RE: [PATCH v6 1/2] irqchip: imx-gpcv2: IMX GPCv2 driver for wakeup sources

[Shenwei Wang]

> -----Original Message-----
> From: Shawn Guo [mailto:shawnguo@xxxxxxxxxx]
> Sent: 2015å7æ27æ 9:35
> To: Wang Shenwei-B38339
> Cc: jason@xxxxxxxxxxxxxx; Huang Yongcai-B20788;
> linux-kernel@xxxxxxxxxxxxxxx; tglx@xxxxxxxxxxxxx; shawn.guo@xxxxxxxxxx;
> linux-arm-kernel@xxxxxxxxxxxxxxxxxxx
> Subject: Re: [PATCH v6 1/2] irqchip: imx-gpcv2: IMX GPCv2 driver for wakeup
> sources
> > > On Wed, Jul 22, 2015 at 12:07:38PM -0500, Shenwei Wang wrote:
> > > > diff --git a/include/soc/imx/gpcv2.h b/include/soc/imx/gpcv2.h new
> > > > file mode 100644 index 0000000..73d6e75
> > > > --- /dev/null
> > > > +++ b/include/soc/imx/gpcv2.h
> > >
> > > I do not like this header, which couples imx7d irqchip and pm driver
> > > so much.  Can you please elaborate why we have to have this header?
> >
> > PM driver does depend on the irqchip driver. It needs some input like
> > enabled irqs and wakeup irqs to decide which module to be powered off
> > in low power states. I am also considering if the header file could be
> > removed or not. So far it seems a common place to define a structure which is
> used in both drivers is still required.
> 
> Please be more specific.  Trimming the header down to the macros and
> structures/fields that are necessary to be in the header might be a good idea.
> And then we can go through them one by one to see if there is a way to avoid
> them being in the header.

The following structure is currently used in both drivers. The members "gpc_base/
wakeup_sources/enabled_irqs" are now shared to PM driver. And the macro IMR_NUM
will be referred by both drivers too.
 
struct imx_gpcv2_irq {
	spinlock_t lock;
	void __iomem *gpc_base;
	u32 wakeup_sources[IMR_NUM];
	u32 enabled_irqs[IMR_NUM];
	u32 cpu2wakeup;
};

Thanks,
Shenwei

> Shawn