Blue Swirl <blauwir...@gmail.com> writes: > On Tue, Feb 15, 2011 at 12:22 PM, Markus Armbruster <arm...@redhat.com> wrote: >> Blue Swirl <blauwir...@gmail.com> writes: >> >>> On Sat, Feb 12, 2011 at 6:57 PM, Markus Armbruster <arm...@redhat.com> >>> wrote: [...] >>>> Old code has pc_init1() call vmport_init(). Where does your code create >>>> qdev "vmport"? And what's happening with port_state? It's still used >>>> by vmport_register(), but no longer connected to the I/O ports. Can't >>>> see how vmport_register() has any effect anymore. >>> >>> I fixed it in the committed version. >> >> Did you post v2 to the list for review? > > No, since v1 got no review.
*Please* don't do that. >>>> Have you tested this? >>> >>> Sure. >> >> Here's how your v2 creates and initializes the qdev: [...] >> Maybe I'm dense, but I can't see how this can work. > > Good catch, it doesn't. Probably vmport_register() should take > ISADevice* parameter to provide the state, instead of using static > state (which would be easy one-line change). > > But if all this is going to be thrown into ps2.c, it may not be > necessary. The whole concept of registration may become useless. Yes.