diff options
| author |   jaszha03 <jason.zhang2@arm.com> | 2019-06-26 17:06:59 -0500 |
|---|---|---|
| committer |   Dave Barach <openvpp@barachs.net> | 2019-08-01 11:54:17 +0000 |
| commit | fb1ccc7c369d879a20567df609d04cfdcec5f8c1 (patch) | |
| tree | c2f289132da3c8e0a71c186df990270bf9c26043 /src/vppinfra/atomics.h | |
| parent | 635acbf4b76038023b0b5a9bdae0dc87e13fd308 (diff) | |
vppinfra: refactor clib_spinlock_t to use compare and swap
Tested performance of a CAS implementation (using __atomic_compare_exchange)
against a TAS implementation (using __atomic_exchange) using test_spinlock.c
and found some performance improvement.
Generated assembly for CAS and TAS implementations show that TAS always
executes with a load-store dependency, but CAS moves a branch condition
between the load and store so that only a load occurs when the lock is free.
Benchmarking:
The results below are the cycle counts from test_spinlock.c, configured so
that for 10000 iterations, 12 threads on separate cores are spawned, each of
which increments a global counter 10000 times in each iteration. For
A72, 8 threads are spawned in each test.
x86 Xeon TAS: 7.333e8, 7.605e8, 7.535e8, 7.485e8, 7.321e8
x86 Xeon CAS: 5.842e8, 5.433e8, 5.389e8, 5.983e8, 5.552e8
Aarch64 ThX2* TAS: 9.852e7, 10.209e7, 9.190e7, 9.600e7, 9.224e7
Aarch64 ThX2* CAS: 7.640e7, 7.486e7, 7.425e7, 7.269e7, 7.534e7
A72 TAS: 7.289e6, 6.963e6, 7.208e6, 6.976e6, 7.200e6
A72 CAS: 1.695e6, 1.608e6, 1.600e6, 1.634e6, 1.746e6
*ThunderX2 used additional gcc options "-march=armv8.1-a+crc+crypto+lse"
Type: refactor
Change-Id: Ic5cd97991804f6b012707fad1a5d1a6edb96cd3d
Signed-off-by: Jason Zhang <jason.zhang2@arm.com>
Reviewed-by: Honnappa Nagarahalli <honnappa.nagarahalli@arm.com>
Reviewed-by: Lijian Zhang <Lijian.Zhang@arm.com>
Diffstat (limited to 'src/vppinfra/atomics.h')
| -rw-r--r-- | src/vppinfra/atomics.h | 3 |
1 files changed, 3 insertions, 0 deletions
diff --git a/src/vppinfra/atomics.h b/src/vppinfra/atomics.h index df8e534b683..ce2177df5c2 100644 --- a/src/vppinfra/atomics.h +++ b/src/vppinfra/atomics.h @@ -37,11 +37,14 @@ #define clib_atomic_cmp_and_swap(addr,old,new) __sync_val_compare_and_swap(addr, old, new) #define clib_atomic_bool_cmp_and_swap(addr,old,new) __sync_bool_compare_and_swap(addr, old, new) +#define clib_atomic_cmp_and_swap_acq_relax_n(addr,exp,new,weak) __atomic_compare_exchange_n ((addr), (exp), (new), (weak), __ATOMIC_ACQUIRE, __ATOMIC_RELAXED) + #define clib_atomic_test_and_set(a) __atomic_exchange_n(a, 1, __ATOMIC_ACQUIRE) #define clib_atomic_release(a) __atomic_store_n(a, 0, __ATOMIC_RELEASE) #define clib_atomic_fence_rel() __atomic_thread_fence(__ATOMIC_RELEASE); +#define clib_atomic_load_relax_n(a) __atomic_load_n((a), __ATOMIC_RELAXED) #define clib_atomic_load_acq_n(a) __atomic_load_n((a), __ATOMIC_ACQUIRE) #define clib_atomic_store_rel_n(a, b) __atomic_store_n ((a), (b), __ATOMIC_RELEASE) |