[PATCH] [POWERPC v2] Add a comment about the Book-e tophys/tovirt macros

[PATCH] [POWERPC v2] Add a comment about the Book-e tophys/tovirt macros

[Kumar Gala]

The Book-E versions of tophys/tovirt are not simple noops to
ensure compatibility with their "classic" ppc brethren.

---

Fix my grammar.. why I bother is beyond me.

- k

 include/asm-powerpc/ppc_asm.h |    5 +++++
 1 files changed, 5 insertions(+), 0 deletions(-)

diff --git a/include/asm-powerpc/ppc_asm.h b/include/asm-powerpc/ppc_asm.h
index 2dbd4e7..668c887 100644
--- a/include/asm-powerpc/ppc_asm.h
+++ b/include/asm-powerpc/ppc_asm.h
@@ -356,6 +356,11 @@ END_FTR_SECTION_IFCLR(CPU_FTR_601)
 #define toreal(rd)
 #define fromreal(rd)

+/*
+ * We use addis to ensure compatiblity with the "classic" ppc versions
+ * of these macros that might have rs as 0, which means a value of 0
+ * not r0
+ */
 #define tophys(rd,rs)				\
 	addis	rd,rs,0

-- 
1.5.4.5

Re: [PATCH] [POWERPC v2] Add a comment about the Book-e tophys/tovirt macros

[Paul Mackerras]

Kumar Gala writes:

> +/*
> + * We use addis to ensure compatiblity with the "classic" ppc versions
> + * of these macros that might have rs as 0, which means a value of 0
> + * not r0
> + */
>  #define tophys(rd,rs)				\
>  	addis	rd,rs,0

Still seems clear as mud.  I think you're trying to say that some
callers use rs = 0 to get the tophys offset in rd, rather than
converting the address in r0, and so this version has to do that too
(i.e. set register rd to 0 when rs == 0)?

Paul.