On 08/16/2017 03:14 PM, Steven Rostedt wrote:

On Wed, 16 Aug 2017 15:07:25 -0400
Waiman Long [off-list ref] wrote:

quoted

I am thinking about maybe letting a RT task to sleep a tiny amount of
time instead of calling schedule(). Hopefully, that will allow a
lower-priority task to make forward progress.

That is actually one of the ugly hacks we have in the PREEMPT_RT patch
(and one of the things we need to fix before getting it to mainline).

A simple msleep(1).

-- Steve

I am planning to use usleep_range(). That will allow a short sleep in
the us range instead of ms. The documentation suggest that for a
10us-20ms sleep.

Cheers,
Longman