Description of problem: Could not hotplug core to the target numa node Version-Release number of selected component (if applicable): qemu-kvm-3.1.0-3.module+el8+2614+d714d2bb.ppc64le kernel-4.18.0-57.el8.ppc64le How reproducible: 5/5 Steps to Reproduce: 1.boot up a rhel8.0 guest with similar cli ...-m 16G -smp 1,maxcpus=64,cores=1,threads=1,sockets=1 -numa node,nodeid=0 -numa node,nodeid=1 ... 2.hotplug a core to numa node 1 (qemu) info hotpluggable-cpus *Choose core 30 and 62 to hotplug, (qemu) device_add host-spapr-cpu-core,core-id=30,id=test30,node-id=1 node-id=1 must match numa node specified with -numa option (qemu) device_add host-spapr-cpu-core,core-id=62,id=test62,node-id=1 node-id=1 must match numa node specified with -numa option 3. Actual results: Fail to hotplug Expected results: Since node-id 0 owns core-id 0 to 31 and node-id 1 owns core-id 32 to 63,at least,it should work when hot-plug core 62 to numa node 1. Additional info:
x86 doesn't have this feature,so it's a specific ppc problem.
After discussing the bug with Laurent,it's output problem.So we had better make the error output more directly and effectively. Please change the message "node-id=1 must match numa node specified with -numa option" to an proper one.It looks like that it has no directly relation to cli's "-numa option" in this scenario.Thanks.
Sent patch upstream. numa: improve cpu hotplug error message with a wrong node-id https://patchwork.ozlabs.org/patch/1102588/
Patch merged upstream: a5bf9fbcc923 numa: improve cpu hotplug error message with a wrong node-id
Move to "POST" as the fix will come with the rebase to QEMU 4.1
Verified the bug on the following build qemu-kvm-4.1.0-1.module+el8.1.0+3966+4a23dca1.ppc64le The error message becomes more understandable and the original problem has been fixed,thanks a lot. (qemu) device_add host-spapr-cpu-core,core-id=30,id=test30,node-id=1 Error: invalid node-id, must be 0 (qemu) device_add host-spapr-cpu-core,core-id=30,id=test30,node-id=0 (qemu) device_add host-spapr-cpu-core,core-id=62,id=test62,node-id=1 Error: invalid node-id, must be 0 (qemu) device_add host-spapr-cpu-core,core-id=62,id=test62,node-id=0
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHBA-2019:3723