Fix 'ALIGN()' macro, take 2

You wouldn't think that doing an ALIGN() macro that aligns something up
to a power-of-two boundary would be likely to have bugs, would you?

But hey, in the wonderful world of mixing integer types, you have to be
careful.  This just makes sure that the alignment is interpreted in the
same type as the thing to be aligned.

Thanks to Roland Dreier, who noticed that the amso1100 driver got broken
by the previous fix (that just extended the mask to "unsigned long", but
was still broken in "unsigned long long" - it just happened to be the
same on 64-bit architectures).

See commit 4c8bd7e for the history of
bugs here...

Acked-by: Roland Dreier <rdreier@cisco.com>
Cc: Andrew Morton <akpm@osdl.org>
Cc: David Miller <davem@davemloft.net>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

include/linux/kernel.h

@@ -30,8 +30,10 @@ extern const char linux_banner[];
 
 #define STACK_MAGIC	0xdeadbeef
 
+#define ALIGN(x,a)		__ALIGN_MASK(x,(typeof(x))(a)-1)
+#define __ALIGN_MASK(x,mask)	(((x)+(mask))&~(mask))
+
 #define ARRAY_SIZE(x) (sizeof(x) / sizeof((x)[0]))
-#define ALIGN(x,a) (((x)+(a)-1UL)&~((a)-1UL))
 #define FIELD_SIZEOF(t, f) (sizeof(((t*)0)->f))
 #define DIV_ROUND_UP(n,d) (((n) + (d) - 1) / (d))
 #define roundup(x, y) ((((x) + ((y) - 1)) / (y)) * (y))