Re: [vfio-users] Guest does not boot OS with GPU passed through, stuck at TianoCore boot splash

[Will Marler]

On Mon, Apr 18, 2016 at 8:46 AM, Alex Williamson <
alex.l.william...@gmail.com> wrote:

> The 750 and 750Ti are not only a good value, but have very modest power
> requirements, so it's a win all around for less demanding/casual gaming.


As a less-demanding/casual gamer with a 750Ti, I fully endorse this
endorsement.
___
vfio-users mailing list
vfio-users@redhat.com
https://www.redhat.com/mailman/listinfo/vfio-users

Re: [vfio-users] Guest does not boot OS with GPU passed through, stuck at TianoCore boot splash

[Alex Williamson]

On Mon, Apr 18, 2016 at 8:35 AM, Blank Field 
wrote:

>
> On Apr 18, 2016 5:31 PM, "Laszlo Ersek"  wrote:
>
> > I'm very pleased with my GTX750.
>
> I guess most 7XX nvidia cards work rather well, no? 750 being the most
> balanced in performance-per-cost.
>

I'd generally suggest Kepler or newer with UEFI support in the ROM.  That
UEFI part can be dicey on Kepler, but is more prevalent on Maxwell (ex. I
had to ask for a UEFI ROM for my EVGA GTX660, but the EVGA GTX750 came with
it).  The 750 and 750Ti are not only a good value, but have very modest
power requirements, so it's a win all around for less demanding/casual
gaming.
___
vfio-users mailing list
vfio-users@redhat.com
https://www.redhat.com/mailman/listinfo/vfio-users

Re: [vfio-users] Guest does not boot OS with GPU passed through, stuck at TianoCore boot splash

[Laszlo Ersek]

On 04/18/16 16:08, Stefan Seil wrote:
> Thank you for your detailed response!
> 
> So if I understand you correctly, you are saying that from now on the
> issue lies within the guest OS,

Yes.

> and thus I am rather limited in what I
> can do on my end.

You could talk to Linux kernel maintainers that own the graphics drivers
for your card, I guess :)

> The good news is that I have actually gotten it to
> work with a Windows 10 guest. Running Windows was my actual goal, so not
> being able to pass the GPU to a Linux guest is not the end of the world
> for me.

Okay.

> I realized that the current setup is probably not very practical for me
> in the long run, though, as I now only have the integrated Intel GPU for
> my host to use, which will be quite a limitation for games and other
> graphics-intense applications. Therefore I am now considering getting a
> second card. You mentioned the GTX 750 (and Alex seems to have used this
> one successfully, too). Does it work well for you? If so, I might just
> try to get one of those. It should also be more or less on the same
> level of performance compared to the HD 7850 I am currently using.

I'm very pleased with my GTX750.

Thanks
Laszlo

___
vfio-users mailing list
vfio-users@redhat.com
https://www.redhat.com/mailman/listinfo/vfio-users

Re: [vfio-users] Guest does not boot OS with GPU passed through, stuck at TianoCore boot splash

[Stefan Seil]

Thank you for your detailed response!

So if I understand you correctly, you are saying that from now on the 
issue lies within the guest OS, and thus I am rather limited in what I 
can do on my end. The good news is that I have actually gotten it to 
work with a Windows 10 guest. Running Windows was my actual goal, so not 
being able to pass the GPU to a Linux guest is not the end of the world 
for me.


I realized that the current setup is probably not very practical for me 
in the long run, though, as I now only have the integrated Intel GPU for 
my host to use, which will be quite a limitation for games and other 
graphics-intense applications. Therefore I am now considering getting a 
second card. You mentioned the GTX 750 (and Alex seems to have used this 
one successfully, too). Does it work well for you? If so, I might just 
try to get one of those. It should also be more or less on the same 
level of performance compared to the HD 7850 I am currently using.


On Wed, Apr 31, 2016 at 12:03 AM, Laszlo Ersek wrote:

On 04/12/16 20:42, Stefan Seil wrote:

As far as I can tell, the other PCI device is the HDMI audio controller
of the graphics card. It is listed as a seperate PCI device and I
assigned it to the VM, too.

I think that makes sense. I have the same situation with my GTX750. I
believe those two functions belong to the same IOMMU domain, so they
have to be assigned together anyway.


I haven't explicitly added any USB devices, although it seems like there
are some per default. I have linked the respective domain XMLs below.

Hmm no, as far as device assignment goes, you have only two 
elements.

