Noticed the following command failure while testing CPU hotplug. $ qemu-system-x86_64 -machine q35,accel=kvm -smp 1,maxcpus=2, cores=1, threads=1,sockets=2 -cpu EPYC -device EPYC-x86_64- cpu,core-id=0,socket-id=1,thread-id=0
qemu-system-x86_64: -device EPYC-x86_64-cpu,core-id=0,socket-id=1, thread-id=0: Invalid CPU [socket: 21855, die: 0, core: 0, thread: 0] with APIC ID 21855, valid index range 0:1 This happens because APIC ID is calculated using uninitialized memory. This is happening after the addition of new field node_id in X86CPUTopoIDs structure. The node_id field is uninitialized while calling apicid_from_topo_ids. The problem is discussed in the thread below. https://lore.kernel.org/qemu-devel/20200602171838.gg577...@habkost.net/ Fix the problem by initializing the node_id from the device being added. Fixes: Link: https://bugzilla.redhat.com/show_bug.cgi?id=1828750 Signed-off-by: Babu Moger <babu.mo...@amd.com> --- hw/i386/pc.c | 10 ++++++++++ 1 file changed, 10 insertions(+) diff --git a/hw/i386/pc.c b/hw/i386/pc.c index e613b2299f..aa9fb48834 100644 --- a/hw/i386/pc.c +++ b/hw/i386/pc.c @@ -1553,6 +1553,15 @@ static void pc_cpu_pre_plug(HotplugHandler *hotplug_dev, cpu->die_id = 0; } + /* + * If node_id is not set, initialize it to zero for now. If the user + * does not pass the correct node in case of numa configuration, it + * will be rejected eventually. + */ + if (cpu->node_id < 0) { + cpu->node_id = 0; + } + if (cpu->socket_id < 0) { error_setg(errp, "CPU socket-id is not set"); return; @@ -1587,6 +1596,7 @@ static void pc_cpu_pre_plug(HotplugHandler *hotplug_dev, } topo_ids.pkg_id = cpu->socket_id; + topo_ids.node_id = cpu->node_id; topo_ids.die_id = cpu->die_id; topo_ids.core_id = cpu->core_id; topo_ids.smt_id = cpu->thread_id;
