Bug 1702608 - [virtual-network][mq] prompt warning "qemu-kvm: unable to start vhost net: 14: falling back on userspace virtio" when boot with win8+ guests with multi-queue
Summary: [virtual-network][mq] prompt warning "qemu-kvm: unable to start vhost net: 14...
Keywords:
Status: POST
Alias: None
Product: Red Hat Enterprise Linux Advanced Virtualization
Classification: Red Hat
Component: qemu-kvm
Version: 8.0
Hardware: Unspecified
OS: Unspecified
high
high
Target Milestone: rc
: 8.0
Assignee: ybendito
QA Contact: Lei Yang
URL:
Whiteboard:
Depends On: 1608226
Blocks: 1649160 1824126
TreeView+ depends on / blocked
 
Reported: 2019-04-24 09:01 UTC by lijin
Modified: 2020-08-05 06:00 UTC (History)
20 users (show)

Fixed In Version:
Doc Type: If docs needed, set a value
Doc Text:
Clone Of: 1608226
: 1824126 (view as bug list)
Environment:
Last Closed:
Type: Bug
Target Upstream Version:


Attachments (Terms of Use)

Comment 1 lijin 2019-04-24 09:08:44 UTC
win8-32 hit the same issue with qemu-kvm-3.1.0-23.module+el8+3081+58d4aeb5.x86_64

qemu cli:
    -smp 32  \
    -device virtio-net-pci,mac=9a:f6:f7:f8:f9:fa,id=id4BOtfH,netdev=idMGpOHX,bus=pci.0,addr=0x5,mq=on,vectors=66  \
    -netdev tap,id=idMGpOHX,vhost=on,script=/etc/qemu-ifup,queues=32 \

As Yuri said, this issue can easily reproduced if queues are larger than guest cpu numbers:
-smp 32          ---> guest only has 2 cpus ---> this issue reproduced
-smp 32,cores=32 ---> guest has 32 cpus     ---> this issue NOT reproduced

Comment 2 Ademar Reis 2020-02-05 22:56:31 UTC
QEMU has been recently split into sub-components and as a one-time operation to avoid breakage of tools, we are setting the QEMU sub-component of this BZ to "General". Please review and change the sub-component if necessary the next time you review this BZ. Thanks

Comment 3 Peixiu Hou 2020-04-10 08:49:39 UTC
Hi,

I Hit the same issue on rhel8.2.0 slow train on win2019 guest.

qemu cli:
-cpu Skylake-Server,hv_stimer,hv_synic,hv_time,hv_relaxed,hv_vpindex,hv_spinlocks=0xfff,hv_vapic,hv_reset -enable-kvm 
-m 5800G -smp 128 \
-device pcie-root-port,port=0x10,chassis=1,id=pci.1,bus=pcie.0,multifunction=on,addr=0x3 \
-device pcie-root-port,port=0x11,chassis=2,id=pci.2,bus=pcie.0,addr=0x3.0x1 \
-device pcie-root-port,port=0x12,chassis=3,id=pci.3,bus=pcie.0,addr=0x3.0x2 \
-netdev tap,script=/etc/qemu-ifup1,id=hostnet0,vhost=on,queues=128 -device virtio-net-pci,netdev=hostnet0,id=net0,mac=00:52:52:22:1d:a8,mq=on,vectors=258,bus=pci.3 \

Used versions:
qemu-kvm-2.12.0-99.module+el8.2.0+5827+8c39933c.x86_64.
kernel-4.18.0-193.el8.x86_64

This bug is reported on RHEL8 Fast train only, if I need clone one on RHEL8 Slow train?

Thanks~
Peixiu

Comment 4 lulu@redhat.com 2020-04-15 02:12:02 UTC
(In reply to Peixiu Hou from comment #3)
> Hi,
> 
> I Hit the same issue on rhel8.2.0 slow train on win2019 guest.
> 
> qemu cli:
> -cpu
> Skylake-Server,hv_stimer,hv_synic,hv_time,hv_relaxed,hv_vpindex,
> hv_spinlocks=0xfff,hv_vapic,hv_reset -enable-kvm 
> -m 5800G -smp 128 \
> -device
> pcie-root-port,port=0x10,chassis=1,id=pci.1,bus=pcie.0,multifunction=on,
> addr=0x3 \
> -device pcie-root-port,port=0x11,chassis=2,id=pci.2,bus=pcie.0,addr=0x3.0x1 \
> -device pcie-root-port,port=0x12,chassis=3,id=pci.3,bus=pcie.0,addr=0x3.0x2 \
> -netdev tap,script=/etc/qemu-ifup1,id=hostnet0,vhost=on,queues=128 -device
> virtio-net-pci,netdev=hostnet0,id=net0,mac=00:52:52:22:1d:a8,mq=on,
> vectors=258,bus=pci.3 \
> 
> Used versions:
> qemu-kvm-2.12.0-99.module+el8.2.0+5827+8c39933c.x86_64.
> kernel-4.18.0-193.el8.x86_64
> 
> This bug is reported on RHEL8 Fast train only, if I need clone one on RHEL8
> Slow train?
> 
> Thanks~
> Peixiu

ThanksPeixu, please clone the bug on RHEL8 Slow train

Comment 7 Gal Hammer 2020-07-20 09:39:15 UTC
(In reply to Peixiu Hou from comment #3)
> Hi,
> 
> I Hit the same issue on rhel8.2.0 slow train on win2019 guest.

Which virtio-net-pci driver version is installed in the guest?

> qemu cli:
> -cpu
> Skylake-Server,hv_stimer,hv_synic,hv_time,hv_relaxed,hv_vpindex,
> hv_spinlocks=0xfff,hv_vapic,hv_reset -enable-kvm 
> -m 5800G -smp 128 \
> -device
> pcie-root-port,port=0x10,chassis=1,id=pci.1,bus=pcie.0,multifunction=on,
> addr=0x3 \
> -device pcie-root-port,port=0x11,chassis=2,id=pci.2,bus=pcie.0,addr=0x3.0x1 \
> -device pcie-root-port,port=0x12,chassis=3,id=pci.3,bus=pcie.0,addr=0x3.0x2 \
> -netdev tap,script=/etc/qemu-ifup1,id=hostnet0,vhost=on,queues=128 -device
> virtio-net-pci,netdev=hostnet0,id=net0,mac=00:52:52:22:1d:a8,mq=on,
> vectors=258,bus=pci.3 \
> 
> Used versions:
> qemu-kvm-2.12.0-99.module+el8.2.0+5827+8c39933c.x86_64.
> kernel-4.18.0-193.el8.x86_64
> 
> This bug is reported on RHEL8 Fast train only, if I need clone one on RHEL8
> Slow train?
> 
> Thanks~
> Peixiu

Comment 13 ybendito 2020-07-27 14:48:01 UTC
This is the regression in latest qemu.
Buggy commit is https://github.com/qemu/qemu/commit/f19bcdfedd53ee93412d535a842a89fa27cae7f2
Fix posted upstream
https://lists.nongnu.org/archive/html/qemu-devel/2020-07/msg07508.html


Note You need to log in before you can comment on or make changes to this bug.