46dcfc8 (ARM: Update mach-types) removes used machine types

[linux@arm.linux.org.uk (Russell King - ARM Linux)]

On Mon, Aug 22, 2011 at 01:49:30PM +0100, Russell King - ARM Linux wrote:
> On Mon, Aug 22, 2011 at 01:18:25PM +0200, Marek Vasut wrote:
> > On Monday, August 22, 2011 10:33:29 AM Uwe Kleine-K?nig wrote:
> > > On Mon, Aug 22, 2011 at 09:17:49AM +0100, Russell King - ARM Linux wrote:
> > > > On Mon, Aug 22, 2011 at 10:10:00AM +0200, Uwe Kleine-K?nig wrote:
> > > > > Hello Russell,
> > > > > 
> > > > > your commit 46dcfc8 that is in today's next among other changes does:
> > > > > 
> > > > > -eukrea_cpuimx25sd      MACH_EUKREA_CPUIMX25    EUKREA_CPUIMX25 2820
> > > > > -eukrea_cpuimx35sd      MACH_EUKREA_CPUIMX35    EUKREA_CPUIMX35 2821
> > > > > 
> > > > > These two are both marked as "mainlined" on
> > > > > http://www.arm.linux.org.uk/developer/machines/ and both are actually
> > > > > used (arch/arm/mach-imx/mach-{eukrea_cpuimx25,cpuimx35.c})
> > > > > 
> > > > > Can you please add them back? Maybe it's worth to double-check the
> > > > > other deletions?!
> > > > 
> > > > No.  They're removed because they're broken (look at them) - I've added
> > > > additional checks so I don't have to waste _significant_ time fixing
> > > > the entries time and time again.
> > > > 
> > > > The requirement is that the machine_is_xxx() must match the MACH_xxx
> > > > parts.  If they don't, they will not exist in future updates.
> > > 
> > > Ah, I wasn't aware of that requirement.
> > > 
> > > Eric, these machine types are owned by you. Do you care to follow up?
> > 
> > Hi Russell,
> > 
> > maybe it'd be worth at least pinging the people before you kill off the machine 
> > so they can come up with fix?
> 
> I've no idea which get killed off, sorry.

... and how about people *TALK* to me when they fuck up their entries in
the machine database.

TBH, the way I feel _right_ _now_ over it is that people who don't get
what they deserve, and their stuff breaks.  Again, my perspective is
that that is their own fault, and they will have to live with it.

Given all this whinging - first about people throwing stuff into the
kernel without the corresponding stuff in the mach-types file first, and
now this, I'm *seriously* considering killing the damned thing off
completely and not doing any further updates to the file anymore.  It's
just *far* too much hastle to deal with given all the absolute shite
that's now there.

And if I'm not allowed to sort it out in an automated way then I'm
just going to give up with it and shut the entire thing down.