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

Re: [Xen-devel] [PATCH] x86_64 trap table fixup




On 4 Jul 2005, at 08:46, Jan Beulich wrote:

One remark since this code is getting touched now anyway: You appear to
have taken the same (somewhat broken) approach of handling #BR as Linux
does. Architecturally, IDT entry 5 should be a DPL=0 one (permitting
BOUND, but disallowing INT 05); is there any particular reason you want
to allow INT 05 here (which could get you in conflict with hosting
Windows, which appears to deal with this the architecture intended way)?

The guest OS is allowed to set the DPL as it wants -- in the case of xenlinux we want to retain maximum compatibility with native Linux. Other paravirtualised OSes can set the DPL to zero if they wish. In the case of Windows, that will only run on VT/SVM hardware anyway, and that deals with handling the guest's IDT pretty much automatically, so these paravirtual interfaces aren;t used at all.

 -- Keir


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


 


Rackspace

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