since conceptually it is reasonable to say that all existing
CPUs have a single die. It simply that 99% of CPUs don't support
more than 1 die.