Regarding USB devices, you have a few emulated controllers. As far as
actual devices go, you seem to be using Spice USB redirection (see
). I had to read up on that:
.

I can't tell if you actually redirected some host-side USB devices. I
think it's unlikely; Cole's blog post says "When connected to a VM's
graphical display", and in your first domain XML, your only display is
the assigned GPU (no  element).


I have played around with the configuration a bit and removed the rom
file for the HDMI audio controller of the graphics card (previously I
had given it the same rom file as the one I passed to the GPU)

Aha! So that is why I saw two instances of AmdGop.efi.

(I think there may be a bug in AmdGop.efi. A UEFI_DRIVER must cope with
the situation when the PciIo protocol interface on the device handle is
already open by another (or the same) driver.)


and now
the VM actually does not stop at the TianoCore logo anymore. The monitor
now displays the GRUB menu

Great!


and turns black after choosing the first boot
entry. This is the new log now: http://pastebin.com/2dxBqP6h

Yes, this one dispatches AmdGop.efi only once. I can also see
shimx64.efi and grubx64.efi in the log.


I have also configured the VM with emulated graphics using QXL (just the
standard configuration you get with the setup wizard). If I try to boot
with the GPU passed through in addition to having the emulated graphics,
I can see it progressing further on the emulated display. It looks like
it starts booting the OS and then stops at this line:

intel_rapl: no valid rapl domains found in package 0

I don't see any of that on the monitor connected to the graphics card,
though.

The point is, by eliminating the romfile for the HDMI audio controller,
you managed to reach grub, and grub actually started your kernel. (The
above intel_rapl message comes from "drivers/powercap/intel_rapl.c".)

 From this point on, it is a guest kernel / guest GPU driver question.
Since you do have functional grub, I recommend the following:

* Edit your domain configuration so that it has an emulated serial
console. Virtio-serial is *not* good enough. Run "virsh edit", and add
the following:

 
   
 
 
   
 

... Actually, I'm just noticing that your second XML paste, at
, already contains this XML fragment. Good.

* Create the $HOME/.screenrc file, with the following content:

logfile /tmp/screen.log
logfile flush 1

* Start the VM like this, from the shell:

   # launch "screen" first
   screen

   # now press [Ctrl-A Shift-H]
   # this will start logging terminal traffic to the logfile

   # launch the guest
   virsh start --console --paused xubuntu-clone

* Wait until the line "Connected to domain xubuntu-clone" appears.

* Now open the guest in virt-manager as well, and click the Pause button
to unpause the guest.

* As soon as the grub menu appears, choose to edit the kernel command
line, under the first grub menu entry (i.e., the one that you would
normally boot).

If the "quiet" parameter is there, remove it.

Independently, add the following three parameters:

   console=tty console=ttyS0,115200 ignore_loglevel

* Boot the edited command line.

* When the guest becomes stuck again, force it off (using 

Re: [vfio-users] Guest does not boot OS with GPU passed through, stuck at TianoCore boot splash

[Laszlo Ersek]

On 04/12/16 20:42, Stefan Seil wrote:
> As far as I can tell, the other PCI device is the HDMI audio controller
> of the graphics card. It is listed as a seperate PCI device and I
> assigned it to the VM, too.

I think that makes sense. I have the same situation with my GTX750. I
believe those two functions belong to the same IOMMU domain, so they
have to be assigned together anyway.

> I haven't explicitly added any USB devices, although it seems like there
> are some per default. I have linked the respective domain XMLs below.

Hmm no, as far as device assignment goes, you have only two 
elements.

Regarding USB devices, you have a few emulated controllers. As far as
actual devices go, you seem to be using Spice USB redirection (see
). I had to read up on that:
.

I can't tell if you actually redirected some host-side USB devices. I
think it's unlikely; Cole's blog post says "When connected to a VM's
graphical display", and in your first domain XML, your only display is
the assigned GPU (no  element).

> I have played around with the configuration a bit and removed the rom
> file for the HDMI audio controller of the graphics card (previously I
> had given it the same rom file as the one I passed to the GPU)

Aha! So that is why I saw two instances of AmdGop.efi.

(I think there may be a bug in AmdGop.efi. A UEFI_DRIVER must cope with
the situation when the PciIo protocol interface on the device handle is
already open by another (or the same) driver.)

> and now
> the VM actually does not stop at the TianoCore logo anymore. The monitor
> now displays the GRUB menu

Great!

> and turns black after choosing the first boot
> entry. This is the new log now: http://pastebin.com/2dxBqP6h

Yes, this one dispatches AmdGop.efi only once. I can also see
shimx64.efi and grubx64.efi in the log.

