bhyve passthrough fails

2017-09-07 Thread Farhan Khan
Hi,

I am trying to pass an mPCIe device to a guest bhyve VM. I have the device
set to use the ppt driver:

# devctl set driver pci0:2:0:0 ppt

pciconf -lv gives me the following output:

ppt0@pci0:2:0:0: class=0x028000 card=0x197d103c chip=0x817910ec rev=0x01
hdr=0x00
vendor = 'Realtek Semiconductor Co., Ltd.'
device = 'RTL8188EE Wireless Network Adapter'
class  = networ

But when I start a FreeBSD guest with the vmrun script (sudo sh vmrun.sh -p
2/0/0 -c 2 -m 512M -t tap0 -d freebsd12.img freebsd1), I immediately get
this error:

bhyve: PCI device at 2/0/0 is not using the ppt(4) driver.

I have virtualization set in the BIOS and am not certain what may be
causing this issue. Any ideas?

Thank you,

--
Farhan Khan
PGP Fingerprint: 782F 342B 5B08 0D2F F4E8 82C3 FFA1 CAE1 6536 51CA
___
freebsd-virtualization@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-virtualization
To unsubscribe, send any mail to 
"freebsd-virtualization-unsubscr...@freebsd.org"


Re: openbsd 6.0 or 6.1 guest on freebsd-12 (current) host

2017-09-07 Thread Alan Somers
On Wed, Sep 6, 2017 at 3:53 PM, Jason Tubnor  wrote:
> On 6 September 2017 at 22:15, tech-lists  wrote:
>
>>
>>
>> Have you encountered anything on openbsd in a bhyve that you've found
>> not to work?
>>
>>
> As Thomas mentioned, there is/was a bug with certain CPUs, but this was due
> to the strict checking of CPU features that OpenBSD introduced (
> https://marc.info/?l=openbsd-misc&m=149136173520510&w=2).  Peter explained
> the switch to get around that issue (-w).
>
> The only other thing is Jumbo frame support, but that is really down to
> limits in VirtIO and not really a show stopper and doesn't affect anything
> in production for us.
>
> Enjoy bhyve, it has been a drop in replacement for our individual ESXi
> instances and has been bulletproof, which is good as we have hypervisor
> hosts over 6 hour drives away! (Oh the uplift from 11.0 to 11.1 was done
> remotely without nuking the host).
>
> Cheers,
>
> Jason.

Do you have a semiautomated way to move bhyve instances between
different hosts?  Without such a feature, we can't replace all of our
ESXi machines.

-Alan
___
freebsd-virtualization@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-virtualization
To unsubscribe, send any mail to 
"freebsd-virtualization-unsubscr...@freebsd.org"


Re: bhyve passthrough fails

2017-09-07 Thread Peter Grehan

Hi Farhan,

bhyve: PCI device at 2/0/0 is not using the ppt(4) driver.


 Just noticed this myself: looks like a regression :( I'll see if I can 
work it out and get a fix in.


later,

Peter.
___
freebsd-virtualization@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-virtualization
To unsubscribe, send any mail to 
"freebsd-virtualization-unsubscr...@freebsd.org"


Re: bhyve passthrough fails

2017-09-07 Thread Rodney W. Grimes
> Hi Farhan,
> > bhyve: PCI device at 2/0/0 is not using the ppt(4) driver.
> 
>   Just noticed this myself: looks like a regression :( I'll see if I can 
> work it out and get a fix in.

What is the state of bhyve in 10.4?  Do we need to fire off some
quick tests

-- 
Rod Grimes rgri...@freebsd.org
___
freebsd-virtualization@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-virtualization
To unsubscribe, send any mail to 
"freebsd-virtualization-unsubscr...@freebsd.org"


Re: bhyve passthrough fails

2017-09-07 Thread Farhan Khan
I would not mind sending you on-going feedback or debugging information if
it helps.

Could it be a chipset issue? This is a lower-end laptop. It can do
virtualization, but the passthrough issue seemed to fail. I was on FreeBSD
11.1.

--
Farhan Khan
PGP Fingerprint: 782F 342B 5B08 0D2F F4E8 82C3 FFA1 CAE1 6536 51CA

On Thu, Sep 7, 2017 at 10:52 AM, Peter Grehan  wrote:

> Hi Farhan,
>
>> bhyve: PCI device at 2/0/0 is not using the ppt(4) driver.
>>
>
>  Just noticed this myself: looks like a regression :( I'll see if I can
> work it out and get a fix in.
>
> later,
>
> Peter.
>
___
freebsd-virtualization@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-virtualization
To unsubscribe, send any mail to 
"freebsd-virtualization-unsubscr...@freebsd.org"


Re: bhyve passthrough fails

2017-09-07 Thread Peter Grehan
I would not mind sending you on-going feedback or debugging information  > if it helps.> > Could it be a chipset issue? This is a lower-end 
laptop. It can do > virtualization, but the passthrough issue seemed to 
fail. I was on > FreeBSD 11.1.
 It may be due to this (and also my issue which was on AMD, may be due 
to a reboot where I didn't enable the iommu). If that is the case, the 
error handling could certainly be improved :(


 You can check to see if there is an IOMMU available by looking for the 
APCI DMAR table:


  acpidump -t | grep DMAR

later,

Peter.
___
freebsd-virtualization@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-virtualization
To unsubscribe, send any mail to 
"freebsd-virtualization-unsubscr...@freebsd.org"


Re: bhyve passthrough fails

2017-09-07 Thread Peter Grehan

Hi Rod,


Hi Farhan,

bhyve: PCI device at 2/0/0 is not using the ppt(4) driver.


   Just noticed this myself: looks like a regression :( I'll see if I can
work it out and get a fix in.


What is the state of bhyve in 10.4?  Do we need to fire off some
quick tests


 Always worth some quick tests before a release.

later,

Peter.

___
freebsd-virtualization@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-virtualization
To unsubscribe, send any mail to 
"freebsd-virtualization-unsubscr...@freebsd.org"


Re: openbsd 6.0 or 6.1 guest on freebsd-12 (current) host

2017-09-07 Thread Jason Tubnor
On 8 September 2017 at 00:28, Alan Somers  wrote:

>
>
> Do you have a semiautomated way to move bhyve instances between
> different hosts?  Without such a feature, we can't replace all of our
> ESXi machines.
>

We are not using bhyve for those workloads (yet).  Our vsphere environment
still has life left in it but we never want to bet on one horse (I've been
in too many positions where vendor lock has occurred and its been painful
and expensive).  However, where there are single hosts instances out at
remote branches, it is bhyve all the way where we use to use individual
instances of ESXi.

So for your use case, it isn't there yet, but, if you don't need live
migration, using zfs and/or iSCSI you could tool something together for
offline migration of guests between hosts.

Allowing you to get closer to the operating system (unlike the abstraction
that is vsphere/ESXi), gives you the ability to take the excellent tools
that FreeBSD has to offer and make them fit your application.
___
freebsd-virtualization@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-virtualization
To unsubscribe, send any mail to 
"freebsd-virtualization-unsubscr...@freebsd.org"