On Tue, 15 Jan 2019, Sebastian Andrzej Siewior wrote:
> There two sleep functions which ensure that the forked function sleeps
> and does not spin until everythin is setup. There is no need for that,
> the scheduler will take care of that anyway. Also that sleep may
> complete before or after the testcase starts. If it completes afterwards
> then the testcase waits to start so…
>
> Remove the sleep, it does change the outcome of the testcase.
>
> Signed-off-by: Sebastian Andrzej Siewior <bigeasy@xxxxxxxxxxxxx>
> ---
> src/ssdd/ssdd.c | 9 +--------
> 1 file changed, 1 insertion(+), 8 deletions(-)
>
> diff --git a/src/ssdd/ssdd.c b/src/ssdd/ssdd.c
> index 9b85dfa9d2236..68a426266dd90 100644
> --- a/src/ssdd/ssdd.c
> +++ b/src/ssdd/ssdd.c
> @@ -132,12 +132,8 @@ static void sigchld(int sig, unused siginfo_t * info, unused void *arg)
>
> static void child_process(void)
> {
> - unused volatile int i;
> -
> - /* wait for ptrace attach */
> - usleep(100000);
> while (1)
> - i = 0;
> + ;
> }
>
> static int forktests(int testid)
> @@ -170,9 +166,6 @@ static int forktests(int testid)
> exit(1);
> }
>
> - /* give both our child and parent time to set things up */
> - usleep(125000);
> -
> /*
> * Attach to the child.
> */
> --
> 2.20.1
>
>
Signed-off-by: John Kacur <jkacur@xxxxxxxxxx>
