COLO's iperf3 test leads to "KVM: entry failed, hardware error 0x80000021" and segmentation fault

[Jing-Wei Su]

Hello, guys

I was testing QEMU-4.2.0's COLO using iperf3.

Primary VM is an iper3 receiver, and the client is an iperf3 sender.

PVM host information:
   model name : Intel(R) Core(TM) i5-4590S CPU @ 3.00GHz
   Linux 5.0.0-38-generic #41-Ubuntu SMP Tue Dec 3 00:27:35 UTC 2019
x86_64 x86_64 x86_64 GNU/Linux
   VM (guest) memory: 1024 MB

SVM host information:
   model name : Intel(R) Core(TM) i3-6100 CPU @ 3.70GHz
   Linux 5.3.0-42-generic #34~18.04.1-Ubuntu SMP Fri Feb 28 13:42:26
UTC 2020 x86_64 x86_64 x86_64 GNU/Linux
    VM (guest) memory: 1024 MB

After one hour iperf3 test, the SVM crashed and PVM's network IP is missing.
The QMP in SVM side shows

```
qemu-system-x86_64: warning: TSC frequency mismatch between VM
(2993069 kHz) and host (3696011 kHz), and TSC scaling unavailable
{"timestamp": {"seconds": 1584685696, "microseconds": 736356},
"event": "RESUME"}
{"timestamp": {"seconds": 1584685716, "microseconds": 642200}, "event": "STOP"}
qemu-system-x86_64: warning: TSC frequency mismatch between VM
(2993069 kHz) and host (3696011 kHz), and TSC scaling unavailable
{"timestamp": {"seconds": 1584685716, "microseconds": 995300},
"event": "RESUME"}
{"timestamp": {"seconds": 1584685729, "microseconds": 166332}, "event": "STOP"}
qemu-system-x86_64: warning: TSC frequency mismatch between VM
(2993069 kHz) and host (3696011 kHz), and TSC scaling unavailable
{"timestamp": {"seconds": 1584685729, "microseconds": 257427},
"event": "RESUME"}
{"timestamp": {"seconds": 1584685749, "microseconds": 165485}, "event": "STOP"}
qemu-system-x86_64: warning: TSC frequency mismatch between VM
(2993069 kHz) and host (3696011 kHz), and TSC scaling unavailable
{"timestamp": {"seconds": 1584685749, "microseconds": 466703},
"event": "RESUME"}
KVM: entry failed, hardware error 0x80000021

If you're running a guest on an Intel machine without unrestricted mode
support, the failure can be most likely due to the guest entering an invalid
state for Intel VT. For example, the guest maybe running in big real mode
which is not supported on less recent Intel processors.

RAX=0000000080080783 RBX=ffff8972b4060000 RCX=0000000000000002
RDX=ffffa7e2c05a0000
RSI=ffff8972b4060000 RDI=000000000000000b RBP=ffff8972bfc03c70
RSP=ffff8972bfc03c60
R8 =0000000000000000 R9 =0000000000000000 R10=0000000000000000
R11=0000000000000000
R12=0000000000000000 R13=ffffffff82e5d560 R14=ffffffff82c03d78
R15=ffff8972b46b2200
RIP=ffffffffc03b74dc RFL=00000046 [---Z-P-] CPL=0 II=0 A20=1 SMM=0 HLT=0
ES =0000 0000000000000000 000fffff 00000000
CS =0010 0000000000000000 ffffffff 00a09b00 DPL=0 CS64 [-RA]
SS =0018 0000000000000000 ffffffff 00c09300 DPL=0 DS   [-WA]
DS =0000 0000000000000000 000fffff 00000000
FS =0000 0000000000000000 ffffffff 00000000
GS =0000 ffff8972bfc00000 000fffff 00000000
LDT=0000 0000000000000000 000fffff 00000000
TR =0040 fffffe0000003000 0000206f 00008b00 DPL=0 TSS64-busy
GDT=     fffffe0000001000 0000007f
IDT=     fffffe0000000000 00000fff
CR0=80050033 CR2=00005626cf302028 CR3=000000003d238000 CR4=000006f0
DR0=0000000000000000 DR1=0000000000000000 DR2=0000000000000000
DR3=0000000000000000
DR6=00000000ffff0ff0 DR7=0000000000000400
EFER=0000000000000d01
Code=ff ff ff ff 89 82 d8 00 00 00 48 8b 83 68 0c 00 00 8b 40 08 <4c>
8d a3 70 0b 00 00 4c 89 e7 e8 15 de ca c1 84 c0 74 63 48 c7 83 e0 0a
00 00 00 00 00 00
{"timestamp": {"seconds": 1584685749, "microseconds": 472013}, "event": "STOP"}

qemu-system-x86_64: invalid runstate transition: 'internal-error' -> 'colo'

secondary-nonshared.sh: line 34:  9052 Aborted                 (core
dumped) qemu-system-x86_64 -name secondary -enable-kvm -cpu
qemu64,+kvmclock -m 1024 -global kvm-apic.vapic=false -netdev
tap,id=hn0,vhost=off,helper=/usr/local/libexec/qemu-bridge-helper
-device e1000,id=e0,netdev=hn0 -chardev
socket,id=red0,host=${primary_ip},port=9003,reconnect=1 -chardev
socket,id=red1,host=${primary_ip},port=9004,reconnect=1 -object
filter-redirector,id=f1,netdev=hn0,queue=tx,indev=red0 -object
filter-redirector,id=f2,netdev=hn0,queue=rx,outdev=red1 -object
filter-rewriter,id=rew0,netdev=hn0,queue=all -drive
if=none,id=parent0,file.filename=$imagefolder/secondary.qcow2,driver=qcow2
-drive 
if=none,id=childs0,driver=replication,mode=secondary,file.driver=qcow2,top-id=colo-disk0,file.file.filename=$imagefolder/secondary-active.qcow2,file.backing.driver=qcow2,file.backing.file.filename=$imagefolder/secondary-hidden.qcow2,file.backing.backing=pa
```

There are some errors in above logs
"KVM: entry failed, hardware error 0x80000021"
invalid runstate transition: 'internal-error' -> 'colo'

Should the CPU on both PVM and SVM  be identical for the COLO?
Or, is the error caused by other issues?

Thanks.
Sincerely,

Jing-Wei