[PATCH 3.16.y-ckt 053/133] ARM: orion5x: fix legacy orion5x IRQ numbers

[Luis Henriques]

On Thu, Oct 15, 2015 at 11:03:19AM +0200, Benjamin Cama wrote:
> Hi everyone,
> 
> Le 2015-10-15 08:22, Gregory CLEMENT a écrit :
> > On jeu., oct. 15 2015, Ben Hutchings <ben at decadent.org.uk> wrote:
> >
> >>On Wed, 2015-09-30 at 10:50 +0100, Luis Henriques wrote:
> >>>3.16.7-ckt18 -stable review patch.  If anyone has any objections,
> >>>please let me know.
> >>>
> >>>------------------
> >>>
> >>>From: Benjamin Cama <benoar at dolka.fr>
> >>>
> >>>commit 5be9fc23cdb42e1d383ecc8eae8a8ff70a752708 upstream.
> >>>
> >>>Since v3.18, attempts to deliver IRQ0 are rejected, breaking orion5x.
> >>[...]
> >>
> >>But I don't think this was needed for 3.16-ckt, was it?
> >
> >Indeed it fixes a commit (a71b092a9c68 ("ARM: Convert handle_IRQ to use
> >__handle_domain_irq")) that was introduced in 3.18:
> >
> >git tag --contains a71b092a9c68 | grep "^v3\." | sort | head -1
> >v3.18
> >
> >So please Luis do not apply it on 3.16-ckt unless you already back port
> >the commit a71b092a9c68 ("ARM: Convert handle_IRQ to use
> >__handle_domain_irq") on this branch.
> 
> Indeed it is not completely needed, but to be clear it "just" shifts the
> IRQs by one, so it shouldn't do any harm either. So it is up to you. I would
> understand that you choose not to apply it to be on the safe side. BTW, it
> seems f28d7de is required by my patch, appearing in 3.14; as I don't follow
> kernel development closely, I could not assure you my patch is completly
> safe for 3.16.
>

*Sigh*
Sorry for letting this commit getting in.  I'm reverting it for the next
3.16 kernel release.  Thank you all!

Cheers,
--
Luís