> I have also configured the VM with emulated graphics using QXL (just the
> standard configuration you get with the setup wizard). If I try to boot
> with the GPU passed through in addition to having the emulated graphics,
> I can see it progressing further on the emulated display. It looks like
> it starts booting the OS and then stops at this line:
> 
> intel_rapl: no valid rapl domains found in package 0
> 
> I don't see any of that on the monitor connected to the graphics card,
> though.

The point is, by eliminating the romfile for the HDMI audio controller,
you managed to reach grub, and grub actually started your kernel. (The
above intel_rapl message comes from "drivers/powercap/intel_rapl.c".)

>From this point on, it is a guest kernel / guest GPU driver question.
Since you do have functional grub, I recommend the following:

* Edit your domain configuration so that it has an emulated serial
console. Virtio-serial is *not* good enough. Run "virsh edit", and add
the following:


  


  


... Actually, I'm just noticing that your second XML paste, at
, already contains this XML fragment. Good.

* Create the $HOME/.screenrc file, with the following content:

logfile /tmp/screen.log
logfile flush 1

* Start the VM like this, from the shell:

  # launch "screen" first
  screen

  # now press [Ctrl-A Shift-H]
  # this will start logging terminal traffic to the logfile

  # launch the guest
  virsh start --console --paused xubuntu-clone

* Wait until the line "Connected to domain xubuntu-clone" appears.

* Now open the guest in virt-manager as well, and click the Pause button
to unpause the guest.

* As soon as the grub menu appears, choose to edit the kernel command
line, under the first grub menu entry (i.e., the one that you would
normally boot).

If the "quiet" parameter is there, remove it.

Independently, add the following three parameters:

  console=tty console=ttyS0,115200 ignore_loglevel

* Boot the edited command line.

* When the guest becomes stuck again, force it off (using virt-manager),
same as before. This will also make "virsh" return in your "screen"
window, and then you can type "exit" there.

* Email the /tmp/screen.log file, with the guest kernel's debug output,
to whoever provided your guest kernel. :)

Thanks
Laszlo

> These are the two domain XMLs now:
> Without additional emulated graphics: http://pastebin.com/hBTdP6Sq
> With additional emulated graphics: http://pastebin.com/Wm3E23nW


___
vfio-users mailing list
vfio-users@redhat.com
https://www.redhat.com/mailman/listinfo/vfio-users

Re: [vfio-users] Guest does not boot OS with GPU passed through, stuck at TianoCore boot splash

[Stefan Seil]

Oh, interesting. This seems to be what I am experiencing now (see my 
answer to Laszlo). When I am now using additional emulated graphics, I 
can see the boot output (on the emulated display) until it stops at a 
certain point .


On Mon, Apr 11, 2016 at 10:47 PM, Alex Williamson wrote:
On Mon, Apr 11, 2016 at 1:47 PM, Stefan Seil > wrote:


I am already running on 4.5, so that should be fine. The OVMF
version I was using was a couple of months old, so I just got a
newer one from https://www.kraxel.org/repos/jenkins/edk2/ and
pointed the respective XML tag to the newer OVMF_CODE-pure-efi.fd,
but that did not change anything.
I guess your card is from a different manufacturer then and thus
has a different ROM? Because if I try to use the standard ROM of
my card (as in not explicitly passing any rom file to the VM at
all), I don't get any output on the monitor connected to the card
at all.
What exactly do you mean with "AMD cards also have the option of
running as secondary graphics devices in many cases"?


You have two options when you assign a GPU to a VM, you can do it in 
addition to the emulated graphics (cirrus/VGA/QXL) or instead of the 
emulated graphics.  You indicated following my blog, so I assume you 
removed the emulated graphics when you added the assigned GPU.  This 
is necessary for NVIDIA GeForce, but many users have found that AMD 
cards also work when added to the VM configuration with the emulated 
graphics.  No ROM is necessary on the assigned GPU in this case.  The 
emulated graphics will show the BIOS and early boot output, then the 
assigned graphics can be enabled.  This mode doesn't actually require 
OVMF, it works with SeaBIOS too.


