Re: Diff between Linus' and linux-mips git: tulip

[Peter Horton <pdh@colonel-panic.org>]

On Fri, Feb 24, 2006 at 01:13:27AM +0000, Martin Michlmayr wrote:
> * Peter Horton <pdh@colonel-panic.org> [2006-02-20 23:03]:
> > > -                       /* No media table either */
> > > -                       tp->flags &= ~HAS_MEDIA_TABLE;
> > > +		       /* Ensure our media table fixup get's applied */
> > > +		       memcpy(ee_data + 16, ee_data, 8);
> > >  #endif
> > >  #ifdef CONFIG_MIPS_COBALT
> > Didn't the memcpy() used to be inside the CONFIG_MIPS_COBALT section ?
> > Looking at tulip/eeprom.c I can't work out why it was ever there though
> 
> Yeah, and it's still there in the Cobalt section.  But now (in the
> mips tree) it's _also_ there for CONFIG_DDB5477.  So I have several
> questions:
>  - can we just get rid of the code between CONFIG_MIPS_COBALT?
>  - should the CONFIG_DDB5477 change be reverted (probably), and do we
>    need these special cases for CONFIG_DDB* anyway or can they be
>    solved in a better way (e.g. by putting something in eeprom.c).
> 
> It seems mips is the only arch that mucks around with "#ifdef CONFIG_"
> in this file...

Looks like both the "sa_offset = 0" and "memcpy(...)" are required to
ensure our fixup from tulip/eeprom.c gets applied. I don't know why it
only seems to be us that needs it :-(

P.