On Wed, 2013-02-20 at 20:25 +0000, David Woodhouse wrote: > On Wed, 2013-02-20 at 19:20 +0100, Andreas Färber wrote: > > > And then all the *existing* uses of DO_UPCAST should be fixed to > use it > > > too? Such as in i440fx_write_config() ? > > > > Outside the scope of this patch. > > Well, maybe. But I might as well start by cleaning up the rest of the > file that was setting me such a bad example. And then maybe I'll be > familiar enough with the current coding requirements that I'll be > capable of making the simple change I *want* without it being heckled > to death :)
I've posted a patch series to fix most of this. What I *haven't* done is the instances DO_UPCAST(PIIX3State, …) because there are *two* PIIX3 classes with the same PIIX3State, so any sanity checking on the base class would have to cope with it being *either* "PIIX3" or "PIIX3-xen". Admittedly, two of them are in the creation of each and we *do* know which one we have. But two aren't... http://git.infradead.org/users/dwmw2/qemu.git (gitweb) or git://git.infradead.org/users/dwmw2/qemu.git -- dwmw2
smime.p7s
Description: S/MIME cryptographic signature
