Re: [Xen-devel] [PATCH] MAINTAINERS: Add myself as x86 I/O emulation and viridian maintainer

["Jan Beulich" <JBeulich@xxxxxxxx>]

>>> On 31.03.16 at 10:00, <Paul.Durrant@xxxxxxxxxx> wrote:
>>  -----Original Message-----
>> From: Xen-devel [mailto:xen-devel-bounces@xxxxxxxxxxxxx] On Behalf Of Jan
>> Beulich
>> Sent: 31 March 2016 08:52
>> To: Paul Durrant
>> Cc: xen-devel@xxxxxxxxxxxxxxxxxxxx 
>> Subject: Re: [Xen-devel] [PATCH] MAINTAINERS: Add myself as x86 I/O
>> emulation and viridian maintainer
>> 
>> >>> On 31.03.16 at 09:28, <paul.durrant@xxxxxxxxxx> wrote:
>> > I have made many modifications to this code over the past few years
>> > so I'm probably the one most familiar with it.
>> >
>> > Signed-off-by: Paul Durrant <paul.durrant@xxxxxxxxxx>
>> > ---
>> >  MAINTAINERS | 16 ++++++++++++++++
>> >  1 file changed, 16 insertions(+)
>> >
>> > diff --git a/MAINTAINERS b/MAINTAINERS
>> > index e765311..43bbb8f 100644
>> > --- a/MAINTAINERS
>> > +++ b/MAINTAINERS
>> > @@ -384,6 +384,16 @@ F:    xen/include/asm-x86/
>> >  F:        tools/firmware/hvmloader/
>> >  F:        tools/tests/x86_emulator/
>> >
>> > +X86 I/O EMULATION
>> > +M:        Paul Durrant <paul.durrant@xxxxxxxxxx>
>> > +S:        Supported
>> > +F:        xen/arch/x86/hvm/emulate.c
>> > +F:        xen/arch/x86/hvm/hvm.c
>> 
>> This file doesn't really belong here (as containing a lot of other
>> stuff).
> 
> I'm concerned about missing any changes to the ioreq server code that's in 
> there. Would it perhaps be better to split that code out at this point?

Yes, splitting out at least all the ioreq server code is likely a good
idea, and once split out the new file should be added here.

>> Instead I think you also want ...
>> 
>> > +F:        xen/arch/x86/hvm/intercept.c
>> 
>> F:   xen/arch/x86/hvm/io.c
> 
> Yes, I missed that.
> 
>> 
>> > +F:        xen/include/asm-x86/hvm/hvm.h
>> 
>> And this again doesn't belong here, while
>> 
>> F:   xen/include/asm-x86/hvm/emulate.h
>> 
>> seems missing.
>> 
> 
> Yes, that should be there too.
> 
>> If you agree, I can do these adjustments while committing.
>> 
> 
> I agree to the additions and dropping hvm.h, but I'd like to keep hvm.c 
> pending potentially splitting out the ioreq server code.

One of the topics I intend to bring up on the hackathon is the
ambiguous meaning of such an entry: It's never been clear to me
whether a more narrow F: in one section would override a wider
F: in another. In the case here such an override would not be
intended. Otoh x86/mm/shadow/ vs x86/mm/ vs x86/ all are more
likely to imply such an override. Until that's fully clarified I'm not
agreeing to add hvm.c here.

Jan


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