Re: [Sched-tools] [PATCH 09/14] reorder the start sequence of threads

Hi,

On 20/03/15 08:45, pi-cheng.chen wrote:

From: Vincent Guittot vincent.guittot@linaro.org

Signed-off-by: Vincent Guittot vincent.guittot@linaro.org

src/rt-app.c | 27 ++++++++++++++------------- 1 file changed, 14 insertions(+), 13 deletions(-)

diff --git a/src/rt-app.c b/src/rt-app.c index 0c11c72..235c622 100644 --- a/src/rt-app.c +++ b/src/rt-app.c @@ -270,19 +270,6 @@ void *thread_body(void *arg) } }

• if (data->wait_before_start > 0) {

• log_notice("[%d] Waiting %ld usecs... ", data->ind,
  

• 	 data->wait_before_start);
  

• clock_gettime(CLOCK_MONOTONIC, &t_now);
  

• t_next = usec_to_timespec(data->wait_before_start);
  

• t_next = timespec_add(&t_now, &t_next);
  

• clock_nanosleep(CLOCK_MONOTONIC,
  

• 		TIMER_ABSTIME,
  

• 		&t_next,
  

• 		NULL);
  

• log_notice("[%d] Starting...", data->ind);
  

• }
• /* if we know the duration we can calculate how many periods we will
  • do at most, and the log to memory, instead of logging to file.
  */

@@ -299,6 +286,20 @@ void *thread_body(void *arg) "\t\tend\t\tdeadline\tdur.\tslack" "\tBudget\tUsed Budget\n");

• if (data->wait_before_start > 0) {

• log_notice("[%d] Waiting %ld usecs... ", data->ind,
  

• 	 data->wait_before_start);
  

• clock_gettime(CLOCK_MONOTONIC, &t_now);
  

• t_next = usec_to_timespec(data->wait_before_start);
  

• t_next = timespec_add(&t_now, &t_next);
  

• clock_nanosleep(CLOCK_MONOTONIC,
  

• 		TIMER_ABSTIME,
  

• 		&t_next,
  

• 		NULL);
  

• log_notice("[%d] Starting...", data->ind);
  

• }

#ifdef DLSCHED /* TODO find a better way to handle that constraint */ /*

Can you explain in the changelog why we need this fix and what it changes?

Thanks,

- Juri