Re: [PATCH] s390/mm: Make virt_to_pfn() a static inline

[Alexander Gordeev]

On Fri, Aug 11, 2023 at 07:49:01PM +0200, Linus Walleij wrote:
> On Fri, Aug 11, 2023 at 3:44 PM Alexander Gordeev
> <agordeev@xxxxxxxxxxxxx> wrote:
> 
> > Funnily enough, except drivers/s390/char/vmcp.c none of affected
> > code pieces below is an offender. But anyway, to me it looks like
> > a nice improvement.
> 
> I'm puzzled, vmcp.c is a char * so actually not an offender
> (I am trying to push a version without casting to the compile farm),
> the rest are unsigned long passed to the function which now
> (after my change) has const void * as argument?
> 
> Example:
> 
> > > @@ -90,7 +90,7 @@ static long cmm_alloc_pages(long nr, long *counter,
> 
> unsigned long addr;
> 
> > > +             diag10_range(virt_to_pfn((void *)addr), 1);

I only tried to say that these pieces weren't offenders before
you patch and turned ones after. But that seems like what your
commit message says.

> Yours,
> Linus Walleij

Thanks!