[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-users] PCI passthrough partial success


  • To: "Andy Burns" <lists.xensource.com@xxxxxxxxxxxxxx>
  • From: "Christopher Isip" <cmisip@xxxxxxxxx>
  • Date: Wed, 2 Jul 2008 18:27:46 -0400
  • Cc: xen-users@xxxxxxxxxxxxxxxxxxx
  • Delivery-date: Wed, 02 Jul 2008 15:28:24 -0700
  • Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=message-id:date:from:to:subject:cc:in-reply-to:mime-version :content-type:references; b=hZH1g+RWi7akz8pVksClmyPDCZCadlbOGzQranRj9/rGg+LKtuooi+SZ5bJp4IvOUE atQU8lDn81LkstiSTgjlNvGctd7HF83KK6DxH6ZrMMSz5q7SDrPkq4NcibNzjthRbGYQ i+Cvjdk26kM/t0k2PnJvXMMQr4hGQ9xWicDGE=
  • List-id: Xen user discussion <xen-users.lists.xensource.com>



On Sun, Jun 29, 2008 at 7:48 AM, Andy Burns <lists.xensource.com@adslpipe.co.uk> wrote:
On 29/06/2008 09:49, Andy Burns wrote:

# dmesg|grep -i saa

saa7130/34: v4l2 driver version 0.2.14 loaded
saa7130[0]: found at 0000:00:00.0, rev: 1, irq: 17, latency: 64, mmio: 0xfebffc00
saa7130[0]: subsystem: 185b:c901, board: Compro Videomate DVB-T200 [card=71,autodetected]
saa7130[0]: can't ioremap() MMIO memory
saa7134: probe of 0000:00:00.0 failed with error -5

Sorry, I meant to include another message I see in the domU each time the domU boots.

# xm dmesg

(XEN) mm.c:625:d6 Non-privileged (6) attempt to map I/O space 000fec00




_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-users

I have mythbackend running in an Ubuntu-Hardy domU with a Centos 5.1 dom0.  Here are the settings that work for me:

In dom0 /boot/grub/:

title CentOS (2.6.18-xen_3.2.0)
    root (hd0,5)
    kernel /boot/xen.gz-3.2 swiotlb=256 noirqdebug
    module /boot/vmlinuz-2.6-xen ro root=LABEL=/  max_loop=32
    module /boot/initrd-2.6-xen.img   

In /etc/xen/Ubuntu-Hardy-Mythtv.cfg:

bootloader = '/usr/bin/pygrub'
memory = 640
name = "Ubuntu-Mythtv"
disk = ['file:/mnt/VM/Ubuntu-Hardy-Mythtv.img,hda1,w','file:/mnt/VM/Ubuntu-Hardy-Mythtv-swap.img,hda2,w']
vif = [ '' ]
pci = [ "0000:03:02.0","0000:04:08.0","0000:04:09.0","0000:02:00.0","0000:02:00.1" ]
extra = "swiotlb=force all-generic-ide=1 clock=jiffies"
sdl = 1

In the domU /boot/grub/menu.1st:

title        Ubuntu 8.04, kernel 2.6.24-16-xen
root        (hd0,0)
kernel        /boot/vmlinuz-2.6.24-16-xen root=/dev/hda1 ro console=xvc0 swiotlb=128,force
initrd        /boot/initrd.img-2.6.24-16-xen

I have a jmicron pci express SATA-PATA combo card pcibacked, hence the all-generic-ide line above.

For some reason jiffies in extra doesn't stick. Also, I followed a recommendation from the mythv list to chrt the ivtv pids. So I have a boot script in the domU to run:
#!/bin/bash
/bin/echo jiffies > /sys/devices/system/clocksource/clocksource0/current_clocksource
/usr/bin/chrt -rp 99 `pgrep ivtv0`
/usr/bin/chrt -rp 99 `pgrep ivtv1`
/usr/bin/chrt -rp 99 `pgrep ivtv2`

#It probably wont hurt to increase the domU's scheduling priority, run this in dom0:

xm sched-credit -d Ubuntu-Mythtv -w 512

I would suggest having the Mythtv backend not do any transcoding or commflagging.  Instead, create slave mythtv backends without tuners and have them do the transcode and commflagging.  This effectively compartmentalizes the commflagger and transcoder so that the main mythtv backend will not drop frames when writing to the disk.

My only problem with the system is I get dma timeout errors in dmesg.  These were present even in a nonxen setup, so I think it is related to my hardware and ivtv.

Hope that helps.  I am still trying to find ways to optimize the mythbackend domU so If you have any ideas, please let me know.

Chris

_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-users

 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.