On Mon, 9 Mar 2015, Markos Chandras wrote: > > Since this operates on addresses shouldn't PTR_ADDIU be used instead? > > > > Maciej > > > > I don't know. I thought PTR_ADDIU should be used for pointers but the > arguments in these macros are "unsigned long". Hmm, good point. I think we should match the C data type used even though we have an assumption that sizeof(long) == sizeof(void *), so your change looks right to me as it stands. I think we have a convention to separate `linux' from `asm' inclusions by an empty line though, so I suggest that you add one here: > diff --git a/arch/mips/include/asm/r4kcache.h b/arch/mips/include/asm/r4kcache.h > index 1b22d2da88a1..d329f7328bd4 100644 > --- a/arch/mips/include/asm/r4kcache.h > +++ b/arch/mips/include/asm/r4kcache.h > @@ -12,6 +12,7 @@ > #ifndef _ASM_R4KCACHE_H > #define _ASM_R4KCACHE_H > > +#include <linux/stringify.h> > #include <asm/asm.h> > #include <asm/cacheops.h> > #include <asm/compiler.h> as well. I can offer you my review tag if you repost the change with this trivial update. Maciej
