public inbox for kvm@vger.kernel.org
 help / color / mirror / Atom feed
From: "Ryan C. Underwood" <nemesis@icequake.net>
To: kvm@vger.kernel.org
Subject: PCI passthrough resource remapping
Date: Fri, 8 Jan 2010 20:45:01 -0600	[thread overview]
Message-ID: <20100109024500.GB4727@localhost.localdomain> (raw)


I have a multifunction PCI device that I'd like to pass through to KVM.
In order to do that, I'm reading that the PCI memory region must be 4K-page
aligned and the PCI memory resources itself must also be exact multiples
of 4K pages.

I have added the following on my kernel command line:
reassign_resources reassigndev=08:09.0,08:09.1,08:09.2,08:09.3,08:09.4

But I don't know if it has any effect.  The resources are still not
sized in 4K pages.  Also, this seems to screw up the last device.

Am I doing the right thing.  I am on amd64 and using the latest qemu-kvm
release 0.12.1.2 with kernel 2.6.31.  It seems very hard to find
documentation on the kernel parameters related to pci-passthrough.


08:09.0 0c00: 1180:0832 (rev 05) (prog-if 10)
        Subsystem: 103c:30cd
        Flags: bus master, medium devsel, latency 32, IRQ 16
        Memory at f4400000 (32-bit, non-prefetchable) [size=2K]
        Capabilities: <access denied>
        Kernel driver in use: ohci1394
        Kernel modules: firewire-ohci, ohci1394

08:09.1 0805: 1180:0822 (rev 22)
        Subsystem: 103c:30cd
        Flags: bus master, medium devsel, latency 32, IRQ 18
        Memory at f4400800 (32-bit, non-prefetchable) [size=256]
        Capabilities: <access denied>
        Kernel driver in use: sdhci-pci
        Kernel modules: sdhci-pci

08:09.2 0880: 1180:0843 (rev 12)
        Subsystem: 103c:30cd
        Flags: bus master, medium devsel, latency 32, IRQ 10
        Memory at f4400c00 (32-bit, non-prefetchable) [size=256]
        Capabilities: <access denied>
        Kernel driver in use: ricoh-mmc
        Kernel modules: ricoh_mmc

08:09.3 0880: 1180:0592 (rev 12)
        Subsystem: 103c:30cd
        Flags: bus master, medium devsel, latency 32, IRQ 10
        Memory at f4401000 (32-bit, non-prefetchable) [size=256]
        Capabilities: <access denied>

08:09.4 0880: 1180:0852 (rev ff) (prog-if ff)
        !!! Unknown header type 7f


-- 
Ryan C. Underwood, <nemesis@icequake.net>

             reply	other threads:[~2010-01-09  3:09 UTC|newest]

Thread overview: 39+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2010-01-09  2:45 Ryan C. Underwood [this message]
2010-01-09  3:22 ` PCI passthrough resource remapping Alexander Graf
2010-01-10 21:53   ` Ryan C. Underwood
2010-01-10 22:15   ` Ryan C. Underwood
2010-01-14 13:59     ` Avi Kivity
2010-01-14 15:26       ` Ryan C. Underwood
2010-01-14 15:34         ` Avi Kivity
2010-01-14 15:47           ` Michael S. Tsirkin
2010-01-14 15:54             ` Avi Kivity
2010-01-14 18:31               ` Ryan C. Underwood
2010-01-14 19:09                 ` Avi Kivity
2010-01-14 19:34                   ` Ryan C. Underwood
2010-01-16  9:23                     ` Avi Kivity
2010-01-15 13:11                 ` Pasi Kärkkäinen
2010-01-15 13:15                   ` Alexander Graf
2010-03-26  2:37   ` Kenni Lund
2010-03-26  3:00     ` Brian Jackson
2010-03-29 17:23       ` Kenni Lund
2010-03-29 19:17         ` Alexander Graf
2010-03-29 23:00           ` Kenni Lund
2010-03-29 23:12             ` Alexander Graf
2010-03-29 23:47               ` Chris Wright
2010-03-30  0:21                 ` Kenni Lund
2010-03-30  2:08                   ` Chris Wright
2010-03-30 22:27                     ` Kenni Lund
2010-03-30 22:29                       ` Alexander Graf
2010-03-30 23:52                         ` Kenni Lund
2010-03-31  0:59                           ` Chris Wright
2010-03-30 23:58                       ` Chris Wright
2010-03-31  0:47                         ` Kenni Lund
2010-03-31  1:32                           ` Chris Wright
2010-03-31 10:07                             ` Kenni Lund
2010-03-31 15:15                               ` Chris Wright
2010-03-31 11:43                           ` Kenni Lund
2010-03-31 12:24                             ` Alexander Graf
2010-03-31 13:04                               ` Kenni Lund
2010-03-31 15:18                               ` Chris Wright
2010-03-31 15:23                                 ` Alexander Graf
2010-04-07  5:52                                 ` Avi Kivity

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=20100109024500.GB4727@localhost.localdomain \
    --to=nemesis@icequake.net \
    --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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox