|
|
|
|
|
|
|
|
|
|
xen-users
[Xen-users] VT-d on Gigabyte H55M-UD2H
Hello,
I have been experimenting with VT-d on a Gigabyte H55M board. I realize that
Gigabyte's support of VT-d is "flaky at best", but I though I would give it a
try.
After contacting gigabyte tech support, they sent me a custom BIOS with VT-d
support enabled (which I found quite nice, considering they don't officially
offer VT-d).
xen now tries to activate I/O virtualisation but fails (note: this is with Xen
4.1.0):
xm dmesg:
(XEN) [VT-D]iommu.c:1078: drhd->address = fed90000 iommu->reg = ffff82c3fff57000
(XEN) [VT-D]iommu.c:1080: cap = ffffffffffffffff ecap = ffffffffffffffff
(XEN) [VT-D]iommu.c:1086: IOMMU: unsupported
(XEN) ---- print_iommu_regs ----
(XEN) drhd->address = fed90000
(XEN) VER = ffffffff
(XEN) CAP = ffffffffffffffff
(XEN) n_fault_reg = 100
(XEN) fault_recording_offset = 3ff0
(XEN) ECAP = ffffffffffffffff
(XEN) GCMD = ffffffff
(XEN) GSTS = ffffffff
(XEN) RTADDR = ffffffffffffffff
(XEN) CCMD = ffffffffffffffff
(XEN) FSTS = ffffffff
(XEN) FECTL = ffffffff
(XEN) FEDATA = ffffffff
(XEN) FEADDR = ffffffff
(XEN) FEUADDR = ffffffff
(XEN) I/O virtualisation disabled
Now, I don't know very much about VT-d and what the BIOS's part in this is, but
all these ffff's don't look good. The question is, is the BIOS faulty or is xen
looking in the wrong place?
I send the above report to the gigabyte techs, and they said they don't support
linux. They did however, make an identical setup with Win7 64 and VMWare and
confirm that VT-d is working for them.
Does this mean that VMWare is doing something that xen isn't?
Cheers,
Martin
_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-users
|
<Prev in Thread] |
Current Thread |
[Next in Thread> |
- [Xen-users] VT-d on Gigabyte H55M-UD2H,
Martin Skinner <=
|
|
|
|
|