Re: ed(4) broken?

In message: <20050615202355.GA12986_at_odin.ac.hmc.edu>
            Brooks Davis <brooks_at_one-eyed-alien.net> writes:
: On Wed, Jun 15, 2005 at 01:55:52PM -0600, Warner Losh wrote:
: > From: Maxim Konovalov <maxim_at_macomnet.ru>
: > Subject: Re: ed(4) broken?
: > Date: Wed, 15 Jun 2005 23:14:52 +0400 (MSD)
: > 
: > > On Wed, 15 Jun 2005, 00:27-0600, M. Warner Losh wrote:
: > > 
: > > > In message: <20050614231344.F76340_at_mp2.macomnet.net>
: > > >             Maxim Konovalov <maxim_at_macomnet.ru> writes:
: > > > : Latest GENERIC in qemu produces
: > > > : Jun 14 19:12:04 qemu6 kernel: ed_start(0xc12e5000) GONE
: > > > : Jun 14 19:12:07 qemu6 kernel: ed0: NIC memory corrupt - invalid packet length 45
: > > > : No real hardware though.
: > > >
: > > > ed works for me on all the real hardware I have.  Maybe you could give
: > > > me more details.
: > > 
: > > Something between 2005/06/10 00:00 UTC and 2005/06/11 00:00 UTC broke
: > > qemu's ed(4).  I'm still not sure it's not qemu problem.  No
: > > additional info so far, sorry.
: > 
: > That's when the big, jumbo ifnet patch went in.  I eyeballed the
: > changes and can see nothing wrong by visual inspection.
: 
: I've looked to code over closely again and I'm supprised it works for
: anyone.  I found four instances of casts of sc to get ifp including one
: if sc_xmit.  I must have missed them when reconverting the driver after
: your cleanup. :(  I've commited a fix for them.

I don't think I've actually tested it on real hardware since June 8th,
so this isn't surprising...

Warner