Sad Clouds <cryintotheblue...@googlemail.com> wrote:
> Well, quite often you need to call time function in a loop. For
> example, you have a loop that accepts network packets, or connections
> and every so often you need to have current system time. The resolution
> of the timer can be specific to application, i.e. it could be seconds,
> milliseconds, or microseconds.a
>
> uint64_t current_time;
> while(1)
> {
> get_network_data();
> current_time = update_current_time();
> ...
> use_current_time();
> }
>
> So in a tight loop, if update_current_time() calls clock_gettime() on
> every loop iteration, what would kill performance.
The advantage of using something like a time() syscall is that it could allow
the scheduler to give the CPU to another process. If you did implement a
mmap()ed time interface, this would not happen and the process would loop
in user space and thus be allowed to keep the CPU until the quantum is expired.
Jörg
--
EMail:jo...@schily.isdn.cs.tu-berlin.de (home) Jörg Schilling D-13353 Berlin
j...@cs.tu-berlin.de (uni)
joerg.schill...@fokus.fraunhofer.de (work) Blog:
http://schily.blogspot.com/
URL: http://cdrecord.berlios.de/private/ ftp://ftp.berlios.de/pub/schily
_______________________________________________
opensolaris-code mailing list
opensolaris-code@opensolaris.org
http://mail.opensolaris.org/mailman/listinfo/opensolaris-code
