Re: [PATCH] asm-generic: Fix build when __set_fixmap is absent

[Mark Rutland <mark.rutland@arm.com>]

On Tue, Feb 09, 2016 at 04:52:34PM +0000, Catalin Marinas wrote:
> On Tue, Feb 09, 2016 at 04:33:34PM +0000, Mark Rutland wrote:
> > > >> include/asm-generic/fixmap.h:72:6: error: conflicting types for '__set_fixmap'
> > >     void __set_fixmap(enum fixed_addresses idx, phys_addr_t phys, pgprot_t prot);
> > >          ^
> > >    In file included from arch/um/include/asm/pgtable.h:11:0,
> > >                     from include/linux/mm.h:67,
> > >                     from include/linux/ring_buffer.h:5,
> > >                     from include/linux/trace_events.h:5,
> > >                     from include/trace/syscall.h:6,
> > >                     from include/linux/syscalls.h:81,
> > >                     from init/main.c:18:
> > >    arch/um/include/asm/fixmap.h:39:13: note: previous declaration of '__set_fixmap' was here
> > >     extern void __set_fixmap (enum fixed_addresses idx,
> > >                 ^
> > 
> > The conflict is the type of 'phys'. In arch/um that's an unsigned long
> > rather than a phys_addr_t as it is elsewhere.
> 
> At a quick grep, we also have:
> 
> arch/sh/include/asm/fixmap.h
> arch/sh/mm/init.c
> arch/sh/mm/nommu.c
> 
> > If I convert that to a phys_addr_t the build goes along happily.
> > 
> > Should we change set_fixmap_offset back to a macro function for now, or
> > is it simple/correct to change arch/um to use phys_addr_t in
> > __set_fixmap?
> 
> And sh. I prefer the static inline, though there is more effort needed
> to test and get acks ;) (I really don't mind either way).

I would also prefer to make this a static inline, but it looks like we
need to sort out some cross-architecture cleanup first. I'm happy to
have a go at that.

In the meantime, so as to allow linux-next to build, and to save us from
merge hell, let's follow the usual idiom and hope that underscores will
protect us.

Hopefully this is the last time I ask this today: Arnd, Catalin, are you
happy with the below patch?

Mark.

---->8----
From 6d283603d18071690dc138e4a0591a445a1d1e30 Mon Sep 17 00:00:00 2001
From: Mark Rutland <mark.rutland@arm.com>
Date: Tue, 9 Feb 2016 17:08:26 +0000
Subject: [PATCH] asm-generic: make __set_fixmap_offset a macro again

Turning __set_fixmap_offset into a static inline breaks the build for
several architectures. Fixing this properly requires updates to a number
of architectures to make them agree on the prototype of __set_fixmap.

For the timebeing, restore __set_fixmap_offset to its prior state as a
macro function, reverting commit ac4c0ac73485867c ("asm-generic: make
__set_fixmap_offset a static inline"). To avoid the original issue with
namespace clashes, 'addr' is prefixed with a liberal sprinking of
underscores.

Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Catalin Marinas <catalin.marinas@arm.com>
---
 include/asm-generic/fixmap.h | 14 +++++++-------
 1 file changed, 7 insertions(+), 7 deletions(-)

diff --git a/include/asm-generic/fixmap.h b/include/asm-generic/fixmap.h
index f9c27b6..827e4d3 100644
--- a/include/asm-generic/fixmap.h
+++ b/include/asm-generic/fixmap.h
@@ -70,13 +70,13 @@ static inline unsigned long virt_to_fix(const unsigned long vaddr)
 #endif
 
 /* Return a pointer with offset calculated */
-static inline unsigned long __set_fixmap_offset(enum fixed_addresses idx,
-						phys_addr_t phys,
-						pgprot_t flags)
-{
-	__set_fixmap(idx, phys, flags);
-	return fix_to_virt(idx) + (phys & (PAGE_SIZE - 1));
-}
+#define __set_fixmap_offset(idx, phys, flags)				\
+({									\
+	unsigned long ________addr;					\
+	__set_fixmap(idx, phys, flags);					\
+	________addr = fix_to_virt(idx) + ((phys) & (PAGE_SIZE - 1));	\
+	________addr;							\
+})
 
 #define set_fixmap_offset(idx, phys) \
 	__set_fixmap_offset(idx, phys, FIXMAP_PAGE_NORMAL)
-- 
1.9.1