Samuel, I was actually referring to the equivalent of dedicating a PCI HBA to a
paravirtualized guest(using pciback) in an HVM guest. I believe this is
possible with VT-d hardware and with the Xen 3.2.1 kernel for an HVM as well,
although I haven't been able to do this myself yet to an HVM. In any case 2^57
is large enough for any domain to keep going for some time, with h/w dedicated
or not.
Regards,
Bhaskar.
-----Original Message-----
From: Samuel Thibault [mailto:samuel.thibault@xxxxxxxxxxxxx]
Sent: Tuesday, July 22, 2008 2:59 PM
To: Jayaraman, Bhaskar
Cc: Sebastian Reitenbach; xen-devel@xxxxxxxxxxxxxxxxxxx
Subject: Re: [Xen-devel] maximum size of a block device in a domU
Jayaraman, Bhaskar, le Tue 22 Jul 2008 13:18:23 +0800, a écrit :
> Hi, if I were using Xen 3.2.1 and above and if I were to direct assign a
> block device to a HVM guest will I still play by the rules mentioned below
> for maximum size of the device?
If by "direct assign" you mean setting disk = [ 'phy:sda4,hda,w' ] in
the config file, then yes you are still going through the IDE emulation,
limited to 2^48 sectors, i.e. 2^57B, plus the dom0 limitations (and the
HVM domU limitations of course).
Samuel
_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-devel
|
Home