[ofa-general] Problem while running ib tests

[Jason Gunthorpe]

On Wed, Sep 23, 2009 at 06:43:12PM +0300, Moni Shoua wrote:

> In the man page for clock_gettime() I see that
> CLOCK_PROCESS_CPUTIME_ID may be more accurate for (high) performance
> measurements. What do you think?

The CPUTIME counters are of the 'time spent running code' variety,
they do not time absolute time and they are executed in the
kernel. CLOCK_MONOTONIC is the closest thing to tsc 

The man page is a bit misleading.. See the SUSv3 spec:
https://www.opengroup.org/onlinepubs/000095399/functions/clock_getres.html

Jason