RE: [Xen-devel] [PATCH] Fix build failure in 32 environment

["Jiang, Yunhong" <yunhong.jiang@xxxxxxxxx>]

This patch itself is only for 32-bit version compile success.

>From this bug, seems the 32-bit xen hypervisor is really not so populated.

--jyh

>-----Original Message-----
>From: xen-devel-bounces@xxxxxxxxxxxxxxxxxxx
>[mailto:xen-devel-bounces@xxxxxxxxxxxxxxxxxxx] On Behalf Of Dan Magenheimer
>Sent: Tuesday, December 15, 2009 12:53 AM
>To: Keir Fraser; Jiang, Yunhong
>Cc: Xu, Dongxiao; xen-devel@xxxxxxxxxxxxxxxxxxx
>Subject: RE: [Xen-devel] [PATCH] Fix build failure in 32 environment
>
>> > If I remember correctly, rdtscp is a valid instruction
>> > in 32-bit mode so why are you disabling support for
>> > 32-bit Xen?
>> >
>> > Personally, I don't care if 32-bit Xen is supported,
>> > but either we need to say it is supported or it is
>> > not supported, not have new features randomly choose
>> > to support it or not, especially features that have
>> > impact on userland applications.
>>
>> It can be supported for existign features but not all new ones. We can
>> perhaps retire it all together after 4.0.
>
>IMHO, this is not a new feature, this is an ABI incompatibility
>issue. If rdtscp is a valid 32-bit instruction, IMHO it
>should either work on both 32-bit Xen and 64-bit Xen or
>neither.
>
>Was this patch a quick hack to allow the code to compile
>for 32-bit Xen or was it a reasoned decision to disallow
>support on 32-bit Xen?
>
>(And, in any case, if rdtscp is a "new feature", why was
>the patch accepted post-freeze, especially when there
>is still lively debate going on? ;-)
>
>_______________________________________________
>Xen-devel mailing list
>Xen-devel@xxxxxxxxxxxxxxxxxxx
>http://lists.xensource.com/xen-devel

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