Anyway, I'm curious to the answers to Laszlo's questions.  I've also 
seen hangs at the TianoCore splash with certain USB host controllers 
assigned.  Thanks,


Alex


___
vfio-users mailing list
vfio-users@redhat.com
https://www.redhat.com/mailman/listinfo/vfio-users

Re: [vfio-users] Guest does not boot OS with GPU passed through, stuck at TianoCore boot splash

[Stefan Seil]

As far as I can tell, the other PCI device is the HDMI audio controller 
of the graphics card. It is listed as a seperate PCI device and I 
assigned it to the VM, too.
I haven't explicitly added any USB devices, although it seems like there 
are some per default. I have linked the respective domain XMLs below.
I have played around with the configuration a bit and removed the rom 
file for the HDMI audio controller of the graphics card (previously I 
had given it the same rom file as the one I passed to the GPU) and now 
the VM actually does not stop at the TianoCore logo anymore. The monitor 
now displays the GRUB menu and turns black after choosing the first boot 
entry. This is the new log now: http://pastebin.com/2dxBqP6h


I have also configured the VM with emulated graphics using QXL (just the 
standard configuration you get with the setup wizard). If I try to boot 
with the GPU passed through in addition to having the emulated graphics, 
I can see it progressing further on the emulated display. It looks like 
it starts booting the OS and then stops at this line:


intel_rapl: no valid rapl domains found in package 0

I don't see any of that on the monitor connected to the graphics card, 
though.



These are the two domain XMLs now:
Without additional emulated graphics: http://pastebin.com/hBTdP6Sq
With additional emulated graphics: http://pastebin.com/Wm3E23nW

On Mon, Apr 11, 2016 at 10:15 PM, Laszlo Ersek wrote:

On 04/11/16 22:11, Laszlo Ersek wrote:

On 04/11/16 21:48, Stefan Seil wrote:

Thank you for the detailed explanation!
I did this and have uploaded the log here: http://pastebin.com/BXU9UqiG
I don't really see any apparent errors in there...

Right, I expected a failed ASSERT() or something similar, but I'm not
seeing a direct error anywhere.

What I can see from the log is:

- Two instances of "AmdGop.efi" starting. (Note: that doesn't mean they
   are actually binding the card(s) yet.) The number two is strange. Do
   you have two assigned GPUs? Can you share your domain XML?

- I can see three instances of EfiGraphicsOutputProtocol being
   produced. One of them belongs to the QXL driver built into OVMF.
   Again, do you have two assigned GPUs?

Hm, one of those GOPs is probably produced by the console splitter
driver, so please ignore this one paragraph.


- The boot process advances pretty far beyond the above. The last
   message is related to the USB bus driver
   (MdeModulePkg/Bus/Usb/UsbBusDxe). Are you assigning USB devices too?

   In your email, you wrote, "If I remove the PCI devices from the VM,
   the OS boots without problems." -- device*s*, plural. Did you perhaps
   mean "PCI devices" generally as "assigned devices"? If so, what
   happens if you keep the GPU(s), but remove the assigned USB devices
   from the domain XML?

Thanks
Laszlo

___
vfio-users mailing list
vfio-users@redhat.com
https://www.redhat.com/mailman/listinfo/vfio-users



___
vfio-users mailing list
vfio-users@redhat.com
https://www.redhat.com/mailman/listinfo/vfio-users

Re: [vfio-users] Guest does not boot OS with GPU passed through, stuck at TianoCore boot splash

[Jayme Howard]

Possibly more appropriate to the spirit of another thread, but for what
it's worth Alex, I'm using both emulated graphics and an NVIDIA GTX 970
together, successfully.  Not going to try to derail this further, just felt
that it's worth noting for anybody who catches this in the archive.

On Mon, Apr 11, 2016 at 3:36 PM, Alex Williamson <
alex.l.william...@gmail.com> wrote:

> On Mon, Apr 11, 2016 at 1:47 PM, Stefan Seil  wrote:
>
>> I am already running on 4.5, so that should be fine. The OVMF version I
>> was using was a couple of months old, so I just got a newer one from
>> https://www.kraxel.org/repos/jenkins/edk2/ and pointed the respective
>> XML tag to the newer OVMF_CODE-pure-efi.fd, but that did not change
>> anything.
>> I guess your card is from a different manufacturer then and thus has a
>> different ROM? Because if I try to use the standard ROM of my card (as in
>> not explicitly passing any rom file to the VM at all), I don't get any
>> output on the monitor connected to the card at all.
>> What exactly do you mean with "AMD cards also have the option of running
>> as secondary graphics devices in many cases"?
>
>
> You have two options when you assign a GPU to a VM, you can do it in
> addition to the emulated graphics (cirrus/VGA/QXL) or instead of the
> emulated graphics.  You indicated following my blog, so I assume you
> removed the emulated graphics when you added the assigned GPU.  This is
> necessary for NVIDIA GeForce, but many users have found that AMD cards also
> work when added to the VM configuration with the emulated graphics.  No ROM
> is necessary on the assigned GPU in this case.  The emulated graphics will
> show the BIOS and early boot output, then the assigned graphics can be
> enabled.  This mode doesn't actually require OVMF, it works with SeaBIOS
> too.
>
> Anyway, I'm curious to the answers to Laszlo's questions.  I've also seen
> hangs at the TianoCore splash with certain USB host controllers assigned.
> Thanks,
>
> Alex
>
> ___
> vfio-users mailing list
> vfio-users@redhat.com
> https://www.redhat.com/mailman/listinfo/vfio-users
>
>
___
vfio-users mailing list
vfio-users@redhat.com
https://www.redhat.com/mailman/listinfo/vfio-users

