Hi,
I have a debian (testing/Etch) system with Xen (3.0.3 as in Debian) on
it using the debian 2.6.18 kernel from debian on the host and using
almost the same on the guest (adm8211 driver added).
One of the user domains is in use as the firewall and has two PCI
devices exported into it. One is a adm NC100 regular ethernet NIC (which
works without any problems). The other is an adm8211 based wireless card
that I''m trying to get to work.
Pciback is warning that the driver is attempting to write to the
configuration space of this device and that I should report it (the
device is not working because, I guess, the driver has some issues).
I think it is in fact the guest linux kernel that is attempting to write
to the configuration space of this device, because pciback also reported
this back when I didn''t yet have a proper driver for that device.
I''ll
double check on that and also set the loglevel high to get any messages
from the kernel concerning this.
Something else that''s weird is that halting the guest doesn''t
work
properly... please take a look at the last few lines of the guest
shutting down (although, since the driver doesn''t work properly, this
might be a result of that). Note that, in the guest, pci 0:0 is the
ethernet nic and 0:1 is the wireless nic. I didn''t notice this when the
wireless nic was not being used (I mean I don''t remember seeing them;
they could''ve been there).
Unmounting local filesystems...done.
Will now halt.
pcifront pci-0: pciback not responding!!!
pcifront pci-0: pciback not responding!!!
pcifront pci-0: pciback not responding!!!
xenbus_dev_shutdown: device/vbd/2050 timeout closing device
System halted.
Also the host is an amd64 machine (UP) with a 64 bit userspace. The
guest is also of the amd64 architecture.
This is the lspci as seen on the host (after some testing, so it''s been
used):
00:08.0 Network controller: ADMtek ADM8211 802.11b Wireless Interface
(rev 11)
Subsystem: Standard Microsystems Corp [SMC] Unknown device 2602
Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop-
ParErr- Stepping- SERR- FastB2B-
Status: Cap+ 66MHz- UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort-
<TAbort- <MAbort- >SERR- <PERR-
Latency: 96 (16000ns min, 32000ns max), Cache Line Size: 32 bytes
Interrupt: pin A routed to IRQ 16
Region 0: I/O ports at dc00 [size=256]
Region 1: Memory at cfffe800 (32-bit, non-prefetchable) [size=1K]
Expansion ROM at cff80000 [disabled] [size=128K]
Capabilities: [c0] Power Management version 2
Flags: PMEClk- DSI- D1+ D2+ AuxCurrent=100mA
PME(D0+,D1+,D2+,D3hot+,D3cold+)
Status: D0 PME-Enable- DSel=0 DScale=0 PME-
This is the quirk that prevents the warning (and no others appear):
(adm8211
(pci_ids
# Entries are formated as follows:
#
<vendor>:<device>[:<subvendor>:<subdevice>]
(''1317:8201''
)
)
(pci_config_space_fields
# Entries are formated as follows:
# <register>:<size>:<mask>
# size is measured in bytes (1,2,4 are valid sizes)
# mask is currently unused; use all zero''s
(''000000c4:2:00000000''
)
)
)
I don''t know what is at 0xc4 in the configuration space of the adm8211
(and couldn''t easily find it from the sources).
Anything else you need? Please ask.
Please CC me as I''m not subscribed.
Many thanks,
Jan Evert
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel