Re: [patch 37/38] mips: vr41xx: Cleanup the direct access to irq_desc[]

[Thomas Gleixner <tglx@linutronix.de>]

On Thu, 24 Mar 2011, Sergei Shtylyov wrote:

> Hello.
> 
> On 24-03-2011 16:00, Thomas Gleixner wrote:
> 
> > > > -               if (!(desc->status&   IRQ_DISABLED)&&   
> > > > desc->chip->unmask)
> > > > -                       desc->chip->unmask(source_irq);
> > > > +               if (!(desc->status&   IRQ_DISABLED)&&   
> > > > chip->irq_unmask)
> > > > +                       chip->irq_unmask(idata);
> 
> > >     Hm, doesn't this (I mean the old) code break after the previous patch?
> 
> > Not as long as the compat functions are active in the core.
> 
>    I've looked at compat_*() before replying: it seems that they work vice
> versa, i.e. the new functions are emulated by calling the old, and you're
> moving away from old to new in the previous patch. Maybe I miss something...

Oops. Yes. So the patches should be folded