Re: [vfio-users] Guest does not boot OS with GPU passed through, stuck at TianoCore boot splash

[Alex Williamson]

On Mon, Apr 11, 2016 at 1:47 PM, Stefan Seil  wrote:

> I am already running on 4.5, so that should be fine. The OVMF version I
> was using was a couple of months old, so I just got a newer one from
> https://www.kraxel.org/repos/jenkins/edk2/ and pointed the respective XML
> tag to the newer OVMF_CODE-pure-efi.fd, but that did not change anything.
> I guess your card is from a different manufacturer then and thus has a
> different ROM? Because if I try to use the standard ROM of my card (as in
> not explicitly passing any rom file to the VM at all), I don't get any
> output on the monitor connected to the card at all.
> What exactly do you mean with "AMD cards also have the option of running
> as secondary graphics devices in many cases"?


You have two options when you assign a GPU to a VM, you can do it in
addition to the emulated graphics (cirrus/VGA/QXL) or instead of the
emulated graphics.  You indicated following my blog, so I assume you
removed the emulated graphics when you added the assigned GPU.  This is
necessary for NVIDIA GeForce, but many users have found that AMD cards also
work when added to the VM configuration with the emulated graphics.  No ROM
is necessary on the assigned GPU in this case.  The emulated graphics will
show the BIOS and early boot output, then the assigned graphics can be
enabled.  This mode doesn't actually require OVMF, it works with SeaBIOS
too.

Anyway, I'm curious to the answers to Laszlo's questions.  I've also seen
hangs at the TianoCore splash with certain USB host controllers assigned.
Thanks,

Alex
___
vfio-users mailing list
vfio-users@redhat.com
https://www.redhat.com/mailman/listinfo/vfio-users

Re: [vfio-users] Guest does not boot OS with GPU passed through, stuck at TianoCore boot splash

[Laszlo Ersek]

On 04/11/16 22:11, Laszlo Ersek wrote:
> On 04/11/16 21:48, Stefan Seil wrote:
>> Thank you for the detailed explanation!
>> I did this and have uploaded the log here: http://pastebin.com/BXU9UqiG
>> I don't really see any apparent errors in there...
> 
> Right, I expected a failed ASSERT() or something similar, but I'm not
> seeing a direct error anywhere.
> 
> What I can see from the log is:
> 
> - Two instances of "AmdGop.efi" starting. (Note: that doesn't mean they
>   are actually binding the card(s) yet.) The number two is strange. Do
>   you have two assigned GPUs? Can you share your domain XML?
> 
> - I can see three instances of EfiGraphicsOutputProtocol being
>   produced. One of them belongs to the QXL driver built into OVMF.
>   Again, do you have two assigned GPUs?

Hm, one of those GOPs is probably produced by the console splitter
driver, so please ignore this one paragraph.

> - The boot process advances pretty far beyond the above. The last
>   message is related to the USB bus driver
>   (MdeModulePkg/Bus/Usb/UsbBusDxe). Are you assigning USB devices too?
> 
>   In your email, you wrote, "If I remove the PCI devices from the VM,
>   the OS boots without problems." -- device*s*, plural. Did you perhaps
>   mean "PCI devices" generally as "assigned devices"? If so, what
>   happens if you keep the GPU(s), but remove the assigned USB devices
>   from the domain XML?
> 
> Thanks
> Laszlo
> 
> ___
> vfio-users mailing list
> vfio-users@redhat.com
> https://www.redhat.com/mailman/listinfo/vfio-users
> 

___
vfio-users mailing list
vfio-users@redhat.com
https://www.redhat.com/mailman/listinfo/vfio-users

Re: [vfio-users] Guest does not boot OS with GPU passed through, stuck at TianoCore boot splash

[Stefan Seil]

Thank you for the detailed explanation!
I did this and have uploaded the log here: http://pastebin.com/BXU9UqiG
I don't really see any apparent errors in there...

