Date: Tue, 2 Oct 2001 19:04:40 -0700 (PDT) From: Matthew Jacob <mjacob@feral.com> To: "Matthew N. Dodd" <winter@jurai.net> Cc: Julian Elischer <julian@elischer.org>, Garrett Wollman <wollman@khavrinen.lcs.mit.edu>, <freebsd-net@FreeBSD.ORG> Subject: Re: review of minor clarifying comments Message-ID: <20011002190406.W4030-100000@wonky.feral.com> In-Reply-To: <Pine.BSF.4.21.0110022142450.38937-100000@sasami.jurai.net>
next in thread | previous in thread | raw e-mail | index | archive | help
On Tue, 2 Oct 2001, Matthew N. Dodd wrote: > On Mon, 1 Oct 2001, Matthew Jacob wrote: > > Yes- this may have been true for a long time - but quite a number of > > Unix variants don't require it, so I forgot :-).. For example, neither > > OpenBSD nor NetBSD seem to require it. > > > > It turns out that this is only stumbled on rarely- but if it's all the same to > > you, a comment would help a lot. > > It looks like storing a pointer to struct arpcom in struct ifnet might > solve most of the cases where ifp->if_softc == struct arpcom/struct > ifnet. I'm not exactly sure how to deal with the miibus stuff at this > point; the following, in mii.c:miibus_linkchg(): > > mii = device_get_softc(dev); > ifp = device_get_softc(parent); > > can be changed to: > > mii = device_get_softc(dev); > ifp = mii->mii_ifp; > > I'm thinking we could pass struct ifnet in to mii.c:mii_phy_probe() and > attach it to the device ivars. > I would think that the latter makes the most sense. It all doesn't matter *too* much as long as what the requirements are are reasonably documented or known. -matt To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20011002190406.W4030-100000>