[PATCH v2 3/4] drm/msm: get rid of msm_iomap_size

Bjorn Andersson bjorn.andersson at linaro.org
Wed Apr 28 13:59:01 UTC 2021


On Wed 28 Apr 08:41 CDT 2021, Dmitry Baryshkov wrote:

> On 28/04/2021 05:47, Bjorn Andersson wrote:
> > On Mon 26 Apr 19:18 CDT 2021, Dmitry Baryshkov wrote:
> > [..]
> > > diff --git a/drivers/gpu/drm/msm/msm_drv.c b/drivers/gpu/drm/msm/msm_drv.c
> > > index 92fe844b517b..be578fc4e54f 100644
> > > --- a/drivers/gpu/drm/msm/msm_drv.c
> > > +++ b/drivers/gpu/drm/msm/msm_drv.c
> > > @@ -124,7 +124,7 @@ struct clk *msm_clk_get(struct platform_device *pdev, const char *name)
> > >   }
> > >   static void __iomem *_msm_ioremap(struct platform_device *pdev, const char *name,
> > > -				  const char *dbgname, bool quiet)
> > > +				  const char *dbgname, bool quiet, phys_addr_t *psize)
> > 
> > size_t sounds like a better fit for psize...
> 
> I was trying to select between size_t and phys_addr_t, settling on the
> latter one because it is used for resource size.
> 

I always thought resource_size_t was an alias for size_t, now I know :)

That said, I still think that size_t (in line with resource_size_t)
gives a better hint about what the parameter represents...

Regards,
Bjorn


More information about the dri-devel mailing list