From: Heiko Harders <heiko.harders@gmail.com>
To: kvm@vger.kernel.org
Subject: pcidevice: failed to assign irq / hang on Intel nic boot message
Date: Thu, 11 Jun 2009 12:10:37 +0200 [thread overview]
Message-ID: <4A30D81D.80100@gmail.com> (raw)
Hello,
I'm trying to setup a virtual machine with my onboard nic passed
through. Unfortunately I get the message:
Failed to assign irq for "2:00.0": Input/output error
Perhaps you are assigning a device that shares an IRQ with another device?
I'm using a clean Ubuntu 9.04 installation (64 bits version) which comes
with kernel `2.6.28-11-generic'. The kvm version is 84 and I'm using an
AMD cpu.
The relevant section from `dmesg' shows that IRQ 18 is used, but that by
MSI/MSI-X also IRQ 2300 is assigned. IRQ 18 is also used for onboard usb
devices and the graphics card. I need these, so I can't shut them down
unfortunately.
# relevant snipped from dmesg after booting the machine:
[ 4.123477] r8169 Gigabit Ethernet driver 2.3LK-NAPI loaded
[ 4.123498] r8169 0000:02:00.0: PCI INT A -> GSI 18 (level, low) ->
IRQ 18
[ 4.123514] r8169 0000:02:00.0: setting latency timer to 64
[ 4.123593] r8169 0000:02:00.0: irq 2300 for MSI/MSI-X
The procedure I followed was:
# unbind the onboard nic
cd /sys/bus/pci/devices/0000:02:00.0/driver
echo -n "0000:02:00.0" > unbind
# try to run kvm
kvm -m 512 -hda /dev/vg/vm1 -cdrom ubuntu-9.04-server-amd64.iso
-pcidevice host=2:00.0 -boot d
Failed to assign irq for "2:00.0": Input/output error
Perhaps you are assigning a device that shares an IRQ with another device?
# The following appeared in dmesg:
[ 1220.744178] r8169 0000:02:00.0: PCI INT A disabled
[ 1242.066374] pci 0000:02:00.0: PCI INT A -> GSI 18 (level, low) -> IRQ 18
[ 1242.066662] pci 0000:02:00.0: Invalid ROM contents
[ 1242.761820] kvm: 5930: cpu0 unhandled wrmsr: 0xc0010117 data 0
[ 1243.389097] pci 0000:02:00.0: PCI INT A disabled
Does this mean that for my setup it isn't possible to use the onboard
nic as a pcidevice for kvm? Anything I missed, or suggestions to try?
Another thing I tried was using another nic (an intel pro/100) as
pcidevice. This seemed to work, I didn't get any complaints (that nic
didn't share any irq's with other devices). The QEMU window appears and
the virtual machine boots. But the Intel nic shows a message during booting:
Initializing Intel PRO/100 Boot Agent Version 2.0
Press Ctrl+S to enter the Setup Program..
While the progress dots (behind the word `Program' in the last sentence)
appeared, QEMU didn't advance beyond this point. Pressing Ctrl+S brought
me into the set program, but disabling the boot message didn't make any
difference (and while the message was disabled it was still possible to
enter the setup program using Ctrl+S). Is this a known problem? Any work
arround for this?
Thanks for any help/insights,
Heiko
reply other threads:[~2009-06-11 10:10 UTC|newest]
Thread overview: [no followups] expand[flat|nested] mbox.gz Atom feed
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=4A30D81D.80100@gmail.com \
--to=heiko.harders@gmail.com \
--cc=kvm@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.