OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

virtio-comment message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]


Subject: Re: [virtio] [OASIS Issue Tracker] Created: (VIRTIO-9) Change config space, ring and headers to always be little endian


Stefan Hajnoczi <stefanha@redhat.com> writes:
> On Mon, Aug 19, 2013 at 05:57:49PM +0930, Rusty Russell wrote:
>> OASIS Issues Tracker <workgroup_mailer@lists.oasis-open.org> writes:
>
> What is the transition plan for existing virtio implementations?
>
> On x86 it's not an issue since guest endian == little-endian.  Not sure
> about others.

Indeed!  I didn't cover that at all.  We're supposed to produce a
separate transition guide: it's a deliverable after the main spec
itself.

For PCI, we're looking at using PCI capabilities to describe the device:
we want more feature bits and other changes anyway.  The new endianness
(and any other changes) would be tied to that new layout.

For MMIO and CCW, there's no equivalent transition point proposed.  A
feature bit?  Some other boutique marker?  See ongoing discussion...

>>  	struct virtio_blk_req {
>> -		u32 type;
>> -		u32 ioprio;
>> -		u64 sector;
>> +		le32 type;
>> +		le32 ioprio;
>> +		le64 sector;
>>  		char data[][512];
>> -		u8 status;
>> +		le8 status;
>
> spurious change, there is no le8.

Oops, thanks.  Overzealous search/replace.

Cheers,
Rusty.



[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]