Re: [PATCH 0 of 5] PV on HVM Xen

[Stefano Stabellini <stefano.stabellini@eu.citrix.com>]

On Wed, 17 Mar 2010, Sheng Yang wrote:
> I would like to work on the PIRQ porting upstream Linux

I am glad to hear that!


> A simple version would be just using PIRQ instead of VIRQ in my 
> patches, but PIRQ is flexible, I would see if I can do more with it

After the first part of your series gets applied, I'll rebase my pirq
remapping patches on top of it, then you can tell me what's wrong so I
can change it. I am very open to critics or suggestions.
I would also prefer that you take my pirq remapping patches, you make
any change you like, then you send your work to the list, rather than
reinventing the wheel.
I value your contributions so I would be happy if you could work on my
series, that even though is working still misses MSI support and I am
sure you'll be able to make many other important improvements.


> And we may not get a version exactly the same as pv_ops dom0 code in
> the end... I would try to make them similar and make the HVM part
> small enough, then reduce Jeremy's maintain effort.
> 

As pointed out before by Jeremy and Konrad, the best starting point is
probably Konrad's pv/pcifront-2.6.31 tree: it contains most of the pirq
stuff, ready to be upstreamed.
AFAICT the only things required to make pirq mappings work (as in
my series) that are missing are:

- xen_register_gsi

- xen_setup_pirqs

- the xen_register_gsi hook in acpi_register_gsi

the first two should be easy to port because they don't require any
change but the last one definitely needs modifications in order to be
accepted upstream.
I didn't include MSI support because is not required, but that is
another area that needs changes.