I wasn't necessarily expecting this to work with my card, but I thought 
I'd try it anyways. Of course I won't really have another option than 
getting a compatible card if it won't work out (other than not using 
UEFI, but as far as I understand it I would have to patch the kernel 
because I have an Intel CPU and would lose some capabilities of the 
integrated GPU as a result of that, which probably would be quite a 
hassle in the long run).


On Mon, Apr 11, 2016 at 5:48 PM, Laszlo Ersek wrote:

On 04/10/16 16:46, Blank Field wrote:


What you can do - convert libvirt xml into a usual qemu command line,
and add an ISA debug console and check OVMF debug output in it.

* Yes, capturing the OVMF debug log should be the first step when
reporting apparent OVMF issues.

You can get that log without resorting to the raw QEMU command line.
Modify your domain XML like this:

- add the following attribute to the root element of the domain XML:

   

- add the following XML fragment near the end of the domain XML, as a
direct child of the  element:

   
 
 
 
 
   

This will send the OVMF log to "/tmp/guest_name.log".

* Regarding the specific case: a UEFI oprom for a GPU may easily be
buggy even if it matches the card. If you are experimenting with
"similar" drivers, all bets are off. Please buy a decent card (Alex has
recommended the GTX750 several times; I myself bought one on his
recommendation, and have had zero problems with it). Or else, at least
get an officially supported driver binary from the manufacturer, please.

Thanks
Laszlo


___
vfio-users mailing list
vfio-users@redhat.com
https://www.redhat.com/mailman/listinfo/vfio-users

Re: [vfio-users] Guest does not boot OS with GPU passed through, stuck at TianoCore boot splash

[Stefan Seil]

I am already running on 4.5, so that should be fine. The OVMF version I 
was using was a couple of months old, so I just got a newer one from 
https://www.kraxel.org/repos/jenkins/edk2/ and pointed the respective 
XML tag to the newer OVMF_CODE-pure-efi.fd, but that did not change 
anything.
I guess your card is from a different manufacturer then and thus has a 
different ROM? Because if I try to use the standard ROM of my card (as 
in not explicitly passing any rom file to the VM at all), I don't get 
any output on the monitor connected to the card at all.
What exactly do you mean with "AMD cards also have the option of running 
as secondary graphics devices in many cases"?


On Mon, Apr 11, 2016 at 5:55 PM, Alex Williamson wrote:
My guess is the old kernel/OVMF MTRR issues, what kernel version are 
you running?  What OVMF version are you running.  Upgrade both ;)  But 
of course a "close enough" ROM may not really be close enough.  I find 
that my HD7850 works pretty well w/  OVMF even w/o a UEFI compatible 
ROM.  AMD cards also have the option of running as secondary graphics 
devices in many cases, at least old ones like this where reset works.  
Thanks,


Alex


___
vfio-users mailing list
vfio-users@redhat.com
https://www.redhat.com/mailman/listinfo/vfio-users

Re: [vfio-users] Guest does not boot OS with GPU passed through, stuck at TianoCore boot splash

[Blank Field]

He's also should be able to attach just the .efi file or install graphics
drivers via qxl.
On Apr 11, 2016 6:48 PM, "Laszlo Ersek"  wrote:

> On 04/10/16 16:46, Blank Field wrote:
>
> > What you can do - convert libvirt xml into a usual qemu command line,
> > and add an ISA debug console and check OVMF debug output in it.
>
> * Yes, capturing the OVMF debug log should be the first step when
> reporting apparent OVMF issues.
>
> You can get that log without resorting to the raw QEMU command line.
> Modify your domain XML like this:
>
> - add the following attribute to the root element of the domain XML:
>
>   type='kvm'
>xmlns:qemu='http://libvirt.org/schemas/domain/qemu/1.0'  <-- this one
>   >
>
> - add the following XML fragment near the end of the domain XML, as a
> direct child of the  element:
>
>   
> 
> 
> 
> 
>   
>
> This will send the OVMF log to "/tmp/guest_name.log".
>
> * Regarding the specific case: a UEFI oprom for a GPU may easily be
> buggy even if it matches the card. If you are experimenting with
> "similar" drivers, all bets are off. Please buy a decent card (Alex has
> recommended the GTX750 several times; I myself bought one on his
> recommendation, and have had zero problems with it). Or else, at least
> get an officially supported driver binary from the manufacturer, please.
>
> Thanks
> Laszlo
>
___
vfio-users mailing list
vfio-users@redhat.com
https://www.redhat.com/mailman/listinfo/vfio-users

