kernel-ark/arch/x86/lib
Andi Kleen e0a96129db x86: use early clobbers in usercopy*.c
Impact: fix rare (but currently harmless) miscompile with certain configs and gcc versions

Hugh Dickins noticed that strncpy_from_user() was miscompiled
in some circumstances with gcc 4.3.

Thanks to Hugh's excellent analysis it was easy to track down.

Hugh writes:

> Try building an x86_64 defconfig 2.6.29-rc1 kernel tree,
> except not quite defconfig, switch CONFIG_PREEMPT_NONE=y
> and CONFIG_PREEMPT_VOLUNTARY off (because it expands a
> might_fault() there, which hides the issue): using a
> gcc 4.3.2 (I've checked both openSUSE 11.1 and Fedora 10).
>
> It generates the following:
>
> 0000000000000000 <__strncpy_from_user>:
>    0:   48 89 d1                mov    %rdx,%rcx
>    3:   48 85 c9                test   %rcx,%rcx
>    6:   74 0e                   je     16 <__strncpy_from_user+0x16>
>    8:   ac                      lods   %ds:(%rsi),%al
>    9:   aa                      stos   %al,%es:(%rdi)
>    a:   84 c0                   test   %al,%al
>    c:   74 05                   je     13 <__strncpy_from_user+0x13>
>    e:   48 ff c9                dec    %rcx
>   11:   75 f5                   jne    8 <__strncpy_from_user+0x8>
>   13:   48 29 c9                sub    %rcx,%rcx
>   16:   48 89 c8                mov    %rcx,%rax
>   19:   c3                      retq
>
> Observe that "sub %rcx,%rcx; mov %rcx,%rax", whereas gcc 4.2.1
> (and many other configs) say "sub %rcx,%rdx; mov %rdx,%rax".
> Isn't it returning 0 when it ought to be returning strlen?

The asm constraints for the strncpy_from_user() result were missing an
early clobber, which tells gcc that the last output arguments
are written before all input arguments are read.

Also add more early clobbers in the rest of the file and fix 32-bit
usercopy.c in the same way.

Signed-off-by: Andi Kleen <ak@linux.intel.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
[ since this API is rarely used and no in-kernel user relies on a 'len'
  return value (they only rely on negative return values) this miscompile
  was never noticed in the field. But it's worth fixing it nevertheless. ]
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-01-21 09:43:17 +01:00
..
checksum_32.S
clear_page_64.S
copy_page_64.S
copy_user_64.S x86: wrong register was used in align macro 2008-07-30 10:10:39 -07:00
copy_user_nocache_64.S x86: wrong register was used in align macro 2008-07-30 10:10:39 -07:00
csum-copy_64.S
csum-partial_64.c x86: fix csum_partial() export 2008-05-13 19:38:47 +02:00
csum-wrappers_64.c
delay.c x86: integrate delay functions. 2008-07-09 08:52:05 +02:00
getuser.S x86: merge getuser asm functions. 2008-07-09 09:14:05 +02:00
io_64.c
iomap_copy_64.S
Makefile x86: drop -funroll-loops for csum_partial_64.c 2008-09-04 08:42:06 -07:00
memcpy_32.c
memcpy_64.S
memmove_64.c
memset_64.S
mmx_32.c
msr-on-cpu.c x86: msr-on-cpu: remove unnecessary level of abstraction 2008-08-25 22:45:50 -07:00
putuser.S x86: merge putuser asm functions. 2008-07-09 09:14:13 +02:00
rwlock_64.S
semaphore_32.S
string_32.c x86: coding style fixes to arch/x86/lib/string_32.c 2008-08-15 16:53:25 +02:00
strstr_32.c x86: coding style fixes to arch/x86/lib/strstr_32.c 2008-08-15 16:53:24 +02:00
thunk_32.S ftrace: trace irq disabled critical timings 2008-05-23 20:32:46 +02:00
thunk_64.S ftrace: trace irq disabled critical timings 2008-05-23 20:32:46 +02:00
usercopy_32.c x86: use early clobbers in usercopy*.c 2009-01-21 09:43:17 +01:00
usercopy_64.c x86: use early clobbers in usercopy*.c 2009-01-21 09:43:17 +01:00