On Sat, Nov 03, 2001 at 02:32:18PM -0500, Jim McGrath wrote:
> It makes the assumption that the ifnet/arpcom structure is at the beginning
> of the softc structure. In needs to do this to maintain media independence.
whoops, you are right... it is really bad that dev->softc
is a void* and there
TED]]
Sent: Saturday, November 03, 2001 2:06 PM
To: Jim McGrath
Cc: [EMAIL PROTECTED]; [EMAIL PROTECTED]
Subject: Re: softc, ifnet and arpcom overlay
On Sat, Nov 03, 2001 at 01:17:24PM -0500, Jim McGrath wrote:
> Check mii.c, miibus_attach()
I don't see anything strange there, the
On Sat, Nov 03, 2001 at 01:17:24PM -0500, Jim McGrath wrote:
> Check mii.c, miibus_attach()
I don't see anything strange there, the access is through
device_get_softc(dev) which presumably does the right thing
and makes no assumption on structure layout.
cheers
luigi
Check mii.c, miibus_attach()
Jim
-Original Message-
From: Luigi Rizzo [mailto:[EMAIL PROTECTED]]
Sent: Saturday, November 03, 2001 12:12 PM
To: [EMAIL PROTECTED]; [EMAIL PROTECTED]
Subject: softc, ifnet and arpcom overlay
Hi,
i remember early in october this topic came out,
but while
Hi,
i remember early in october this topic came out,
but while i can easily find out where the arpcom/ifnet overlay is used,
i am not totally certain about the softc/ifnet overlay.
For first, there is a reference from ifnet to the corresponding
softc, so in principle the overlay is totally unnece
