The lo,hi order is different from the comments. And in commit
1ec182c33379 ("target/arm: Convert to HAVE_CMPXCHG128"), it changes
the original code logic. So just restore the old code logic before this 
commit:
do_paired_cmpxchg64_be():
    cmpv = int128_make128(env->exclusive_high, env->exclusive_val);
    newv = int128_make128(new_hi, new_lo);

Fixes: 1ec182c33379 ("target/arm: Convert to HAVE_CMPXCHG128")

Signed-off-by: Catherine Ho <catherine.hecx@gmail.com>
---
I didn't see any obvious real error case here, so set it as RFC

 target/arm/helper-a64.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/target/arm/helper-a64.c b/target/arm/helper-a64.c
index 101fa6d3ea..70850e564d 100644
--- a/target/arm/helper-a64.c
+++ b/target/arm/helper-a64.c
@@ -583,8 +583,8 @@ uint64_t HELPER(paired_cmpxchg64_be)(CPUARMState *env, uint64_t addr,
      * High and low need to be switched here because this is not actually a
      * 128bit store but two doublewords stored consecutively
      */
-    Int128 cmpv = int128_make128(env->exclusive_val, env->exclusive_high);
-    Int128 newv = int128_make128(new_lo, new_hi);
+    Int128 cmpv = int128_make128(env->exclusive_high, env->exclusive_val);
+    Int128 newv = int128_make128(new_hi, new_lo);
     Int128 oldv;
     uintptr_t ra = GETPC();
     uint64_t o0, o1;
-- 
2.17.1

On 1/31/19 5:40 PM, Catherine Ho wrote:
> The lo,hi order is different from the comments. And in commit
> 1ec182c33379 ("target/arm: Convert to HAVE_CMPXCHG128"), it changes
> the original code logic. So just restore the old code logic before this 
> commit:
> do_paired_cmpxchg64_be():
>     cmpv = int128_make128(env->exclusive_high, env->exclusive_val);
>     newv = int128_make128(new_hi, new_lo);
> 
> Fixes: 1ec182c33379 ("target/arm: Convert to HAVE_CMPXCHG128")
> 
> Signed-off-by: Catherine Ho <catherine.hecx@gmail.com>
> ---


Reviewed-by: Richard Henderson <richard.henderson@linaro.org>


r~

On Mon, 11 Feb 2019 at 16:11, Richard Henderson
<richard.henderson@linaro.org> wrote:
>
> On 1/31/19 5:40 PM, Catherine Ho wrote:
> > The lo,hi order is different from the comments. And in commit
> > 1ec182c33379 ("target/arm: Convert to HAVE_CMPXCHG128"), it changes
> > the original code logic. So just restore the old code logic before this
> > commit:
> > do_paired_cmpxchg64_be():
> >     cmpv = int128_make128(env->exclusive_high, env->exclusive_val);
> >     newv = int128_make128(new_hi, new_lo);
> >
> > Fixes: 1ec182c33379 ("target/arm: Convert to HAVE_CMPXCHG128")
> >
> > Signed-off-by: Catherine Ho <catherine.hecx@gmail.com>
> > ---
>
>
> Reviewed-by: Richard Henderson <richard.henderson@linaro.org>

Thanks. Richard: is this effectively a bugfix for big-endian guest code ?

-- PMM

On 2/11/19 8:19 AM, Peter Maydell wrote:
> Thanks. Richard: is this effectively a bugfix for big-endian guest code ?

Yes.


r~

On Mon, 11 Feb 2019 at 16:49, Richard Henderson
<richard.henderson@linaro.org> wrote:
>
> On 2/11/19 8:19 AM, Peter Maydell wrote:
> > Thanks. Richard: is this effectively a bugfix for big-endian guest code ?
>
> Yes.

Thanks, applied to target-arm.next; I've added a line to the commit
message to note that it only affects BE guests.

-- PMM