[cgl_discussion] Some POSIX Timers testing data

Selbak, Rolla N rolla.n.selbak at intel.com
Mon Jan 20 14:49:31 PST 2003


These tests look really great Julie, thanks!


* my views are not necessarily my employer's *


-----Original Message-----
From: Fleischer, Julie N [mailto:julie.n.fleischer at intel.com]
Sent: Monday, January 20, 2003 3:05 PM
To: 'rddunlap at osdl.org'
Cc: 'cgl_discussion at osdl.org'
Subject: [cgl_discussion] Some POSIX Timers testing data


Randy/CGL -
On the POSIX Test Suite website, we've published our results of running the
latest POSIX conformance Timers tests against the two different Timers
implementations (high-res-timers and the alternate POSIX timers).  If you
are interested, you can see them at URL:
http://posixtest.sourceforge.net/testpass.html.

These tests are conformance tests designed to test that the lines of the
POSIX System Interfaces specification applying to Timers are implemented.
This is just a small set of the eventual POSIX Test Timers suite, which
would be one piece of the evaluation criteria CGL might use to evaluate a
distribution.  The final suite will include additional conformance tests
(for process CPU Time plus any test case enhancements made), functional
tests (to test not just conformance, but full functionality), and
performance and stress tests (which will show how well the distributions
handle stress conditions).  I would also imagine the evaluation criteria
would involve prioritizing which of these areas is the most important to CGL
and then examining the data with that priority ordering in mind.

So, these results definitely do not give a full picture comparison (There's
still quite a bit of work to do there.), but they show the progress made so
far.  Next steps are functional and stress tests of Timers.

Thanks.
- Julie

**These views are not necessarily those of my employer.**
_______________________________________________
cgl_discussion mailing list
cgl_discussion at lists.osdl.org
http://lists.osdl.org/mailman/listinfo/cgl_discussion



More information about the cgl_discussion mailing list