Messages in this thread | | | From | "David Schwartz" <> | Subject | RE: usleep() (in)accuracy | Date | Sun, 13 Aug 2000 22:03:01 -0700 |
| |
Animesg Singh writes:
Very True , but the resolution between 3ms to 20ms varies wildly even with SCHED_RR, for eg. for 5ms delay the one obtained is anything between 20ms to 30ms depending on number of threads operational (more the number of threads more is the degardation) but even simple linux should provide predictable delay mechanism , so many application require it (eg RTP Servers where packet delay is in this range). Can We do something about this regards Animesh -
The low latency patches might help a little bit. It also might help to undersleep and then busy wait the rest of the sleep once you wake up. But the basic answer is, no, there's nothing you can do about it. Linux is not an RTOS.
DS
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |