Re: [PATCH 00/48] Make the IRQCHIP_DECLARE macro globally accessible

[Thomas Gleixner]

On Thu, 2 Jul 2015, JoÃl Porquet wrote:
> On Thursday, July 02, 2015 12:33:05 PM Vineet Gupta wrote:
> > On Thursday 02 July 2015 04:02 AM, Joel Porquet wrote:
> > > At the moment the IRQCHIP_DECLARE macro is only declared locally in
> > > 'drivers/irqchip/irqchip.h'. That prevents from using it directly in
> > > arch/*
> > > directories whenever irqchip drivers only exist there, which happens in a
> > > few cases (e.g. arc, arm, microblaze and mips).
> > > 
> > > This patch makes the macro to be globally defined, in
> > > include/linux/irqchip.h, and thus usable for arch-specific declarations
> > > of irqchip drivers. In this way, it is very similar to what clocksource
> > > does (ie CLOCKSOURCE_OF_DECLARE is defined in
> > > include/linux/clocksource.h).
> > > 
> > > I split up everything into patches to make the integration easier. Please
> > > let me know if it's not, and in such case how to make it better.
> > > 
> > > For now, patch 01 of this series transfers the declaration of the macro
> > > IRQCHIP_DECLARE to the global header 'include/linux/irqchip.h'. The
> > > following patches, from 02 to 47, modify all the irqchip drivers that use
> > > IRQCHIP_DECLARE, one by one. And finally, the last patch 48 removes the
> > > private and now useless header 'drivers/irqchip/irqchip.h'.
> > 
> > Hi Joel,
> > 
> > I don't see the rest of series on lkml and/or the patch which touches
> > arch/arc. Also, you may wanna redo this after 4.2-rc1 anyways. For ARC
> > atleast, there's a new intc which would also require similar fixup. There
> > might be others ....
> > 
> > Thx,
> > -Vineet
> 
> Hi Vineet (and all),
> 
> Sorry for the mistake, I hope I didn't spam anyone (too much). I realized to 
> late that sending about fifty patches to 26 recipients was probably not a good 
> idea, and my smtp provider would have blocked me before the end anyway.
> 
> Therefore I will follow your suggestion and wait until after 4.2-rc1. Then 
> I'll resubmit a new patchset that takes into account the new intc(s) as well.
> 
> But since this patchset affects many files across several drivers and 
> architectures, what would be the best way to submit it?
> 
> Would it be OK to send the cover to all the maintainers/mailing-lists involved 
> in order to inform them that a patchset is affecting their respective 
> subsystem, but to send the patches only on the kernel mailing-list?
> 
> And/or is there someone in particular who is in charge to integrate such a 
> transversal patchset?

The best thing is to move the macro to include/linux/irqchip.h now and
include linux/irqchip in drivers/irqchip/irqchip.h.

That's a safe change and I can queue it right away and merge it into
rc1. 

So after that I can queue the drivers/irqchip patches in my tree and
remove local header file for 4.3. The changes for stuff which is
outside of drivers/irqchip and depends on the global visible macro can
be queued in the relevant maintainer trees.

Can you send me a patch which moves the macro and includes
linux/irqchip from the local header file?

Thanks,

	tglx