Re: [Xen-devel] [PATCH v1 04/12] xen/hvmlite: Bootstrap HVMlite guest

["H. Peter Anvin" <hpa@xxxxxxxxx>]

On January 23, 2016 8:12:23 AM PST, Konrad Rzeszutek Wilk 
<konrad.wilk@xxxxxxxxxx> wrote:
>On January 23, 2016 11:01:06 AM EST, "H. Peter Anvin" <hpa@xxxxxxxxx>
>wrote:
>>On January 23, 2016 7:34:33 AM PST, Konrad Rzeszutek Wilk
>><konrad.wilk@xxxxxxxxxx> wrote:
>>>
>>>>However, this stub belongs in Linux, not in the Xen toolstack.  That
>>>>way, when the Linux boot protocol is modified, both sides can be
>>>>updated
>>>>accordingly.
>>>
>>>I would add that this idea is borrowed from the EFI stub code that
>>>Linux has which also constructs the boot parameter structure when
>>>invoked (either from firmware or from EFI shell).
>>
>>There is a huge difference though: EFI is a widely used multivendor
>>industry standard.  You are taking about something Xen-specific, and
>>which in good Xen tradition isn't even documented, apparently (did we
>>ever get documentation for the hypervisor ABI?)
>>
>>Asking "why burden Xen with something Linux-specific" is a pretty
>>extreme case of the tail wagging the dog.
>>
>>That being said, before any code can be put anywhere, it needs to be
>>written.  We can argue where to put it later.  We went through this
>>process with the EFI stub, too: a standalone implementation (efilinux)
>>first.
>
>http://lists.xenproject.org/archives/html/xen-devel/2015-12/msg01793.html
>
>I believe is the latest version. Roger (CCed) has probably an updated
>one.

I suspect you should write a noninteractive bootloader as a reference 
implementation, and then consider porting Grub2 and maybe Syslinux to your ABI 
for those that want a full featured interactive bootloader compatible with the 
normal management.
-- 
Sent from my Android device with K-9 Mail. Please excuse brevity and formatting.

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel