[PATCH v2] omap4-fb: add driver

[Christoph Fritz]

*ping*

On Thu, 2013-04-11 at 21:01 +0200, Christoph Fritz wrote:
> Hi Jean-Christophe
> 
> On Tue, 2013-04-09 at 15:29 +0200, Jean-Christophe PLAGNIOL-VILLARD
> wrote:
> > On 09:16 Tue 09 Apr     , Sascha Hauer wrote:
> > > On Mon, Apr 08, 2013 at 09:17:15AM +0200, Jean-Christophe PLAGNIOL-VILLARD wrote:
> > > > On 00:15 Mon 08 Apr     , Christoph Fritz wrote:
> > > > > This patch adds omap4 display controller support.
> > > > > 
> > > > > +
> > > > > +static inline void fb_write(uint32_t v, void __iomem *addr)
> > > > > +{
> > > > > +	__raw_writel(v, addr);
> > > > > +}
> > > > > +
> > > > > +static inline uint32_t fb_read(void const __iomem *addr)
> > > > > +{
> > > > > +	return __raw_readl(addr);
> > > > > +}
> > > > what is the dif?
> > > 
> > > Maybe Christoph understands what you mean here. I don't
> > those 2 inline as useless
> 
> You want me to purge the "inline" because gcc is smart enough to use
> inlining on its own?
> 
> Or are you noting that using a function here is overkill for only one
> call inside?
> 
> I use functions here because it makes debugging more easy and if you
> have a weird hardware config, here you can easily adapt the read and
> write functions.
> 
> > > > > +
> > > > > +	fbi->regs.dss   = dev_request_mem_region(dev, OMAP4_FB_RESOURCE_DSS);
> > > > > +	fbi->regs.dispc = dev_request_mem_region(dev, OMAP4_FB_RESOURCE_DISPC);
> > > > use named dev_request
> > > 
> > > We don't have such a function.
> > yes we do
> > 
> > dev_request_mem_region_by_name
> 
> Ok
> 
>  Thanks
>   -- Christoph
>