Re: [GIT PULL] PCI: dt: Remove magic numbers for legacy PCI IRQ interrupts

[Andrew Murray <andrew.murray@arm.com>]

On Wed, Dec 11, 2019 at 01:33:37PM -0600, Rob Herring wrote:
> On Wed, Dec 11, 2019 at 10:59 AM Olof Johansson <olof@lixom.net> wrote:
> >
> > On Wed, Dec 11, 2019 at 04:18:08PM +0000, Andrew Murray wrote:
> > > Hi Arnd,
> > >
> > > Please consider this pull request.
> > >
> > > The following changes since commit e42617b825f8073569da76dc4510bfa019b1c35a:
> > >
> > >   Linux 5.5-rc1 (2019-12-08 14:57:55 -0800)
> > >
> > > are available in the Git repository at:
> > >
> > >   git://linux-arm.org/linux-am.git tags/pci-dt-intx-defines-5.5-rc1
> > >
> > > for you to fetch changes up to d50e85b9ad3d4287ab3c5108b7b36ad4fd50e5b4:
> > >
> > >   dt-bindings: PCI: Use IRQ flags for legacy PCI IRQ interrupts (2019-12-11 16:05:55 +0000)
> > >
> > > ----------------------------------------------------------------
> > > PCI: dt: Remove magic numbers for legacy PCI IRQ interrupts
> > >
> > > PCI devices can trigger interrupts via 4 physical/virtual lines known
> > > as INTA, INTB, INTC or INTD. Due to interrupt swizzling it is often
> > > required to describe the interrupt mapping in the device tree. Let's
> > > avoid the existing magic numbers and replace them with a #define to
> > > improve clarity.
> > >
> > > This is based on v5.5-rc1. As this series covers multiple architectures
> > > and updates include/dt-bindings it was felt that it may be more
> > > convenient to merge in one go.
> >
> > That's a pretty high-effort way of doing this, with potential for messy
> > conflicts.
> >
> > The standard way of making sweeping changes across the tree is usually to
> > get the new interface/definition added in one release, and then moving
> > usage over through the various maintainers in the release after since
> > the define is then in the base tree for everybody. Would you mind using

OK, thanks for this - noted for next time.


> > the same approach here, please? Especially since this is mostly a cleanup.
> 
> Yeah, it's already going to conflict with some PCI controller schema
> conversions pending.
> 
> I'm happy to apply the header for 5.5-rc2. Then send the dts changes
> to Arnd/Olof and the binding changes to Lorenzo.

Much appreciated.

I'll rebase the dts and doc in a subsequent release.

Thanks,

Andrew Murray

> 
> Rob