diff options
| author | Boqun Feng <[email protected]> | 2015-11-02 09:30:31 +0800 |
|---|---|---|
| committer | Michael Ellerman <[email protected]> | 2015-12-14 20:38:18 +1100 |
| commit | 49e9cf3f0c04bf76ffa59242254110309554861d (patch) | |
| tree | a8dad7893126a02ebb18c88ea0d374502cf68beb /tools/perf/scripts/python | |
| parent | 4dcbd88eb600d52ce52a75c5075c2eff2f6849e6 (diff) | |
powerpc: Make value-returning atomics fully ordered
According to memory-barriers.txt:
> Any atomic operation that modifies some state in memory and returns
> information about the state (old or new) implies an SMP-conditional
> general memory barrier (smp_mb()) on each side of the actual
> operation ...
Which mean these operations should be fully ordered. However on PPC,
PPC_ATOMIC_ENTRY_BARRIER is the barrier before the actual operation,
which is currently "lwsync" if SMP=y. The leading "lwsync" can not
guarantee fully ordered atomics, according to Paul Mckenney:
https://lkml.org/lkml/2015/10/14/970
To fix this, we define PPC_ATOMIC_ENTRY_BARRIER as "sync" to guarantee
the fully-ordered semantics.
This also makes futex atomics fully ordered, which can avoid possible
memory ordering problems if userspace code relies on futex system call
for fully ordered semantics.
Fixes: b97021f85517 ("powerpc: Fix atomic_xxx_return barrier semantics")
Cc: [email protected] # 3.2+
Signed-off-by: Boqun Feng <[email protected]>
Reviewed-by: Paul E. McKenney <[email protected]>
Acked-by: Peter Zijlstra (Intel) <[email protected]>
Signed-off-by: Michael Ellerman <[email protected]>
Diffstat (limited to 'tools/perf/scripts/python')
0 files changed, 0 insertions, 0 deletions