[PATCH v3] Add single-threaded fast path to rand()

[Adhemerval Zanella Netto]

On 29/07/24 11:23, Wilco Dijkstra wrote:
> Improve performance of rand() and __random() by adding a single-threaded fast
> path.  Bench-random-lock shows about 5x speedup on Neoverse V1.

LGTM, although I am not sure which kind of workload does really
benefit from this optimization.

> 
> ---
> 
> diff --git a/stdlib/random.c b/stdlib/random.c
> index 62f22fac8d58c7977f09c134bf80a797750da645..174603a8915fd8aa4b3ae64d023003c9e2c038f2 100644
> --- a/stdlib/random.c
> +++ b/stdlib/random.c
> @@ -51,6 +51,7 @@
>     SUCH DAMAGE.*/
>  
>  #include <libc-lock.h>
> +#include <sys/single_threaded.h>
>  #include <limits.h>
>  #include <stddef.h>
>  #include <stdlib.h>
> @@ -288,6 +289,12 @@ __random (void)
>  {
>    int32_t retval;
>  
> +  if (SINGLE_THREAD_P)
> +    {
> +      (void) __random_r (&unsafe_state, &retval);
> +      return retval;
> +    }
> +
>    __libc_lock_lock (lock);
>  
>    (void) __random_r (&unsafe_state, &retval);