Re: [Xen-devel] [PATCH 07/10] libxl: New API for providing OS events to libxl

[Ian Jackson <Ian.Jackson@xxxxxxxxxxxxx>]

Stefano Stabellini writes ("Re: [Xen-devel] [PATCH 07/10] libxl: New API for 
providing OS events to libxl"):
> I still prefer the approach prototyped in
> http://marc.info/?l=xen-devel&m=132371797519877: avoid introducing
> restrictions on how libxl functions have to be called, unify egc and gc,
> make it possible to take the lock recursively and use the nested counter
> to figure out when to call the callbacks.
> It would make my head hurt less when I have to read/write this code,
> however others might react differently.

I would agree with you if it were the case that editing general code
in libxl might involve adding new event-generating functions, or
indeed might involve turning existing functions into event-generating
ones.

But this is not the case.  The event-generating portions of the code
end up completely separate in any correct implementation - and someone
using this machinery doesn't even need to be very aware of this.

> Only one more thing: I would kindly ask to move all these event related
> functions to a different source file, to make it easier for people to
> understand which ones are different from the rest of the library.

They are, in libxl_event.c.  Unless by "event related" you meant
"event generating", in which case I disagree, because I don't think
that's the most relevant distinction.

Ian.

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