[vfio-users] vfio help needed (nvidia GTX 750 Ti persistent code 43)

[Mogliii]

I am not sure but I can only find vague/general recommendations in the
blog, but no particular models. I want to avoid using ACS patching so I
don't have to worry about it when updating the host kernel.

I'm currently considering the asrock  Z87 Extreme6 as it is reasonably
priced and listed in
https://en.wikipedia.org/wiki/List_of_IOMMU-supporting_hardware.
However, I am not clear if I would need ACS patch with it.

SXX on the google docs list (line 500) is using that motherboard but
with ACS patch. In line 426, SXX is using the motherboard without ACS
patch. The manual of the motherboard states
"""
V T- d
Intel® Virtualization Technology for Directed I/O helps your virtual
machine monitor better utilize hardware by improving application
compatibility andreliability, and providing additional levels of
manageability, security, isolation, andI/O performance.
"""

I don't want to buy the wrong one. Does anyone have an definite answer?

On 9/17/2015 5:12 PM, Blank Field wrote:
>
> We still have i*-4xxx cpus in stock in my country.(hell, even socket
> 775 is still can be found) So it depends on where do you live.
> You will need VT-d. Or AMD-Vi.
> All the recomendations are available on the blog, plus a very verbose
> guide on how to glue it all together.
>
> On Sep 17, 2015 11:05 AM, "Mogliii" <mogliii at gmx.net
> <mailto:mogliii at gmx.net>> wrote:
>
>     On 9/17/2015 2:24 PM, Blank Field wrote:
>>
>>     Read alex's blog at vfio.blogspot.com <http://vfio.blogspot.com>
>>     He has the same GPU and gives you hardware recommendations.
>>
>     I found it, but on the page it also says that "I wouldn't
>     necessarily recommend this particular setup (it's probably only
>     available on ebay anymore anyway)".
>     Any specific chipset I should aim for? I guess the CPU (i3, i5 or
>     i7) won't make a big difference as long as VTx is shown in the
>     datasheet? The google docs file shows many older models that are
>     not not sold anymore. Especially I would like to pick a
>     motherboard that has proper iommu separation of the primary
>     pcie-slot. So I don't need kernel patching.
>>
>>     But it should be possible to make it work, but it took me a year
>>     to figure out how.
>>
>     Sorry, I have other things to do within the next year ^^
>>     On Sep 17, 2015 8:21 AM, "Mogliii" <mogliii at gmx.net
>>     <mailto:mogliii at gmx.net>> wrote:
>>
>>         OK, now this is a new information for me. I thought all along
>>         that my problem is the graphics card.
>>
>>         I tried to understand your whole message but I got lost at
>>         MMIO and PIO.
>>         For the sake of saving my precious time and avoiding
>>         frustration, I should better get a new motherboard and CPU?
>>         How high are the chances it will work with my GXT750 Ti?
>>         Any hardware suggestions (I have DDR3-1600 memory that I
>>         would like to keep using)?
>>
>>         Thank you
>>
>>
>>
>>         On 9/17/2015 1:18 PM, Blank Field wrote:
>>>
>>>         You have an AMD FM2 CPU.
>>>         Do use SeaBIOS and make use of VGA I/O lines, you must have
>>>         a VideoBIOS Extension support in your GPU's firmware. Since
>>>         we are in the 21'st century, you have it.
>>>         But the problem is that in order to make VGA work in a VM we
>>>         must translate MMIO, PIO and IRQs needed into the VM.
>>>         That is where IOMMU kicks in, and the related part of
>>>         software is vfio-pci's x-vga option.
>>>         Usually you enable x-vga, vfio tells IOMMU to translate all
>>>         VGA related stuff from GPU to the VM and vice versa, but on
>>>         that particular platform(AMD FM2) something is broken in
>>>         hardware that crashes the IOMMU and the CPU when trying to
>>>         work with VGA.
>>>
>>>         On Sep 17, 2015 3:15 AM, "Alex Williamson"
>>>         <alex.williamson at redhat.com
>>>         <mailto:alex.williamson at redhat.com>> wrote:
>>>
>>>             On Wed, Sep 16, 2015 at 6:06 PM, Mogliii
>>>             <mogliii at gmx.net <mailto:mogliii at gmx.net>> wrote:
>>>
>>>                 Dear Alex,
>>>
>>>                 What do you mean you are suspicious of pci-stub?
>>>                 All tutorials I've seen so far require you to
>>>                 prevent nuveau from
>>>                 grabbing the card. Also in your very well written
>>>                 tutorial I followed
>>>                 and helped me a lot.
>>>                 (http://vfio.blogspot.jp/2015/05/vfio-gpu-how-to-series-part-3-host.html)
>>>
>>>                 I just tried the MSI GTX 750 Ti with Bios mode (bios
>>>                 selection switch ->
>>>                 1, followed by host reboot). But it behaves the
>>>                 same. I see the ovmf
>>>                 bios and boot splash, but then loss of signal.
>>>
>>>
>>>             It's not that you're using pci-stub, it's that pci-stub
>>>             is built as a module on your kernel and you need to be
>>>             sure to put all the pieces in place to make sure that
>>>             module gets loaded before anything else that might touch
>>>             the device.  As I said, it appears to be working, but
>>>             you're still getting a code 43. 
>>>
>>>             _______________________________________________
>>>             vfio-users mailing list
>>>             vfio-users at redhat.com <mailto:vfio-users at redhat.com>
>>>             https://www.redhat.com/mailman/listinfo/vfio-users
>>>
>>
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listman.redhat.com/archives/vfio-users/attachments/20150917/e569dbc2/attachment.htm>