Re: [vfio-users] Guest does not boot OS with GPU passed through, stuck at TianoCore boot splash

[Alex Williamson]

On Sun, Apr 10, 2016 at 8:04 AM, Stefan Seil  wrote:

> Hello there!
>
> I am trying to set up a PCI passthrough for my graphics card. I followed
> the instructions on vfio.blogspot.com and used vfio-pci and OVMF, and so
> far everything seemed to work fine. When I now try to start the VM with my
> graphics card passed through, though, it shows the TianoCore boot splash
> (with the successfull memory test notice on the bottom) on the monitor
> connected to the card and stops there. I tried waiting for quite a while,
> but it is stuck at this boot splash. If I remove the PCI devices from the
> VM, the OS boots without problems.
>
> I don't see any error messages in any of the relevant logs I can think of
> (libvirt log, dmesg, virt-manager itself). The only line showing up in the
> libvirt log (/var/log/libvirt/qemu/...) is the following:
>
> 2016-04-10 10:45:08.461+: starting up libvirt version: 1.3.2, package:
> 0~15.10~ppa0 (Jacob Zimmermann  Fri, 04 Mar LC_ALL=C
> PATH=/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin 
> QEMU_AUDIO_DRV=spice
> /usr/bin/kvm-spice -name xubuntu2 -S -char device redirected to /dev/pts/9
> (label charserial0)
>
> I don't really have an idea what the problem could be, especially since
> there are no errors showing up. I am using an Intel Core i5-3470, a
> Sapphire HD 7850 (AMD) and Xubuntu 15.10 on the host. I installed all the
> relevant tools from a PPA by Jacob Zimmermann in order to get more recent
> versions of qemu, virt-manager and the likes.
>
> One thing I should note is that the ROM of my graphics card does not
> support UEFI out of the box. As it is, the passthrough did not work at all;
> there was nothing showing up on the monitor connected to the card. I found
> a ROM from a similar model by Gigabyte that supports UEFI, though, and
> passed it to the VM. Now the monitor shows the TianoCore boot splash, but
> stops there. Could this be the problem? I would think that it should work
> this way, though, as the boot splash is already passed to the card;
> otherwise there would be nothing showing up on the monitor, right?
>

My guess is the old kernel/OVMF MTRR issues, what kernel version are you
running?  What OVMF version are you running.  Upgrade both ;)  But of
course a "close enough" ROM may not really be close enough.  I find that my
HD7850 works pretty well w/  OVMF even w/o a UEFI compatible ROM.  AMD
cards also have the option of running as secondary graphics devices in many
cases, at least old ones like this where reset works.  Thanks,

Alex
___
vfio-users mailing list
vfio-users@redhat.com
https://www.redhat.com/mailman/listinfo/vfio-users

Re: [vfio-users] Guest does not boot OS with GPU passed through, stuck at TianoCore boot splash

[Laszlo Ersek]

On 04/10/16 16:46, Blank Field wrote:

> What you can do - convert libvirt xml into a usual qemu command line,
> and add an ISA debug console and check OVMF debug output in it.

* Yes, capturing the OVMF debug log should be the first step when
reporting apparent OVMF issues.

You can get that log without resorting to the raw QEMU command line.
Modify your domain XML like this:

- add the following attribute to the root element of the domain XML:

  

- add the following XML fragment near the end of the domain XML, as a
direct child of the  element:

  




  

This will send the OVMF log to "/tmp/guest_name.log".

* Regarding the specific case: a UEFI oprom for a GPU may easily be
buggy even if it matches the card. If you are experimenting with
"similar" drivers, all bets are off. Please buy a decent card (Alex has
recommended the GTX750 several times; I myself bought one on his
recommendation, and have had zero problems with it). Or else, at least
get an officially supported driver binary from the manufacturer, please.

Thanks
Laszlo

___
vfio-users mailing list
vfio-users@redhat.com
https://www.redhat.com/mailman/listinfo/vfio-users

Re: [vfio-users] Guest does not boot OS with GPU passed through, stuck at TianoCore boot splash

[Blank Field]

100% one core CPU?
What kind of disk device you have?
Are you passing through a physical disk?

