Bug 1847230

Summary: QEMU refuse to start if set hmat=on and set vcpus for both numa nodes
Product: Red Hat Enterprise Linux 9 Reporter: Yumei Huang <yuhuang>
Component: qemu-kvmAssignee: Igor Mammedov <imammedo>
qemu-kvm sub component: Devices QA Contact: Yumei Huang <yuhuang>
Status: CLOSED CURRENTRELEASE Docs Contact:
Severity: high    
Priority: high CC: chayang, imammedo, jinqi, juzhang, mprivozn, plai, virt-maint
Version: unspecifiedKeywords: Triaged
Target Milestone: rcFlags: pm-rhel: mirror+
Target Release: ---   
Hardware: Unspecified   
OS: Unspecified   
Whiteboard:
Fixed In Version: Doc Type: If docs needed, set a value
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2021-12-16 07:27:02 UTC Type: Bug
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:
Bug Depends On:    
Bug Blocks: 1745059, 1749518, 1786303, 1897025    

Description Yumei Huang 2020-06-16 02:46:23 UTC
Description of problem:
Boot guest with hmat=on and two numa nodes, set vcpus for both nodes, qemu refuse to start and print error message.

Version-Release number of selected component (if applicable):
qemu-kvm-4.2.0-25.module+el8.2.1+6985+9fd9d514
kernel-4.18.0-193.7.1.el8_2.x86_64

How reproducible:
always

Steps to Reproduce:
1.  
# /usr/libexec/qemu-kvm -M pc,hmat=on -numa node,cpus=0-1 -numa node,cpus=2-3 -smp 4
qemu-kvm: -numa node,cpus=2-3: The initiator of CPU NUMA node 1 should be itself


If set initiator for node 1 as the error message says, still hit it.

# /usr/libexec/qemu-kvm -M pc,hmat=on -numa node,cpus=0-1 -numa node,cpus=2-3,initiator=1 -smp 4
qemu-kvm: -numa node,cpus=2-3,initiator=1: The initiator of CPU NUMA node 1 should be itself
 

Actual results:
QEMU quit with error message.

Expected results:
QEMU can start without error.

Additional info:
Reproducible on rhel8.3-av, qemu-kvm-5.0.0-0.module+el8.3.0+6620+5d5e1420.

Comment 1 John Ferlan 2020-06-17 14:01:20 UTC
Amnon - there isn't a NUMA specific qemu-kvm subcomponent, but the work seems to be someone from your team.

Comment 4 John Ferlan 2021-09-09 13:25:26 UTC
Bulk update: Move RHEL-AV bugs to RHEL9. If necessary to resolve in RHEL8, then clone to the current RHEL8 release.

Comment 6 RHEL Program Management 2021-12-16 07:27:02 UTC
After evaluating this issue, there are no plans to address it further or fix it in an upcoming release.  Therefore, it is being closed.  If plans change such that this issue will be fixed in an upcoming release, then the bug can be reopened.

Comment 8 Igor Mammedov 2021-12-29 15:11:18 UTC
Works for me on RHEL8.6 with

   qemu-kvm-6.2.0-1.module+el8.6.0+13725+61ae1949.x86_64

qemu-kvm -m 4G -object memory-backend-ram,id=z,size=4G -M pc,hmat=on -numa node,cpus=0-1,memdev=z -numa node,cpus=2-3 -smp 4

Comment 9 Yumei Huang 2021-12-30 02:23:55 UTC
Tested on rhel9.0, the issue is gone too. So change to CURRENTRELEASE.

qemu-kvm-6.2.0-1.el9
5.14.0-39.el9.x86_64

# /usr/libexec/qemu-kvm -M pc,hmat=on -m 4G -object memory-backend-ram,id=mem0,size=2G -object memory-backend-ram,id=mem1,size=2G -numa node,cpus=0-1,memdev=mem0 -numa node,cpus=2-3,memdev=mem1 -smp 4
VNC server running on ::1:5900