linuxppc-2.5 (mvista rsync) -- drivers/ide/ppc/pmac.c:1843: error: request for member `queue_head' in something not a structure or union

[Benjamin Herrenschmidt]

On Sun, 2003-08-10 at 22:51, Miles Lane wrote:
> On Sun August 10 2003 1:38 pm, Benjamin Herrenschmidt wrote:
> > On Sun, 2003-08-10 at 17:51, Miles Lane wrote:
> > >   CC      drivers/ide/ppc/pmac.o
> > > drivers/ide/ppc/pmac.c: In function `idepmac_wake_device':
> > > drivers/ide/ppc/pmac.c:1843: error: request for member `queue_head' in
> > > something not a structure or union
> > > drivers/ide/ppc/pmac.c: In function `idepmac_wake_drive':
> > > drivers/ide/ppc/pmac.c:1927: error: request for member `queue_head' in
> > > something not a structure or union
> >
> > I sent an updated version of this driver to linus today
>
> When you send patches to Linus to fix problems that show up in
> the linuxppc-2.5 tree, you don't also apply the patch directly to
> the linuxppc-2.5 tree?  If you and Paul use BK, couldn't you have
> Paul pull the changeset?  Is our current process the most efficient
> we could use?

Right now, I'm sending massive PowerMac driver updates directly
to Linus. Typically, all of this is available from my linuxppc-2.5-benh
tree, though I don't really recommend for non-hackers to use that.

> I am curious, why do you and Paul not use a system more like Russell's
> (changes move into Linus' tree as soon as they are fairly well tested
> and he maintains the ARM patches in

I'm trying to completely avoid having a "ppc" or "powermac" tree with
2.6, at least once all the "pending" bits from mine have been in Linus.
I want to keep as close as Linus tree as I can, not reproduce what I did
for 2.4 where I had megabytes of diffs piling up.

Ben.


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/