On Dec 3, 10:22am, Konrad Rzeszutek Wilk wrote:
} Subject: Re: [Xen-devel] Updated ATI passthrough patches.
Good morning, hope the day is going well for everyone.
> On Sat, Dec 01, 2012 at 06:24:59PM -0600, Dr. Greg Wettstein wrote:
> > Hi, hope the weekend is going well for everyone.
> >
> > I just put a set of updated patches to support ATI passthrough on a
> > primary video adapter on the FTP site. The URL''s are as
follows:
> >
> > ftp://ftp.enjellic.com/pub/xen/xen-4.1.3.ati-passthrough.patch
> >
> > ftp://ftp.enjellic.com/pub/xen/xen-4.2.0.ati-passthrough.patch
> >
> > These patches have been validated to work up through kernel 3.4.18
> > with the xm control plane. We are currently working on validating
> > whether or not there are passthrough issues with xl.
> >
> > The original ATI pass-through patches posted to xen-devel fail with a
> > qemu-dm segmentation fault on recent kernels. This is caused by
> > changes which have been made with respect to proper enforcement of the
> > permitted port ranges on the ioperm() system call.
> >
> > Since these patchs allow a primary graphics adapter to be used for
> > passthrough I wanted to remind everyone of the availability of the
> > following utility script:
> >
> > ftp://ftp.enjellic.com/pub/xen/run-passthrough
> >
> > Which automates the process of unplugging and re-plugging a video card
> > and optionally a USB controller. A script like this, or a network
> > login, is needed in order to use passthrough on a primary graphics
> > adapter.
> Great! What type of config options do you end up using for your guests?
> gfx_pass.., msix_translate=.. ? Or just normal
>
> pci=[BDF] ?
The configuration is pretty straight forward. The active parameters
are as follows:
----------------------------------------------------------------------------
builder=''hvm''
memory = 3072
name = "Windows"
vif = [ ''type=ioemu, bridge=bridge0, model=e1000'' ]
acpi = 1
apic = 1
disk = [ ''phy:/dev/localvg1/winsnap,hda,w'' ]
boot="c"
sdl=0
vnc=0
soundhw=''ac97''
stdvga=1
gfx_passthru=1
pci=[ ''01:00.0'', ''00:1a.0'' ]
----------------------------------------------------------------------------
The 01:00.0 device is the ATI video card:
01:00.0 VGA compatible controller: ATI Technologies Inc Device 6898
(prog-if 00
[VGA controller])
Subsystem: ASUSTeK Computer Inc. Device 0346
Flags: bus master, fast devsel, latency 0, IRQ 11
Memory at b0000000 (64-bit, prefetchable) [size=256M]
Memory at c1a00000 (64-bit, non-prefetchable) [size=128K]
I/O ports at 3000 [size=256]
Expansion ROM at c1a40000 [disabled] [size=128K]
Capabilities: <access denied>
The 00:1a.0 devie is the USB controller which has the keyboard and
mouse on it:
00:1a.0 USB Controller: Intel Corporation Ibex Peak USB2 Enhanced Host
Controller (rev 05) (prog-if 20 [EHCI])
Subsystem: Intel Corporation Device 34ec
Flags: bus master, medium devsel, latency 0, IRQ 21
Memory at c1b22000 (32-bit, non-prefetchable) [size=1K]
Capabilities: <access denied>
Kernel driver in use: ehci_hcd
This is under a 3.4.18 dom0 kernel which is statically compiled for
the hardware platform being used.
Xen is stock 4.2.0 with the ATI patches noted above applied.
The xm control plane is used since xl appears to be a non-starter with
all this. A bit more on that later.
The above is rock solid under Windows 7. I would judge about 30
extended Windows sessions were run over the last week.
Greg
}-- End of excerpt from Konrad Rzeszutek Wilk
As always,
Dr. G.W. Wettstein, Ph.D. Enjellic Systems Development, LLC.
4206 N. 19th Ave. Specializing in information infra-structure
Fargo, ND 58102 development.
PH: 701-281-1686
FAX: 701-281-3949 EMAIL: greg@enjellic.com
------------------------------------------------------------------------------
"If we thought this was a trap, we wouldn''t be doing it, and as
you know,
we have a lot of lawyers."
-- Irving Wladawsky-Berger