Xen devel - Oct 2009 - [PATCH: IOEMU & XEN] Change ioreq layout and reduce its size

[Ian, please apply the attached ioemu patch and then send me an updated
QEMU_TAG. The change to the ioreq layout requires the ioemu and xen-unstable
patch check-ins to be closely synchronised.]

Reduce size of Xen-qemu shared ioreq structure to 32 bytes. This has two
advantages:
 1. We can support up to 128 VCPUs with a single shared page
 2. If/when we want to go beyond 128 VCPUs, a whole number of ioreq_t
structures will pack into a single shared page, so a multi-page array will
have no ioreq_t straddling a page boundary

Also, while modifying qemu, replace a 32-entry vcpu-indexed array with a
dynamically-allocated array.

Signed-off-by: Keir Fraser <keir.fraser@citrix.com>




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

Keir Fraser wrote:
> [Ian, please apply the attached ioemu patch and then send me an
> updated QEMU_TAG. The change to the ioreq layout requires the ioemu
> and xen-unstable patch check-ins to be closely synchronised.]
> 
> Reduce size of Xen-qemu shared ioreq structure to 32 bytes. This has
> two advantages:
Hi, Keir, 
	Reduce the ioreq structure has these advantages indeed, but do we need to
consider the following compatibility issues? Thanks!
1) HVM guest could migrate/save/restore between the new and old hypervisor if
vcpu number is less than 32.
2) Old qemu could work with new hypervisor if guest vcpu number is less than 32.

Best Regards, 
-- Dongxiao
>  1. We can support up to 128 VCPUs with a single shared page
>  2. If/when we want to go beyond 128 VCPUs, a whole number of ioreq_t
> structures will pack into a single shared page, so a multi-page array
> will have no ioreq_t straddling a page boundary
> 
> Also, while modifying qemu, replace a 32-entry vcpu-indexed array
> with a dynamically-allocated array.
> 
> Signed-off-by: Keir Fraser <keir.fraser@citrix.com>
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

On 29/10/2009 13:21, "Xu, Dongxiao" <dongxiao.xu@intel.com>
wrote:
>> Reduce size of Xen-qemu shared ioreq structure to 32 bytes. This has
>> two advantages:
> 
> Hi, Keir, 
> Reduce the ioreq structure has these advantages indeed, but do we need to
> consider the following compatibility issues? Thanks!
> 1) HVM guest could migrate/save/restore between the new and old hypervisor
if
> vcpu number is less than 32.
It still can, as the ioreq struct is private to the hypervisor and qemu,
neither of which is migrated/saved with the guest.
> 2) Old qemu could work with new hypervisor if guest vcpu number is less
than
> 32. 
Qemu, like the toolstack, is expected to be upgraded with the hypervisor.

 -- Keir



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