>>> On 20.11.15 at 17:03, wrote:
> --- a/xen/common/spinlock.c
> +++ b/xen/common/spinlock.c
> @@ -10,6 +10,8 @@
> #include
> #include
>
> +DEFINE_PER_CPU(cpumask_t, percpu_rwlock_readers);
static (and perhaps even local to _percpu_write_lock()).
> @@ -492,6 +494,42 @@ int _rw_is_write_loc
On 20/11/15 16:03, Malcolm Crossley wrote:
> Per-cpu read-write locks allow for the fast path read case to have
> low overhead by only setting/clearing a per-cpu variable for using
> the read lock. The per-cpu read fast path also avoids locked
> compare swap operations which can be particularly slo
Per-cpu read-write locks allow for the fast path read case to have
low overhead by only setting/clearing a per-cpu variable for using
the read lock. The per-cpu read fast path also avoids locked
compare swap operations which can be particularly slow on coherent
multi-socket systems, particularly if
