The current packed struct implementation of unaligned access adds
the packed attribute only to the field within the unaligned struct
rather than to the struct as a whole. This is not sufficient to
enforce proper behaviour on architectures with a default struct
alignment of more than one byte.

For example, the current implementation of __get_unaligned_cpu16
when compiled for arm with gcc -O1 -mstructure-size-boundary=32
assumes the struct is on a 4 byte boundary so performs the load
of the 16bit packed field as if it were on a 4 byte boundary:

__get_unaligned_cpu16:
        ldrh    r0, [r0, #0]
        bx      lr

Moving the packed attribute to the struct rather than the field
causes the proper unaligned access code to be generated:

__get_unaligned_cpu16:
	ldrb	r3, [r0, #0]	@ zero_extendqisi2
	ldrb	r0, [r0, #1]	@ zero_extendqisi2
	orr	r0, r3, r0, asl #8
	bx	lr

Signed-off-by: Will Newton <will.newton@gmail.com>
---
 include/linux/unaligned/packed_struct.h |    6 +++---
 1 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/include/linux/unaligned/packed_struct.h
b/include/linux/unaligned/packed_struct.h
index 2498bb9..c9a6abd 100644
--- a/include/linux/unaligned/packed_struct.h
+++ b/include/linux/unaligned/packed_struct.h
@@ -3,9 +3,9 @@

 #include <linux/kernel.h>

-struct __una_u16 { u16 x __attribute__((packed)); };
-struct __una_u32 { u32 x __attribute__((packed)); };
-struct __una_u64 { u64 x __attribute__((packed)); };
+struct __una_u16 { u16 x; } __attribute__((packed));
+struct __una_u32 { u32 x; } __attribute__((packed));
+struct __una_u64 { u64 x; } __attribute__((packed));

 static inline u16 __get_unaligned_cpu16(const void *p)
 {
-- 
1.7.0.4