What you can do - convert libvirt xml into a usual qemu command line, and
add an ISA debug console and check OVMF debug output in it.
On Apr 10, 2016 5:04 PM, "Stefan Seil"  wrote:

> Hello there!
>
> I am trying to set up a PCI passthrough for my graphics card. I followed
> the instructions on vfio.blogspot.com and used vfio-pci and OVMF, and so
> far everything seemed to work fine. When I now try to start the VM with my
> graphics card passed through, though, it shows the TianoCore boot splash
> (with the successfull memory test notice on the bottom) on the monitor
> connected to the card and stops there. I tried waiting for quite a while,
> but it is stuck at this boot splash. If I remove the PCI devices from the
> VM, the OS boots without problems.
>
> I don't see any error messages in any of the relevant logs I can think of
> (libvirt log, dmesg, virt-manager itself). The only line showing up in the
> libvirt log (/var/log/libvirt/qemu/...) is the following:
>
> 2016-04-10 10:45:08.461+: starting up libvirt version: 1.3.2, package:
> 0~15.10~ppa0 (Jacob Zimmermann  Fri, 04 Mar LC_ALL=C
> PATH=/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin 
> QEMU_AUDIO_DRV=spice
> /usr/bin/kvm-spice -name xubuntu2 -S -char device redirected to /dev/pts/9
> (label charserial0)
>
> I don't really have an idea what the problem could be, especially since
> there are no errors showing up. I am using an Intel Core i5-3470, a
> Sapphire HD 7850 (AMD) and Xubuntu 15.10 on the host. I installed all the
> relevant tools from a PPA by Jacob Zimmermann in order to get more recent
> versions of qemu, virt-manager and the likes.
>
> One thing I should note is that the ROM of my graphics card does not
> support UEFI out of the box. As it is, the passthrough did not work at all;
> there was nothing showing up on the monitor connected to the card. I found
> a ROM from a similar model by Gigabyte that supports UEFI, though, and
> passed it to the VM. Now the monitor shows the TianoCore boot splash, but
> stops there. Could this be the problem? I would think that it should work
> this way, though, as the boot splash is already passed to the card;
> otherwise there would be nothing showing up on the monitor, right?
>
> ___
> vfio-users mailing list
> vfio-users@redhat.com
> https://www.redhat.com/mailman/listinfo/vfio-users
>
>
___
vfio-users mailing list
vfio-users@redhat.com
https://www.redhat.com/mailman/listinfo/vfio-users

[vfio-users] Guest does not boot OS with GPU passed through, stuck at TianoCore boot splash

[Stefan Seil]

Hello there!

 

I am trying to set up a PCI passthrough for my graphics card. I followed the instructions on vfio.blogspot.com and used vfio-pci and OVMF, and so far everything seemed to work fine. When I now try to start the VM with my graphics card passed through, though, it shows the TianoCore boot splash (with the successfull memory test notice on the bottom) on the monitor connected to the card and stops there. I tried waiting for quite a while, but it is stuck at this boot splash. If I remove the PCI devices from the VM, the OS boots without problems.

 

I don't see any error messages in any of the relevant logs I can think of (libvirt log, dmesg, virt-manager itself). The only line showing up in the libvirt log (/var/log/libvirt/qemu/...) is the following:

 

2016-04-10 10:45:08.461+: starting up libvirt version: 1.3.2, package: 0~15.10~ppa0 (Jacob Zimmermann  Fri, 04 Mar LC_ALL=C PATH=/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin QEMU_AUDIO_DRV=spice /usr/bin/kvm-spice -name xubuntu2 -S -char device redirected to /dev/pts/9 (label charserial0)

 

I don't really have an idea what the problem could be, especially since there are no errors showing up. I am using an Intel Core i5-3470, a Sapphire HD 7850 (AMD) and Xubuntu 15.10 on the host. I installed all the relevant tools from a PPA by Jacob Zimmermann in order to get more recent versions of qemu, virt-manager and the likes.

 

One thing I should note is that the ROM of my graphics card does not support UEFI out of the box. As it is, the passthrough did not work at all; there was nothing showing up on the monitor connected to the card. I found a ROM from a similar model by Gigabyte that supports UEFI, though, and passed it to the VM. Now the monitor shows the TianoCore boot splash, but stops there. Could this be the problem? I would think that it should work this way, though, as the boot splash is already passed to the card; otherwise there would be nothing showing up on the monitor, right?

___
vfio-users mailing list
vfio-users@redhat.com
https://www.redhat.com/mailman/listinfo/vfio-users