Date: Sun, 2 Dec 2007 12:25:16 +0100 From: Marko Zec <zec@imunes.net> To: Perforce Change Reviews <perforce@freebsd.org> Subject: Re: PERFORCE change 129994 for review Message-ID: <200712021225.16846.zec@imunes.net> In-Reply-To: <200712021113.lB2BDefa043362@repoman.freebsd.org> References: <200712021113.lB2BDefa043362@repoman.freebsd.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sunday 02 December 2007 12:13:40 Marko Zec wrote: > http://perforce.freebsd.org/chv.cgi?CH=129994 > > Change 129994 by zec@zec_tpx32 on 2007/12/02 11:13:11 > > IFC > > Affected files ... > > .. //depot/projects/vimage/src/sys/dev/iwi/if_iwi.c#7 edit > .. //depot/projects/vimage/src/sys/net80211/ieee80211.c#8 edit > .. //depot/projects/vimage/src/sys/net80211/ieee80211_var.h#7 edit Ahm, those three changes that unintentionally slipped into an IFC for a private branch were actually intended to provide capability for moving 80211 ifnets among vnets. This doesn't work quite properly yet, and the biggest issue here is that for the first time we are pushing stack virtualization hacks into device driver level, something that I think should be avoided like a plaque in general. Marko > .. //depot/user/zec/forwarding/sys/amd64/amd64/busdma_machdep.c#2 ... > Differences ... > > ==== //depot/projects/vimage/src/sys/dev/iwi/if_iwi.c#7 (text+ko) > ==== > > @@ -35,6 +35,8 @@ > * > http://www.intel.com/network/connectivity/products/wireless/prowirele >ss_mobile.htm */ > > +#include "opt_vimage.h" > + > #include <sys/param.h> > #include <sys/sysctl.h> > #include <sys/sockio.h> > @@ -55,6 +57,7 @@ > #include <sys/firmware.h> > #include <sys/kthread.h> > #include <sys/taskqueue.h> > +#include <sys/vimage.h> > > #include <machine/bus.h> > #include <machine/resource.h> > @@ -193,6 +196,7 @@ > static int iwi_probe(device_t); > static int iwi_attach(device_t); > static int iwi_detach(device_t); > +static void iwi_reassign(struct ifnet *, struct vnet *, char *); > static int iwi_shutdown(device_t); > static int iwi_suspend(device_t); > static int iwi_resume(device_t); > @@ -407,6 +411,9 @@ > ieee80211_ifattach(ic); > ic->ic_bmissthreshold = 10; /* override default */ > /* override default methods */ > +#ifdef VIMAGE > + ifp->if_reassign = iwi_reassign; > +#endif > ic->ic_node_alloc = iwi_node_alloc; > sc->sc_node_free = ic->ic_node_free; > ic->ic_node_free = iwi_node_free; > @@ -505,6 +512,28 @@ > return 0; > } > > +#ifdef VIMAGE > +static void > +iwi_reassign(struct ifnet *ifp, struct vnet *vnet, char *dname) > +{ > + struct iwi_softc *sc = ifp->if_softc; > + struct ieee80211com *ic = &sc->sc_ic; > + IWI_LOCK_DECL; > + > + IWI_LOCK(sc); > + bpfdetach(ifp); > + sc->sc_drvbpf = NULL; > + ieee80211_reassign(ic, vnet, dname); > + > + CURVNET_SET_QUIET(vnet); > + bpfattach2(ifp, DLT_IEEE802_11_RADIO, > + sizeof (struct ieee80211_frame) + sizeof (sc->sc_txtap), > + &sc->sc_drvbpf); > + CURVNET_RESTORE(); > + IWI_UNLOCK(sc); > +} > +#endif > > >>> TRUNCATED FOR MAIL (1000 lines) <<<
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200712021225.16846.zec>