On Sunday, February 25, 2018 at 12:08:37 AM UTC+1, Daniil .Travnikov wrote:
> суббота, 24 февраля 2018 г., 3:52:11 UTC+3 пользователь Yuraeitha написал:
> > On Friday, February 23, 2018 at 8:35:00 PM UTC+1, Daniil .Travnikov wrote:
> > > пятница, 23 февраля 2018 г., 14:07:38 UTC+3 пользователь awokd написал:
> > > > On Fri, February 23, 2018 10:46 am, Daniil .Travnikov wrote:
> 
> Thank you for sharing your experience!
> 
> > Since you have trouble getting the installer to work and install Qubes, you 
> > may not be able to do this fix on your local hardware. You may have to pull 
> > out the drive, put it in another computer and install Qubes there. Update 
> > everything, and then make sure you put sys-net and sys-usb in PV mode with 
> > the "qvm-prefs virt_mode" command in dom0.
> > 
> > Now because sys-net and sys-usb are in PV mode, it may be able to bypass 
> > the missing I/O MMU issue, which as far as I understand it is related to 
> > PCI pass-through. That's why you want sys-net, sys-usb and any other 
> > hardware that is pass-through to be using PV mode.
> > 
> > This isn't a beautiful fix, but it may just work. It's not the first time 
> > I've fixed a Qubes 4 install with this approach, however, I have not yet 
> > tried it on this server hardware, which like yours is missing I/O MMU. I 
> > believe it might work, but it might also not work. While installing on 
> > another machine has in the past worked, I never tried to use it to change 
> > sys-net and sys-usb to PV mode before putting it back.
> 
> I installed one of my drives on a friend's computer today and after 
> installation process I got this on his computer:
> https://prnt.sc/ijbfjq
> 
> You can see, that I have not any NetVM, actually at the and of installation 
> process was been some error, I can repeat tomorrow all installation process 
> for screenshots if it helps to understand.
> 
> 
> When I put "qvm-prefs virt_mode" command in dom0 I got this:
> https://prnt.sc/ijbgmd
> 
> I think that I am missing something, could you please clarify what I must 
> exactly put here?
> 
> 
> One more think, even I have not NetVM, I tried to boot this disk on My 
> computer, but all what I got was been Grub menu:
> http://prntscr.com/ijbhrf
> 
> First and third gave reboot, second just show some error about "first boot 
> must kernel" (or something like that) without reboot.

ah, yes you need to change how you write that command, it's not qvm-block, but 
qvm-prefs. This may be important in order to fix your missing networking, as 
it's evident on your screenshot. So you can write "qvm-prefs sys-net virt_mode" 
in dom0 terminal, and it will print out which virt mode your network VM is 
using. It's currently either PV, HVM or PVH (security in that order, from worst 
to best).

But hardware compatibility suffers as you increase security, that's why you 
need to lower it. PV mode is the lowest, however, it is not lower in security 
in comparison to Qubes 3.2. which entirely uses PV mode. Therefore it's not 
such a bad thing to start on, to get things working, and then later consider 
the HVM/PVH implications once you get it all running.

I can see you have no sys-usb, so (for now) just ignore this isolation aspect 
and re-visit it later when you got other issues sorted out.

For verification, do in dom0:
qvm-prefs sys-net virt_mode

To change virt mode, do in dom0:
qvm-prefs sys-net virt_mode pv

- Now try manually start up sys-net, is it starting?
- If it's not starting, try start or create and start a VM which has no 
networking, does this start in PVH/HVM/PV?

-- 
You received this message because you are subscribed to the Google Groups 
"qubes-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to qubes-users+unsubscr...@googlegroups.com.
To post to this group, send email to qubes-users@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/qubes-users/c2189c72-2646-4336-87f3-86a06f42a8f5%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Reply via email to