diff options
| author | Shell <[email protected]> | 2024-05-16 15:45:12 +0800 |
|---|---|---|
| committer | GitHub <[email protected]> | 2024-05-16 15:45:12 +0800 |
| commit | e25fc8b5118fba87acd5bc677641926a913bb3a5 (patch) | |
| tree | d86b1be0152b2d3f360c83a3922d945f565ecaf0 /src/cpu_mp.c | |
| parent | e46333496f880afe7ae2864967406f758aec55a7 (diff) | |
[libcpu/arm64] add C11 atomic ticket spinlock (#8882)
* [libcpu/arm64] add C11 atomic ticket spinlock
Replace the former implementation of flag-based spinlock which is unfair
Besides, C11 atomic implementation is more readable (it's C anyway),
and maintainable. Cause toolchain can use their builtin optimization and
tune for different micro-architectures. For example armv8.5 introduces a
better instruction. The compiler can help with that when it knows your
target platform in support of it.
Signed-off-by: Shell <[email protected]>
* fixup: RT_CPUS_NR
---------
Signed-off-by: Shell <[email protected]>
Diffstat (limited to 'src/cpu_mp.c')
| -rw-r--r-- | src/cpu_mp.c | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/src/cpu_mp.c b/src/cpu_mp.c index 238ddb420b..a10e35f479 100644 --- a/src/cpu_mp.c +++ b/src/cpu_mp.c @@ -146,7 +146,7 @@ rt_base_t rt_cpus_lock(void) pcpu = rt_cpu_self(); if (pcpu->current_thread != RT_NULL) { - register rt_ubase_t lock_nest = rt_atomic_load(&(pcpu->current_thread->cpus_lock_nest)); + rt_ubase_t lock_nest = rt_atomic_load(&(pcpu->current_thread->cpus_lock_nest)); rt_atomic_add(&(pcpu->current_thread->cpus_lock_nest), 1); if (lock_nest == 0) |