[vfio-users] vfio passthrough devices behind pcie switcher problem

[rhett rhett]

hi Alex， i finally fixed this problem. that's because my code lacks pci
virtual channel support.

after i add drivers\pci\vc.c to support this , everythings runs ok.

thanks a lot.



2017-03-13 10:29 GMT+08:00 rhett rhett <rhett.kernel at gmail.com>:

> it's using qemu 1.2 , i tried the newest qemu , version 2.8 , the problem
> remains.
>
> 2017-03-10 22:08 GMT+08:00 Alex Williamson <alex.williamson at redhat.com>:
>
>> On Fri, 10 Mar 2017 10:42:28 +0800
>> rhett rhett <rhett.kernel at gmail.com> wrote:
>>
>> > sorry , update infomation right now
>> >
>> >
>> > i installed centos7.3 at my 8 gpus machine yesterday, and i made a
>> > successful passthrough, the vm guest os can use gpu with no problem. so
>> i
>> > think this is a software problem, i need to patch some patch.
>> >
>> > i also made a test in my 4 gpus machine without any software change, the
>> > result is success. the 4 gpus are attached at pci root without pcie
>> switch
>> > , so , i think the software problem have some correlation with pcie
>> switch .
>> >
>> > thank you Alex .
>> >
>> > [root at 64 /data]# lspci|grep NV
>> > 04:00.0 3D controller: NVIDIA Corporation Device 17fd (rev a1)
>>
>> Tesla M40
>>
>> > 05:00.0 3D controller: NVIDIA Corporation Device 1b38 (rev a1)
>>
>> Tesla P40
>>
>> > 08:00.0 3D controller: NVIDIA Corporation Device 1b38 (rev a1)
>> > 09:00.0 3D controller: NVIDIA Corporation Device 1b38 (rev a1)
>> > 85:00.0 3D controller: NVIDIA Corporation Device 1b38 (rev a1)
>> > 86:00.0 3D controller: NVIDIA Corporation Device 17fd (rev a1)
>> > 89:00.0 3D controller: NVIDIA Corporation Device 1b38 (rev a1)
>> > 8a:00.0 3D controller: NVIDIA Corporation Device 1b38 (rev a1)
>>
>> Note that you have all of these in 3D mode, you may need to put these
>> into VGA mode.  NVIDIA provides tools to do this in the host.
>>
>> > [root at 64 /data]# lspci -t
>> > -+-[0000:ff]-+-08.0
>> >  |           +-08.2
>> >  |           +-08.3
>> >  |           +-09.0
>>
>> [snip... not useful w/o info per devices]
>>
>> > and the xml
>> >
>> > <domain type='kvm'>
>> >   <name>win</name>
>> >   <uuid>a2021423-89d8-4a33-aaa5-07102ae7ad4e</uuid>
>> >   <memory unit='KiB'>8388608</memory>
>> >   <currentMemory unit='KiB'>8388608</currentMemory>
>> >   <vcpu placement='static' cpuset='0-8'>8</vcpu>
>> >   <sysinfo type='smbios'>
>> >     <system>
>> >       <entry name='serial'>21aa32e5-8233-40d4-b323-128824f6becf</entry>
>> >       <entry name='uuid'>a2021423-89d8-4a33-aaa5-07102ae7ad4e</entry>
>> >     </system>
>> >   </sysinfo>
>> >   <os>
>> >     <type arch='x86_64' machine='pc'>hvm</type>
>> >     <boot dev='hd'/>
>> >     <smbios mode='sysinfo'/>
>> >   </os>
>> >   <features>
>> >     <acpi/>
>> >     <apic/>
>> >     <pae/>
>> >     <hap/>
>> >     <hyperv>
>> >       <relaxed state='on'/>
>> >     </hyperv>
>> >   </features>
>> >   <cpu>
>> >     <topology sockets='2' cores='6' threads='2'/>
>> >   </cpu>
>> >   <clock offset='localtime'>
>> >     <timer name='pit' tickpolicy='delay'/>
>> >     <timer name='rtc' tickpolicy='catchup' track='guest'/>
>> >     <timer name='hpet' present='no'/>
>> >   </clock>
>> >   <on_poweroff>destroy</on_poweroff>
>> >   <on_reboot>restart</on_reboot>
>> >   <on_crash>restart</on_crash>
>> >   <devices>
>> >     <emulator>/usr/local/bin/qemu-system-x86_64</emulator>
>>
>> Running custom QEMU, what version is this?
>>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listman.redhat.com/archives/vfio-users/attachments/20170321/33fe20b6/attachment.htm>