Re: [Qemu-discuss] qemu 2.8.50 Assertion `ret == cpu->kvm_msr_buf->nmsrs' failed

[Bandan Das]

"Matwey V. Kornilov" <address@hidden> writes:

> 2017-02-03 0:13 GMT+03:00 Bandan Das <address@hidden>:
>> "Matwey V. Kornilov" <address@hidden> writes:
>>
>>> 02.02.2017 22:36, Matwey V. Kornilov пишет:
>>>>
>>>> Hello,
>>>>
>>>> I see the following when try to run qemu from master as the following:
>>>>
>>>> # ./x86_64-softmmu/qemu-system-x86_64 --version
>>>> QEMU emulator version 2.8.50 (v2.8.0-1006-g4e9f524)
>>>> Copyright (c) 2003-2016 Fabrice Bellard and the QEMU Project developers
>>>> # ./x86_64-softmmu/qemu-system-x86_64 -machine accel=kvm -nodefaults
>>>> -no-reboot -nographic -cpu host -vga none  -kernel .build.kernel.kvm
>>>> -initrd .build.initrd.kvm -append 'panic=1 no-kvmclock console=ttyS0
>>>> loglevel=7' -m 1024 -serial stdio
>>>> qemu-system-x86_64: /home/matwey/lab/qemu/target/i386/kvm.c:1849:
>>>> kvm_put_msrs: Assertion `ret == cpu->kvm_msr_buf->nmsrs' failed.
>>
>> I think this means that KVM gave you a list of msrs but
>> when you tried to set them, it could not set all of them and hence, the
>> assert triggers.
>>
>> Are you using a modified kernel ? There could be something useful in the host
>> dmesg as well.
>>
>
> I use vanilla 4.1.36
>
> Nothing relevant to kvm in dmesg except to
>
> [    5.265836] kvm: VM_EXIT_LOAD_IA32_PERF_GLOBAL_CTRL does not work
> properly. Using workaround

Can you try using -cpu host,-pmu ? Or do you need vPMU for something ?

> [   43.574791] kvm [1249]: vcpu0 unhandled wrmsr: 0x38d data 0
> [   43.580879] kvm [1249]: vcpu0 unhandled wrmsr: 0x38d data 0
> [   43.706797] kvm [1249]: vcpu0 unhandled rdmsr: 0x38d
>
>
>> Bandan
>>
>>>>
>>>>
>>>
>>>
>>> Hello, I've bisected the first bad commit:
>>>
>>>
>>> 48e1a45c3166d659f781171a47dabf4a187ed7a5 is the first bad commit
>>> commit 48e1a45c3166d659f781171a47dabf4a187ed7a5
>>> Author: Paolo Bonzini <address@hidden>
>>> Date:   Wed Mar 30 22:55:29 2016 +0200
>>>
>>>     target-i386: assert that KVM_GET/SET_MSRS can set all requested MSRs
>>>
>>>     This would have caught the bug in the previous patch.
>>>
>>>     Signed-off-by: Paolo Bonzini <address@hidden>
>>>
>>> :040000 040000 5626b58f0d79261bbe8a4fb52bd2cc19c2aa86ad
>>> d7580353932acab9de5c3d024b3f1215eb5659dc M      target-i386