On 07/18/2013 03:23 AM, Daniel P. Berrange wrote: > On Wed, Jul 17, 2013 at 05:41:56PM -0600, Eric Blake wrote: >> So, although I like the split, I can't help but wonder if your rebase >> should take the road of adjusting things to use a callback struct, >> rather than requiring matching implementations across multiple files. > > I think the callback struct approach is useful for the case where you > have multiple possible implementations compiled in at once and we need > to choose them dynamically. Here though, we're making the choice at > compile time, so I think callback + dynamic dispatch is somewhat overkill. > > Following of the virthread approach of #include'ing .c file was my > suggestion to Roman for how to structure this. Fair enough. It just means that we don't have the compiler helping us remember to add both functions at once, so it raises the probability that someone will add a Linux-only function and forget to add the noop counterpart. But as long as we detect it before a release (that's what release candidate testing is for, right?) we should be okay. Roman, I'll accept your code with the .c inclusion, so you don't have to do the extra work of a dynamic dispatch rewrite; now it's just waiting for a rebased patch. -- Eric Blake eblake redhat com +1-919-301-3266 Libvirt virtualization library http://libvirt.org
Attachment:
signature.asc
Description: OpenPGP digital signature
-- libvir-list mailing list libvir-list@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/libvir-list